TUD-MMC_wilma.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluation of Results of \"TUD-MMC at MediaEval 2016: Context of Experience task\" by Wang & Liem"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Necessary imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os.path\n",
    "import xml.etree.ElementTree as ET\n",
    "\n",
    "train_path = \"res/coe_dataset_icpr/dev_set/\"\n",
    "test_path = \"res/coe_dataset_icpr/test_set/\"\n",
    "\n",
    "audio_folder = \"audio_descriptors/\"\n",
    "text_folder = \"text_descriptors/\"\n",
    "vis_folder = \"vis_descriptors/\"\n",
    "metadata_folder = \"XML/\"\n",
    "\n",
    "train_entries_path = \"res/CoeTraining.csv\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Datasets\n",
    "\n",
    "Features are built in the manner described in the paper of Wang & Liem or \"Right Inflight? A Dataset for Exploring the Automatic\n",
    "Prediction of Movies Suitable for a Watching Situation\" (https://mmsys2016.itec.aau.at/papers/MMSYS/a45-riegler.pdf), if Wang & Liem do not provide any information.\n",
    "\n",
    "This leads to following set-up:\n",
    "\n",
    "Metadata: (language, year published, genre, country, runtime and age rating) - from XML -- 1-Hot Encoding for all categorical values<br>\n",
    "Text: as is td-idf <br>\n",
    "Audio: Averaged of all Frames (NaN to 0) - Mel-Frequency Cepstral Coefficients<br>\n",
    "Visual: as is - Histogram of Oriented Gradients (HOG) gray, Color Moments, local binary patterns (LBP) and Gray Level Run Length Matrix\n",
    "\n",
    "NOTE: Training data - invalid entry (2_states, also in test set), (Moulin_Rouge!.mp4, should be Moulin_Rouge! --> fixed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movie_name</th>\n",
       "      <th>file_name</th>\n",
       "      <th>goodforairplanes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>A Fish Called Wanda</td>\n",
       "      <td>A_Fish_Called_Wanda</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A Goofy Movie</td>\n",
       "      <td>A_Goofy_Movie</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A Million Ways to Die in the West</td>\n",
       "      <td>A_Million_Ways_to_Die_in_the_West</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>A Single Man</td>\n",
       "      <td>A_Single_Man</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>American Gangster</td>\n",
       "      <td>American_Gangster</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                          movie_name                          file_name  \\\n",
       "0                A Fish Called Wanda                A_Fish_Called_Wanda   \n",
       "1                      A Goofy Movie                      A_Goofy_Movie   \n",
       "2  A Million Ways to Die in the West  A_Million_Ways_to_Die_in_the_West   \n",
       "3                       A Single Man                       A_Single_Man   \n",
       "4                  American Gangster                  American_Gangster   \n",
       "\n",
       "  goodforairplanes  \n",
       "0                1  \n",
       "1                0  \n",
       "2                1  \n",
       "3                1  \n",
       "4                1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_base_train = pd.read_csv(train_entries_path)\n",
    "df_base_train = df_base_train[df_base_train['file_name'] != '2_States'] # remove invalid entry\n",
    "df_base_train.sort_values(by='file_name', inplace=True)\n",
    "df_base_train.reset_index(inplace=True, drop=True)\n",
    "df_targets_train = df_base_train['goodforairplanes'].astype(int)\n",
    "df_base_train.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature extractors\n",
    "As the dataset was built in a manner that would have been considered dirty already in 2002 a lot of feature extraction is done"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_audio_features(file_name, use_train=True):\n",
    "    \"\"\"\n",
    "        returns 1x14 dataframe, with averaged Mel-Frequency Cepstral Coefficients + file_name\n",
    "    \"\"\"\n",
    "    base_path = train_path if use_train else test_path\n",
    "    file_path = os.path.join(base_path, audio_folder, file_name + \".csv\")\n",
    "    if not os.path.isfile(file_path):\n",
    "        print(file_name, \" does not exist!\")\n",
    "        return pd.DataFrame(columns=[str(x) for x in range(13)] + ['file_name'])\n",
    "    df_audio = pd.read_csv(file_path, header=None).T # transpose (columns are rows)\n",
    "    df_audio = df_audio.fillna(0) # nan values are treated as 0\n",
    "    df_audio = pd.DataFrame(df_audio.mean(axis=0)).T # average accross columns\n",
    "    df_audio['file_name'] = file_name\n",
    "    return df_audio\n",
    "\n",
    "def get_all_audio_features(df, use_train=True):\n",
    "    \"\"\"\n",
    "        returns nx14 dataframe, containing audio features for all movies\n",
    "    \"\"\"\n",
    "    dfs = []\n",
    "    for file_name in df['file_name']:\n",
    "        dfs.append(get_audio_features(file_name, use_train))\n",
    "    \n",
    "    return pd.concat(dfs).reset_index(drop=True)\n",
    "\n",
    "def get_all_text_features(df, use_train=True):\n",
    "    \"\"\"\n",
    "        returns nx3284 dataframe, containing tf-idf features for all movies\n",
    "        the dataset creators messed up - contains several terms multiple times\n",
    "        ordered alphabetically (?) - Live_Nude_Girls and Transformers__Age_of_Extinction where switched (detected perchance)\n",
    "    \"\"\"\n",
    "    base_path = train_path if use_train else test_path\n",
    "    file_path = os.path.join(base_path, text_folder, \"tdf_idf_dev.csv\")\n",
    "    df_txt = pd.read_csv(file_path)\n",
    "    # the creators of the dataset missed how csv-files work - so we transpose and drop empty rows to get the correct format\n",
    "    cols = df_txt.columns \n",
    "    df_txt = df_txt.T.dropna()\n",
    "    df_txt.columns = cols\n",
    "    df_txt.reset_index(inplace=True, drop=True)\n",
    "    df_txt['file_name'] = sorted(df['file_name']) # we assume the info to be order alphabetically, as we do not have more info\n",
    "    return df_txt.reset_index(drop=True)\n",
    "\n",
    "def get_vis_features(file_name, use_train=True):\n",
    "    \"\"\"\n",
    "        returns 1x1653 dataframe, with unspecified visual features + file_name\n",
    "        we assume that every single value in the csv is one feature\n",
    "        this may be wrong, as there are two rows and no documentation (again)\n",
    "    \"\"\"\n",
    "    base_path = train_path if use_train else test_path\n",
    "    file_path = os.path.join(base_path, vis_folder, file_name + \".csv\")\n",
    "    if not os.path.isfile(file_path):\n",
    "        print(file_name, \" does not exist!\")\n",
    "        return pd.DataFrame(columns=[str(x) for x in range(1652)] + ['file_name'])\n",
    "    df_vis = pd.read_csv(file_path, header=None)\n",
    "    df_vis = pd.DataFrame(pd.concat([df_vis.loc[0,:], df_vis.loc[1,:]])).reset_index(drop=True).T # treat each value as single feature (-> no aggregation)\n",
    "    df_vis['file_name'] = file_name\n",
    "    return df_vis\n",
    "\n",
    "def get_all_vis_features(df, use_train=True):\n",
    "    \"\"\"\n",
    "        returns nx1653 dataframe, containing visual features for all movies\n",
    "    \"\"\"\n",
    "    dfs = []\n",
    "    for file_name in df['file_name']:\n",
    "        dfs.append(get_vis_features(file_name, use_train))\n",
    "    \n",
    "    return pd.concat(dfs).reset_index(drop=True)\n",
    "\n",
    "def get_meta_features(file_name, use_train=True):\n",
    "    \"\"\"\n",
    "        returns 1x7 dataframe, with metadata features + file_name\n",
    "        One Hot Encoding is not applied here, this should happen later\n",
    "    \"\"\"\n",
    "    base_path = train_path if use_train else test_path\n",
    "    file_path = os.path.join(base_path, metadata_folder, file_name + \".xml\")\n",
    "    if not os.path.isfile(file_path):\n",
    "        print(file_name, \" does not exist!\")\n",
    "        return pd.DataFrame(columns=['country', 'genre', 'language', 'rated', 'runtime', 'year', 'file_name'])\n",
    "    etree = ET.parse(file_path)\n",
    "    movie = etree.getroot().find('movie')\n",
    "    mv = {}\n",
    "    mv['language'] = [movie.get('language')]\n",
    "    mv['year'] = [int(movie.get('year'))]\n",
    "    mv['genre'] = [movie.get('genre')]\n",
    "    mv['country'] = [movie.get('country')]\n",
    "    mv['runtime'] = [int(movie.get('runtime')[:-4])]\n",
    "    mv['rated'] = [movie.get('rated')]\n",
    "\n",
    "    df_meta = pd.DataFrame.from_dict(mv)\n",
    "    df_meta['file_name'] = file_name\n",
    "    \n",
    "    return df_meta\n",
    "\n",
    "def get_all_meta_features(df, use_train=True):\n",
    "    \"\"\"\n",
    "        returns nx7 dataframe, containing metadata features for all movies\n",
    "    \"\"\"\n",
    "    dfs = []\n",
    "    for file_name in df['file_name']:\n",
    "        dfs.append(get_meta_features(file_name, use_train))\n",
    "    \n",
    "    df_meta = pd.concat(dfs)\n",
    "    \n",
    "    df_country = df_meta.country.str.replace(' ','').str.get_dummies(sep=',')\n",
    "    df_country.columns = ['country_' + x for x in df_country.columns]\n",
    "\n",
    "    df_genre = df_meta.genre.str.replace(' ','').str.get_dummies(sep=',')\n",
    "    df_genre.columns = ['genre_' + x for x in df_genre.columns]\n",
    "\n",
    "    df_language = df_meta.language.str.replace(' ','').str.get_dummies(sep=',')\n",
    "    df_language.columns = ['language_' + x for x in df_language.columns]\n",
    "\n",
    "    df_rated = df_meta.rated.str.get_dummies(sep=',')\n",
    "    df_rated.columns = ['rated_' + x for x in df_rated.columns]\n",
    "    \n",
    "    return pd.concat([df_country, df_genre, df_language, df_rated, df_meta[['runtime', 'year', 'file_name']]], axis=1).reset_index(drop=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>file_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>66.828814</td>\n",
       "      <td>-5.674521</td>\n",
       "      <td>1.670346</td>\n",
       "      <td>1.143263</td>\n",
       "      <td>-0.635255</td>\n",
       "      <td>1.269376</td>\n",
       "      <td>0.633811</td>\n",
       "      <td>0.012407</td>\n",
       "      <td>0.164257</td>\n",
       "      <td>-1.494813</td>\n",
       "      <td>-0.292583</td>\n",
       "      <td>0.086875</td>\n",
       "      <td>-0.914582</td>\n",
       "      <td>A_Fish_Called_Wanda</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>43.908715</td>\n",
       "      <td>-6.380989</td>\n",
       "      <td>1.361790</td>\n",
       "      <td>-1.483354</td>\n",
       "      <td>-0.670040</td>\n",
       "      <td>-1.626537</td>\n",
       "      <td>0.466197</td>\n",
       "      <td>-1.888169</td>\n",
       "      <td>0.848654</td>\n",
       "      <td>-0.990286</td>\n",
       "      <td>0.673634</td>\n",
       "      <td>-0.972964</td>\n",
       "      <td>-0.141539</td>\n",
       "      <td>A_Goofy_Movie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3.390978</td>\n",
       "      <td>-6.725758</td>\n",
       "      <td>0.579762</td>\n",
       "      <td>-0.271885</td>\n",
       "      <td>-0.175640</td>\n",
       "      <td>-0.845690</td>\n",
       "      <td>-0.699064</td>\n",
       "      <td>-0.578434</td>\n",
       "      <td>0.537249</td>\n",
       "      <td>-1.387373</td>\n",
       "      <td>0.747223</td>\n",
       "      <td>-0.887580</td>\n",
       "      <td>-0.205273</td>\n",
       "      <td>A_Million_Ways_to_Die_in_the_West</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>57.743484</td>\n",
       "      <td>-3.722123</td>\n",
       "      <td>2.780418</td>\n",
       "      <td>0.756402</td>\n",
       "      <td>0.043743</td>\n",
       "      <td>-0.960622</td>\n",
       "      <td>-0.435575</td>\n",
       "      <td>-0.176729</td>\n",
       "      <td>1.665236</td>\n",
       "      <td>-2.068548</td>\n",
       "      <td>1.211791</td>\n",
       "      <td>-0.358194</td>\n",
       "      <td>0.738827</td>\n",
       "      <td>A_Single_Man</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>65.354709</td>\n",
       "      <td>-5.609515</td>\n",
       "      <td>-1.303409</td>\n",
       "      <td>-0.831993</td>\n",
       "      <td>-0.518848</td>\n",
       "      <td>-0.019373</td>\n",
       "      <td>-0.500203</td>\n",
       "      <td>-0.897985</td>\n",
       "      <td>0.148561</td>\n",
       "      <td>-0.666728</td>\n",
       "      <td>0.033135</td>\n",
       "      <td>0.383797</td>\n",
       "      <td>-0.209412</td>\n",
       "      <td>American_Gangster</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0         1         2         3         4         5         6  \\\n",
       "0  66.828814 -5.674521  1.670346  1.143263 -0.635255  1.269376  0.633811   \n",
       "1  43.908715 -6.380989  1.361790 -1.483354 -0.670040 -1.626537  0.466197   \n",
       "2   3.390978 -6.725758  0.579762 -0.271885 -0.175640 -0.845690 -0.699064   \n",
       "3  57.743484 -3.722123  2.780418  0.756402  0.043743 -0.960622 -0.435575   \n",
       "4  65.354709 -5.609515 -1.303409 -0.831993 -0.518848 -0.019373 -0.500203   \n",
       "\n",
       "          7         8         9        10        11        12  \\\n",
       "0  0.012407  0.164257 -1.494813 -0.292583  0.086875 -0.914582   \n",
       "1 -1.888169  0.848654 -0.990286  0.673634 -0.972964 -0.141539   \n",
       "2 -0.578434  0.537249 -1.387373  0.747223 -0.887580 -0.205273   \n",
       "3 -0.176729  1.665236 -2.068548  1.211791 -0.358194  0.738827   \n",
       "4 -0.897985  0.148561 -0.666728  0.033135  0.383797 -0.209412   \n",
       "\n",
       "                           file_name  \n",
       "0                A_Fish_Called_Wanda  \n",
       "1                      A_Goofy_Movie  \n",
       "2  A_Million_Ways_to_Die_in_the_West  \n",
       "3                       A_Single_Man  \n",
       "4                  American_Gangster  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_audio_train = get_all_audio_features(df_base_train)\n",
    "df_audio_train.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>24000</th>\n",
       "      <th>baby</th>\n",
       "      <th>baseball</th>\n",
       "      <th>big</th>\n",
       "      <th>doc</th>\n",
       "      <th>escort</th>\n",
       "      <th>frozen</th>\n",
       "      <th>heroes</th>\n",
       "      <th>high</th>\n",
       "      <th>huck</th>\n",
       "      <th>...</th>\n",
       "      <th>york</th>\n",
       "      <th>yorks</th>\n",
       "      <th>young</th>\n",
       "      <th>young.1</th>\n",
       "      <th>younger</th>\n",
       "      <th>youngja</th>\n",
       "      <th>zebra</th>\n",
       "      <th>zellweger</th>\n",
       "      <th>zoologists</th>\n",
       "      <th>file_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>A_Fish_Called_Wanda</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>A_Goofy_Movie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>A_Million_Ways_to_Die_in_the_West</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>A_Single_Man</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.051657</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>American_Gangster</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 3284 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   24000  baby  baseball  big  doc  escort  frozen  heroes  high  huck  \\\n",
       "0    0.0   0.0       0.0  0.0  0.0     0.0     0.0     0.0   0.0   0.0   \n",
       "1    0.0   0.0       0.0  0.0  0.0     0.0     0.0     0.0   0.0   0.0   \n",
       "2    0.0   0.0       0.0  0.0  0.0     0.0     0.0     0.0   0.0   0.0   \n",
       "3    0.0   0.0       0.0  0.0  0.0     0.0     0.0     0.0   0.0   0.0   \n",
       "4    0.0   0.0       0.0  0.0  0.0     0.0     0.0     0.0   0.0   0.0   \n",
       "\n",
       "                 ...                      york  yorks  young  young.1  \\\n",
       "0                ...                  0.000000    0.0    0.0      0.0   \n",
       "1                ...                  0.000000    0.0    0.0      0.0   \n",
       "2                ...                  0.000000    0.0    0.0      0.0   \n",
       "3                ...                  0.000000    0.0    0.0      0.0   \n",
       "4                ...                  0.051657    0.0    0.0      0.0   \n",
       "\n",
       "   younger  youngja  zebra  zellweger  zoologists  \\\n",
       "0      0.0      0.0    0.0        0.0         0.0   \n",
       "1      0.0      0.0    0.0        0.0         0.0   \n",
       "2      0.0      0.0    0.0        0.0         0.0   \n",
       "3      0.0      0.0    0.0        0.0         0.0   \n",
       "4      0.0      0.0    0.0        0.0         0.0   \n",
       "\n",
       "                           file_name  \n",
       "0                A_Fish_Called_Wanda  \n",
       "1                      A_Goofy_Movie  \n",
       "2  A_Million_Ways_to_Die_in_the_West  \n",
       "3                       A_Single_Man  \n",
       "4                  American_Gangster  \n",
       "\n",
       "[5 rows x 3284 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_txt_train = get_all_text_features(df_base_train)\n",
    "df_txt_train.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>1643</th>\n",
       "      <th>1644</th>\n",
       "      <th>1645</th>\n",
       "      <th>1646</th>\n",
       "      <th>1647</th>\n",
       "      <th>1648</th>\n",
       "      <th>1649</th>\n",
       "      <th>1650</th>\n",
       "      <th>1651</th>\n",
       "      <th>file_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.430310</td>\n",
       "      <td>0.38101</td>\n",
       "      <td>0.34082</td>\n",
       "      <td>0.31642</td>\n",
       "      <td>0.414650</td>\n",
       "      <td>0.385990</td>\n",
       "      <td>0.329380</td>\n",
       "      <td>0.31212</td>\n",
       "      <td>0.350670</td>\n",
       "      <td>0.342460</td>\n",
       "      <td>...</td>\n",
       "      <td>362.83</td>\n",
       "      <td>8.592300</td>\n",
       "      <td>9.142700</td>\n",
       "      <td>8.410100</td>\n",
       "      <td>8.792400</td>\n",
       "      <td>1483.3</td>\n",
       "      <td>417.21</td>\n",
       "      <td>892.59</td>\n",
       "      <td>435.28</td>\n",
       "      <td>A_Fish_Called_Wanda</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.002031</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.07302</td>\n",
       "      <td>0.027533</td>\n",
       "      <td>0.005346</td>\n",
       "      <td>0.006015</td>\n",
       "      <td>0.11824</td>\n",
       "      <td>0.026991</td>\n",
       "      <td>0.005171</td>\n",
       "      <td>...</td>\n",
       "      <td>20278.00</td>\n",
       "      <td>0.972010</td>\n",
       "      <td>1.365400</td>\n",
       "      <td>1.803200</td>\n",
       "      <td>1.463400</td>\n",
       "      <td>168740.0</td>\n",
       "      <td>20896.00</td>\n",
       "      <td>34434.00</td>\n",
       "      <td>20967.00</td>\n",
       "      <td>A_Goofy_Movie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>119950.00</td>\n",
       "      <td>0.000001</td>\n",
       "      <td>0.002466</td>\n",
       "      <td>0.000004</td>\n",
       "      <td>0.002466</td>\n",
       "      <td>729320.0</td>\n",
       "      <td>119950.00</td>\n",
       "      <td>230400.00</td>\n",
       "      <td>119950.00</td>\n",
       "      <td>A_Million_Ways_to_Die_in_the_West</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>9550.70</td>\n",
       "      <td>5.943300</td>\n",
       "      <td>6.021500</td>\n",
       "      <td>4.383900</td>\n",
       "      <td>5.428800</td>\n",
       "      <td>32242.0</td>\n",
       "      <td>10313.00</td>\n",
       "      <td>22308.00</td>\n",
       "      <td>9850.00</td>\n",
       "      <td>A_Single_Man</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.231640</td>\n",
       "      <td>0.28629</td>\n",
       "      <td>0.30068</td>\n",
       "      <td>0.28118</td>\n",
       "      <td>0.244890</td>\n",
       "      <td>0.278490</td>\n",
       "      <td>0.290760</td>\n",
       "      <td>0.29380</td>\n",
       "      <td>0.161040</td>\n",
       "      <td>0.166850</td>\n",
       "      <td>...</td>\n",
       "      <td>109.62</td>\n",
       "      <td>17.237000</td>\n",
       "      <td>15.543000</td>\n",
       "      <td>13.671000</td>\n",
       "      <td>15.231000</td>\n",
       "      <td>53559.0</td>\n",
       "      <td>8637.40</td>\n",
       "      <td>18597.00</td>\n",
       "      <td>8679.00</td>\n",
       "      <td>American_Gangster</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 1653 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          0        1        2        3         4         5         6        7  \\\n",
       "0  0.430310  0.38101  0.34082  0.31642  0.414650  0.385990  0.329380  0.31212   \n",
       "1  0.002031  0.00000  0.00000  0.07302  0.027533  0.005346  0.006015  0.11824   \n",
       "2  0.000000  0.00000  0.00000  0.00000  0.000000  0.000000  0.000000  0.00000   \n",
       "3  0.000000  0.00000  0.00000  0.00000  0.000000  0.000000  0.000000  0.00000   \n",
       "4  0.231640  0.28629  0.30068  0.28118  0.244890  0.278490  0.290760  0.29380   \n",
       "\n",
       "          8         9                ...                       1643  \\\n",
       "0  0.350670  0.342460                ...                     362.83   \n",
       "1  0.026991  0.005171                ...                   20278.00   \n",
       "2  0.000000  0.000000                ...                  119950.00   \n",
       "3  0.000000  0.000000                ...                    9550.70   \n",
       "4  0.161040  0.166850                ...                     109.62   \n",
       "\n",
       "        1644       1645       1646       1647      1648       1649       1650  \\\n",
       "0   8.592300   9.142700   8.410100   8.792400    1483.3     417.21     892.59   \n",
       "1   0.972010   1.365400   1.803200   1.463400  168740.0   20896.00   34434.00   \n",
       "2   0.000001   0.002466   0.000004   0.002466  729320.0  119950.00  230400.00   \n",
       "3   5.943300   6.021500   4.383900   5.428800   32242.0   10313.00   22308.00   \n",
       "4  17.237000  15.543000  13.671000  15.231000   53559.0    8637.40   18597.00   \n",
       "\n",
       "        1651                          file_name  \n",
       "0     435.28                A_Fish_Called_Wanda  \n",
       "1   20967.00                      A_Goofy_Movie  \n",
       "2  119950.00  A_Million_Ways_to_Die_in_the_West  \n",
       "3    9850.00                       A_Single_Man  \n",
       "4    8679.00                  American_Gangster  \n",
       "\n",
       "[5 rows x 1653 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_vis_train = get_all_vis_features(df_base_train)\n",
    "df_vis_train.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>country_Argentina</th>\n",
       "      <th>country_Australia</th>\n",
       "      <th>country_Bahamas</th>\n",
       "      <th>country_Canada</th>\n",
       "      <th>country_China</th>\n",
       "      <th>country_CzechRepublic</th>\n",
       "      <th>country_Egypt</th>\n",
       "      <th>country_France</th>\n",
       "      <th>country_Germany</th>\n",
       "      <th>country_India</th>\n",
       "      <th>...</th>\n",
       "      <th>rated_G</th>\n",
       "      <th>rated_N/A</th>\n",
       "      <th>rated_NOT RATED</th>\n",
       "      <th>rated_PG</th>\n",
       "      <th>rated_PG-13</th>\n",
       "      <th>rated_R</th>\n",
       "      <th>rated_TV-MA</th>\n",
       "      <th>runtime</th>\n",
       "      <th>year</th>\n",
       "      <th>file_name</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>108</td>\n",
       "      <td>1988</td>\n",
       "      <td>A_Fish_Called_Wanda</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>78</td>\n",
       "      <td>1995</td>\n",
       "      <td>A_Goofy_Movie</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>116</td>\n",
       "      <td>2014</td>\n",
       "      <td>A_Million_Ways_to_Die_in_the_West</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>99</td>\n",
       "      <td>2009</td>\n",
       "      <td>A_Single_Man</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>157</td>\n",
       "      <td>2007</td>\n",
       "      <td>American_Gangster</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 76 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   country_Argentina  country_Australia  country_Bahamas  country_Canada  \\\n",
       "0                  0                  0                0               0   \n",
       "1                  0                  0                0               0   \n",
       "2                  0                  0                0               0   \n",
       "3                  0                  0                0               0   \n",
       "4                  0                  0                0               0   \n",
       "\n",
       "   country_China  country_CzechRepublic  country_Egypt  country_France  \\\n",
       "0              0                      0              0               0   \n",
       "1              0                      0              0               0   \n",
       "2              0                      0              0               0   \n",
       "3              0                      0              0               0   \n",
       "4              0                      0              0               0   \n",
       "\n",
       "   country_Germany  country_India                ...                  rated_G  \\\n",
       "0                0              0                ...                        0   \n",
       "1                0              0                ...                        1   \n",
       "2                0              0                ...                        0   \n",
       "3                0              0                ...                        0   \n",
       "4                0              0                ...                        0   \n",
       "\n",
       "   rated_N/A  rated_NOT RATED  rated_PG  rated_PG-13  rated_R  rated_TV-MA  \\\n",
       "0          0                0         0            0        1            0   \n",
       "1          0                0         0            0        0            0   \n",
       "2          0                0         0            0        1            0   \n",
       "3          0                0         0            0        1            0   \n",
       "4          0                0         0            0        1            0   \n",
       "\n",
       "   runtime  year                          file_name  \n",
       "0      108  1988                A_Fish_Called_Wanda  \n",
       "1       78  1995                      A_Goofy_Movie  \n",
       "2      116  2014  A_Million_Ways_to_Die_in_the_West  \n",
       "3       99  2009                       A_Single_Man  \n",
       "4      157  2007                  American_Gangster  \n",
       "\n",
       "[5 rows x 76 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_meta_train = get_all_meta_features(df_base_train)\n",
    "df_meta_train.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Implementation fun"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#warnings.filterwarnings(action='ignore')\n",
    "\n",
    "#dw = DataWrapper(df_audio_train, df_vis_train, df_txt_train, df_meta_train, df_targets_train)\n",
    "#dw.generate_subspace()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#dw.f_audio[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program_Files\\Anaconda3\\lib\\site-packages\\sklearn\\ensemble\\weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "  from numpy.core.umath_tests import inner1d\n"
     ]
    }
   ],
   "source": [
    "import warnings\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.neighbors import NearestCentroid\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.ensemble import BaggingClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble import AdaBoostClassifier\n",
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "from sklearn.naive_bayes import GaussianNB # put NOTE into paper - not sure if correct bayes\n",
    "from sklearn.model_selection import StratifiedKFold # put NOTE into paper - better as common kfold sampling\n",
    "from sklearn.model_selection import cross_validate\n",
    "from sklearn.base import clone\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "from sklearn.metrics import f1_score\n",
    "from functools import reduce\n",
    "\n",
    "CI_95_FACTOR = 1.96\n",
    "\n",
    "class BaseClassifier:\n",
    "    \n",
    "    def __init__(self, clf, clf_name, modality):\n",
    "        self.clf = clf\n",
    "        self.clf_name = clf_name\n",
    "        self.modality = modality\n",
    "        \n",
    "    def fit(self, df_features, df_targets, verbose=True):\n",
    "        self.clf = clone(self.clf) # reset any previously trained model\n",
    "        np.random.seed(32143421)\n",
    "        if verbose:\n",
    "            print(f\"Starting training for classifier {self.clf_name} and modality {self.modality}\")\n",
    "        self.clf.fit(df_features, df_targets)\n",
    "        \n",
    "    def predict(self, df_features, verbose=True):\n",
    "        if verbose:\n",
    "            print(f\"Starting prediction for classifier {self.clf_name} and modality {self.modality}\")\n",
    "        return self.clf.predict(df_features)    \n",
    "        \n",
    "class ClassifierFactory:\n",
    "    \n",
    "    @staticmethod\n",
    "    def get_metadata_classifiers():\n",
    "        return [BaseClassifier(KNeighborsClassifier(), 'k-Nearest neighbors', 'Metadata'),\n",
    "                BaseClassifier(NearestCentroid(), 'Nearest mean classifier', 'Metadata'),\n",
    "                BaseClassifier(DecisionTreeClassifier(), 'Decision tree', 'Metadata'),\n",
    "                BaseClassifier(LogisticRegression(), 'Logistic regression', 'Metadata'),\n",
    "                BaseClassifier(SVC(gamma='auto'), 'SVM (Gaussian Kernel)', 'Metadata'),\n",
    "                BaseClassifier(BaggingClassifier(),'Bagging', 'Metadata'),\n",
    "                BaseClassifier(RandomForestClassifier(n_estimators=10),'Random Forest', 'Metadata'),\n",
    "                BaseClassifier(AdaBoostClassifier(), 'AdaBoost', 'Metadata'),\n",
    "                BaseClassifier(GradientBoostingClassifier(), 'Gradient Boosting Tree', 'Metadata')]\n",
    "    \n",
    "    @staticmethod\n",
    "    def get_text_classifiers():\n",
    "        return [BaseClassifier(GaussianNB(), 'Naive Bayes', 'Textual'),\n",
    "                BaseClassifier(KNeighborsClassifier(), 'k-Nearest neighbors', 'Textual'),\n",
    "                BaseClassifier(SVC(gamma='auto'), 'SVM (Gaussian Kernel)', 'Textual')]\n",
    "    \n",
    "    @staticmethod\n",
    "    def get_visual_classifiers():\n",
    "        return [BaseClassifier(KNeighborsClassifier(), 'k-Nearest neighbors', 'Visual'),\n",
    "                BaseClassifier(DecisionTreeClassifier(), 'Decision tree', 'Visual'),\n",
    "                BaseClassifier(LogisticRegression(), 'Logistic regression', 'Visual'),\n",
    "                BaseClassifier(SVC(gamma='auto'), 'SVM (Gaussian Kernel)', 'Visual'),\n",
    "                BaseClassifier(RandomForestClassifier(n_estimators=10), 'Random Forest', 'Visual'),\n",
    "                BaseClassifier(AdaBoostClassifier(), 'AdaBoost', 'Visual'),\n",
    "                BaseClassifier(GradientBoostingClassifier(), 'Gradient Boosting Tree', 'Visual')]\n",
    "    \n",
    "    @staticmethod\n",
    "    def get_audio_classifiers():\n",
    "        return [BaseClassifier(LogisticRegression(), 'Logistic regression', 'Audio'),\n",
    "                BaseClassifier(GradientBoostingClassifier(), 'Gradient Boosting Tree', 'Audio')]                \n",
    "                \n",
    "\n",
    "class DataWrapper:\n",
    "    \n",
    "    def __init__(self, df_audio, df_vis, df_txt, df_meta, df_targets):\n",
    "        self.df_audio = df_audio\n",
    "        self.df_vis = df_vis\n",
    "        self.df_txt = df_txt\n",
    "        self.df_meta = df_meta\n",
    "        self.df_targets = df_targets\n",
    "        \n",
    "        self.f_audio=[]\n",
    "        self.f_vis=[]\n",
    "        self.f_txt = []\n",
    "        self.f_meta = []\n",
    "        \n",
    "    def generate_subspace(self):\n",
    "        # TODO insert LVW functionality\n",
    "        # N = len(df_meta)\n",
    "        # MAX_TRIES = 77*N^5\n",
    "        # print(lvf(MAX_TRIES,df_meta, df_targets_train,5))\n",
    "        self.df_audio = self.df_audio.drop('file_name', axis=1)\n",
    "        self.df_vis = self.df_vis.drop('file_name', axis=1)\n",
    "        self.df_txt = self.df_txt.drop('file_name', axis=1)\n",
    "        self.df_meta = self.df_meta.drop('file_name', axis=1)       \n",
    "        \n",
    "\n",
    "        #self.f_audio = self.getMBestFeatures(self.df_audio, self.df_targets, ClassifierFactory.get_audio_classifiers())\n",
    "        #self.f_vis   = self.getMBestFeatures(self.df_vis, self.df_targets, ClassifierFactory.get_visual_classifiers())\n",
    "        #self.f_txt   = self.getMBestFeatures(self.df_txt, self.df_targets, ClassifierFactory.get_text_classifiers())\n",
    "        #self.f_meta  = self.getMBestFeatures(self.df_meta, self.df_targets, ClassifierFactory.get_metadata_classifiers())\n",
    "        self.f_audio = [[1], [1, 4, 5]]\n",
    "        self.f_vis = [[0,  2,  4,  6,  11,  14,  16,  19,  25,  26,  32,  33,  36,  40,  42,  43,  46,  48,  50,  56,  57,  59,  65,  66,  67,  68,  69,  77,  79,  86,  91,  93,  94,  95,  96,  99,  101,  102,  103,  104,  105,  106,  108,  109,  110,  112,  120,  122,  125,  126,  127,  132,  137,  140,  144,  148,  150,  151,  153,  154,  155,  156,  158,  159,  160,  162,  174,  175,  176,  178,  179,  185,  186,  191,  196,  202,  208,  212,  215,  221,  227,  228,  229,  237,  239,  241,  244,  248,  251,  253,  254,  257,  262,  265,  281,  285,  289,  291,  293,  295,  297,  298,  301,  310,  312,  316,  317,  318,  319,  322,  324,  326,  331,  334,  335,  336,  338,  343,  344,  346,  348,  350,  355,  365,  367,  372,  374,  375,  381,  385,  386,  395,  397,  398,  399,  401,  404,  408,  409,  412,  414,  415,  416,  420,  421,  422,  424,  429,  432,  438,  440,  442,  443,  445,  447,  454,  455,  457,  462,  463,  465,  467,  468,  469,  477,  479,  480,  487,  491,  492,  493,  494,  496,  497,  498,  499,  501,  503,  507,  509,  515,  519,  527,  528,  529,  530,  533,  540,  551,  553,  554,  555,  556,  559,  562,  563,  567,  579,  580,  581,  585,  589,  592,  594,  595,  598,  599,  602,  610,  611,  615,  620,  621,  625,  626,  627,  630,  632,  633,  638,  642,  647,  649,  653,  654,  656,  658,  666,  671,  676,  681,  682,  686,  688,  694,  704,  710,  712,  716,  717,  718,  721,  722,  723,  727,  733,  741,  742,  744,  751,  754,  759,  761,  764,  766,  767,  772,  782,  792,  794,  798,  800,  801,  807,  808,  810,  811,  816,  817,  820,  821,  826,  828,  829,  830,  833,  840,  845,  846,  851,  856,  858,  865,  867,  868,  869,  870,  871,  874,  875,  877,  880,  882,  884,  891,  895,  898,  899,  901,  905,  907,  908,  910,  917,  918,  920,  922,  927,  928,  929,  930,  932,  937,  938,  949,  952,  954,  959,  960,  963,  965,  972,  973,  975,  976,  977,  981,  984,  987,  989,  990,  995,  996,  999,  1000,  1005,  1007,  1008,  1016,  1017,  1019,  1020,  1025,  1031,  1035,  1038,  1039,  1041,  1042,  1049,  1050,  1051,  1056,  1060,  1061,  1063,  1068,  1069,  1070,  1071,  1072,  1074,  1075,  1080,  1085,  1091,  1092,  1093,  1094,  1095,  1097,  1100,  1103,  1112,  1114,  1117,  1122,  1124,  1125,  1126,  1127,  1134,  1135,  1136,  1137,  1139,  1150,  1157,  1158,  1160,  1162,  1163,  1167,  1172,  1175,  1176,  1178,  1179,  1184,  1191,  1196,  1198,  1203,  1206,  1207,  1209,  1213,  1214,  1220,  1221,  1227,  1229,  1230,  1235,  1236,  1238,  1239,  1240,  1253,  1256,  1261,  1265,  1268,  1272,  1276,  1280,  1286,  1291,  1298,  1299,  1309,  1311,  1317,  1320,  1323,  1324,  1326,  1328,  1333,  1334,  1335,  1337,  1340,  1344,  1350,  1352,  1359,  1363,  1367,  1369,  1374,  1381,  1383,  1385,  1393,  1397,  1401,  1403,  1404,  1406,  1407,  1415,  1417,  1418,  1419,  1421,  1424,  1431,  1432,  1434,  1435,  1439,  1442,  1443,  1447,  1448,  1451,  1452,  1456,  1458,  1460,  1468,  1469,  1470,  1473,  1475,  1484,  1485,  1488,  1489,  1490,  1493,  1494,  1497,  1498,  1509,  1514,  1517,  1521,  1530,  1533,  1537,  1547,  1551,  1552,  1556,  1557,  1559,  1561,  1562,  1564,  1569,  1571,  1579,  1580,  1582,  1583,  1584,  1586,  1592,  1593,  1596,  1600,  1604,  1605,  1606,  1612,  1614,  1616,  1620,  1624,  1628,  1631,  1633,  1638,  1644,  1645,  1647], \n",
    "                        [1,  2,  3,  5,  11,  12,  17,  18,  19,  22,  24,  25,  27,  28,  29,  31,  34,  35,  42,  43,  44,  46,  47,  50,  52,  55,  56,  57,  59,  60,  62,  63,  64,  65,  66,  68,  70,  75,  78,  80,  82,  84,  85,  87,  88,  89,  90,  95,  97,  100,  101,  102,  104,  105,  106,  107,  108,  113,  119,  120,  121,  122,  123,  124,  125,  126,  127,  131,  132,  133,  134,  138,  139,  140,  142,  144,  145,  146,  148,  151,  153,  156,  160,  163,  164,  165,  167,  168,  169,  170,  173,  175,  176,  177,  178,  179,  180,  181,  182,  183,  187,  189,  191,  194,  196,  197,  202,  204,  205,  206,  207,  208,  210,  211,  212,  213,  214,  215,  217,  219,  220,  222,  223,  224,  225,  226,  227,  230,  232,  234,  235,  237,  238,  239,  240,  242,  243,  244,  245,  247,  248,  250,  252,  255,  256,  258,  259,  260,  261,  263,  264,  265,  266,  267,  268,  269,  271,  272,  274,  278,  280,  281,  283,  286,  287,  288,  289,  294,  295,  298,  300,  303,  305,  310,  311,  312,  316,  317,  318,  320,  321,  322,  324,  326,  328,  329,  332,  334,  335,  337,  338,  342,  344,  345,  346,  347,  348,  350,  351,  353,  356,  358,  359,  360,  362,  363,  364,  365,  366,  369,  370,  375,  376,  377,  380,  381,  383,  384,  388,  389,  392,  393,  394,  395,  397,  399,  400,  401,  403,  406,  407,  408,  409,  411,  412,  413,  414,  416,  417,  418,  420,  423,  424,  425,  426,  429,  430,  431,  432,  433,  435,  436,  441,  444,  446,  447,  450,  451,  453,  454,  456,  457,  458,  460,  462,  463,  464,  466,  467,  469,  470,  475,  476,  480,  481,  482,  485,  486,  487,  489,  490,  491,  496,  497,  499,  502,  505,  508,  511,  512,  513,  514,  516,  517,  518,  521,  527,  529,  531,  532,  533,  536,  539,  541,  544,  546,  547,  549,  554,  555,  557,  558,  561,  563,  570,  573,  574,  577,  578,  579,  582,  583,  584,  586,  587,  588,  592,  595,  596,  597,  598,  601,  602,  604,  605,  606,  607,  608,  610,  612,  613,  615,  616,  617,  618,  619,  620,  622,  623,  625,  626,  627,  630,  631,  634,  640,  641,  644,  646,  649,  650,  651,  652,  653,  657,  659,  663,  664,  665,  667,  668,  669,  673,  674,  682,  683,  684,  686,  687,  690,  693,  694,  696,  697,  698,  699,  702,  703,  705,  706,  709,  710,  713,  714,  715,  716,  717,  719,  720,  724,  728,  729,  731,  732,  733,  734,  739,  740,  741,  742,  744,  748,  749,  751,  752,  753,  754,  755,  759,  762,  763,  764,  767,  771,  772,  774,  777,  778,  779,  781,  782,  788,  791,  793,  798,  799,  800,  804,  806,  807,  808,  810,  814,  816,  818,  821,  823,  824,  826,  827,  828,  831,  833,  836,  839,  844,  846,  847,  848,  851,  853,  854,  856,  858,  859,  860,  861,  864,  865,  866,  869,  871,  873,  876,  879,  880,  881,  885,  886,  888,  889,  897,  898,  899,  904,  907,  909,  911,  914,  917,  918,  920,  921,  922,  923,  924,  925,  927,  930,  931,  936,  938,  942,  943,  944,  946,  948,  949,  950,  951,  952,  953,  962,  963,  966,  967,  968,  973,  974,  975,  976,  977,  978,  979,  981,  982,  983,  986,  991,  992,  995,  996,  997,  998,  1000,  1001,  1002,  1005,  1008,  1011,  1013,  1016,  1017,  1020,  1022,  1024,  1026,  1027,  1032,  1034,  1036,  1039,  1040,  1041,  1044,  1045,  1046,  1049,  1050,  1051,  1052,  1054,  1057,  1059,  1060,  1061,  1062,  1065,  1067,  1068,  1069,  1070,  1073,  1074,  1075,  1078,  1080,  1081,  1085,  1086,  1087,  1088,  1091,  1092,  1093,  1094,  1096,  1098,  1099,  1100,  1102,  1103,  1104,  1106,  1107,  1108,  1112,  1113,  1114,  1116,  1117,  1118,  1122,  1123,  1125,  1126,  1128,  1129,  1130,  1131,  1132,  1134,  1137,  1138,  1139,  1142,  1143,  1145,  1147,  1151,  1152,  1153,  1155,  1159,  1160,  1162,  1164,  1165,  1166,  1167,  1168,  1169,  1170,  1171,  1174,  1176,  1178,  1179,  1181,  1182,  1184,  1185,  1186,  1187,  1188,  1189,  1190,  1193,  1195,  1196,  1198,  1199,  1202,  1203,  1205,  1206,  1208,  1209,  1210,  1211,  1212,  1213,  1214,  1215,  1216,  1217,  1218,  1219,  1220,  1223,  1224,  1225,  1226,  1228,  1231,  1233,  1237,  1238,  1239,  1242,  1243,  1250,  1252,  1253,  1254,  1256,  1257,  1260,  1262,  1264,  1265,  1267,  1268,  1270,  1271,  1272,  1273,  1275,  1278,  1280,  1281,  1286,  1290,  1291,  1292,  1296,  1297,  1298,  1303,  1305,  1308,  1309,  1310,  1313,  1314,  1316,  1317,  1319,  1320,  1321,  1322,  1323,  1324,  1325,  1329,  1330,  1331,  1332,  1333,  1335,  1337,  1338,  1346,  1348,  1350,  1352,  1357,  1358,  1361,  1362,  1365,  1366,  1367,  1369,  1370,  1371,  1372,  1374,  1377,  1379,  1381,  1382,  1384,  1385,  1387,  1388,  1390,  1391,  1393,  1394,  1395,  1396,  1397,  1398,  1400,  1403,  1404,  1406,  1407,  1408,  1411,  1412,  1413,  1414,  1416,  1420,  1421,  1422,  1424,  1425,  1426,  1430,  1431,  1432,  1433,  1435,  1437,  1438,  1440,  1443,  1444,  1445,  1446,  1449,  1450,  1451,  1452,  1455,  1457,  1458,  1460,  1463,  1465,  1470,  1471,  1473,  1474,  1475,  1479,  1480,  1481,  1483,  1484,  1487,  1491,  1493,  1495,  1497,  1498,  1500,  1501,  1503,  1506,  1507,  1508,  1510,  1512,  1514,  1515,  1516,  1518,  1519,  1520,  1522,  1523,  1524,  1525,  1526,  1528,  1529,  1530,  1531,  1532,  1534,  1535,  1536,  1537,  1538,  1539,  1540,  1542,  1543,  1545,  1549,  1553,  1555,  1556,  1557,  1558,  1559,  1562,  1563,  1564,  1566,  1572,  1573,  1574,  1575,  1576,  1578,  1579,  1584,  1585,  1588,  1591,  1592,  1593,  1595,  1598,  1600,  1603,  1605,  1606,  1607,  1609,  1611,  1612,  1614,  1615,  1616,  1617,  1619,  1621,  1622,  1623,  1627,  1628,  1631,  1632,  1636,  1637,  1638,  1639,  1640,  1641,  1643,  1644,  1646,  1647,  1650], \n",
    "                        [2,  3,  4,  5,  7,  10,  12,  16,  18,  20,  23,  30,  31,  34,  37,  38,  39,  45,  52,  54,  58,  59,  62,  66,  68,  73,  75,  76,  82,  85,  86,  88,  92,  93,  95,  99,  102,  105,  107,  108,  109,  110,  111,  114,  121,  125,  128,  129,  135,  136,  137,  139,  142,  148,  150,  153,  156,  158,  159,  160,  162,  163,  165,  166,  169,  170,  174,  177,  178,  181,  186,  189,  191,  195,  196,  198,  201,  203,  204,  205,  206,  207,  208,  209,  210,  213,  214,  219,  223,  234,  237,  238,  242,  245,  248,  260,  266,  268,  273,  274,  279,  282,  287,  293,  294,  297,  304,  305,  306,  313,  315,  316,  322,  324,  330,  335,  336,  337,  342,  344,  345,  352,  360,  369,  376,  377,  380,  384,  386,  393,  396,  397,  399,  400,  401,  416,  422,  423,  425,  426,  429,  430,  431,  434,  440,  444,  445,  450,  452,  454,  455,  458,  461,  466,  471,  478,  482,  483,  484,  485,  491,  493,  497,  499,  502,  504,  506,  507,  512,  518,  519,  523,  531,  537,  540,  542,  543,  552,  553,  554,  561,  567,  569,  590,  592,  602,  603,  604,  605,  609,  611,  612,  615,  616,  617,  618,  620,  623,  626,  630,  632,  635,  636,  639,  640,  642,  643,  644,  645,  646,  647,  652,  655,  656,  663,  665,  666,  670,  671,  672,  673,  674,  676,  680,  683,  693,  694,  698,  701,  704,  707,  711,  714,  716,  717,  721,  723,  726,  728,  729,  730,  736,  740,  742,  744,  745,  746,  749,  755,  760,  764,  768,  771,  773,  776,  779,  784,  785,  788,  793,  796,  799,  801,  803,  812,  816,  817,  820,  827,  828,  832,  837,  841,  843,  846,  849,  851,  853,  856,  863,  865,  869,  871,  876,  878,  880,  881,  882,  885,  886,  889,  890,  891,  894,  896,  898,  899,  900,  901,  904,  906,  910,  911,  917,  920,  930,  948,  950,  952,  956,  959,  960,  961,  962,  965,  967,  968,  973,  975,  977,  981,  982,  984,  988,  990,  992,  995,  996,  1000,  1004,  1005,  1010,  1015,  1019,  1020,  1022,  1023,  1029,  1032,  1035,  1036,  1043,  1048,  1049,  1050,  1054,  1055,  1056,  1059,  1065,  1066,  1068,  1069,  1070,  1072,  1073,  1074,  1076,  1081,  1083,  1084,  1087,  1089,  1093,  1095,  1099,  1104,  1119,  1121,  1122,  1123,  1135,  1136,  1147,  1153,  1154,  1156,  1157,  1160,  1161,  1162,  1169,  1173,  1174,  1178,  1180,  1184,  1189,  1191,  1203,  1210,  1214,  1215,  1217,  1225,  1227,  1229,  1236,  1240,  1242,  1243,  1244,  1248,  1249,  1259,  1260,  1263,  1266,  1270,  1275,  1276,  1277,  1278,  1280,  1283,  1284,  1285,  1290,  1297,  1300,  1306,  1307,  1308,  1311,  1312,  1317,  1320,  1322,  1325,  1326,  1327,  1330,  1331,  1332,  1335,  1337,  1342,  1345,  1346,  1349,  1351,  1353,  1355,  1356,  1361,  1363,  1364,  1365,  1367,  1370,  1371,  1372,  1385,  1391,  1392,  1394,  1397,  1398,  1399,  1401,  1403,  1410,  1418,  1431,  1432,  1433,  1436,  1438,  1439,  1440,  1441,  1445,  1450,  1451,  1453,  1456,  1460,  1462,  1463,  1465,  1467,  1469,  1471,  1474,  1479,  1490,  1496,  1498,  1499,  1503,  1505,  1506,  1513,  1516,  1517,  1518,  1521,  1523,  1526,  1527,  1528,  1530,  1531,  1535,  1536,  1537,  1539,  1543,  1548,  1552,  1554,  1555,  1558,  1559,  1562,  1564,  1566,  1567,  1569,  1571,  1572,  1573,  1574,  1575,  1579,  1591,  1592,  1596,  1598,  1602,  1613,  1622,  1624,  1626,  1630,  1634,  1635,  1636,  1640,  1643,  1644,  1647,  1650], \n",
    "                        [47,  65,  71,  72,  101,  135,  151,  161,  169,  181,  196,  222,  228,  240,  255,  263,  283,  327,  335,  375,  396,  400,  433,  449,  464,  466,  495,  505,  524,  531,  551,  582,  599,  666,  743,  751,  764,  774,  776,  849,  851,  856,  866,  873,  874,  881,  893,  895,  899,  915,  934,  936,  962,  963,  970,  1002,  1005,  1032,  1106,  1147,  1196,  1232,  1243,  1268,  1288,  1322,  1323,  1365,  1374,  1380,  1384,  1392,  1395,  1409,  1420,  1458,  1472,  1518,  1547,  1565,  1590,  1601,  1612,  1616,  1629,  1631], \n",
    "                        [0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10,  11,  12,  13,  14,  15,  16,  17,  18,  19,  20,  22,  23,  24,  25,  26,  27,  28,  29,  30,  31,  32,  33,  34,  35,  36,  37,  38,  39,  40,  41,  42,  43,  45,  46,  48,  50,  51,  52,  53,  54,  55,  56,  57,  58,  59,  60,  61,  62,  63,  64,  65,  66,  67,  68,  69,  70,  71,  72,  73,  75,  76,  77,  78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,  92,  93,  94,  95,  96,  97,  98,  99,  100,  102,  103,  104,  105,  106,  107,  108,  109,  110,  111,  112,  113,  114,  116,  117,  118,  119,  120,  121,  123,  124,  125,  126,  127,  128,  129,  130,  131,  132,  133,  134,  135,  136,  137,  138,  139,  140,  142,  143,  144,  145,  146,  147,  148,  149,  151,  152,  153,  154,  155,  156,  157,  158,  159,  160,  161,  162,  163,  164,  165,  166,  167,  168,  169,  170,  171,  172,  173,  174,  175,  176,  177,  178,  179,  180,  181,  182,  183,  184,  185,  186,  187,  188,  189,  190,  191,  192,  193,  194,  195,  196,  197,  198,  199,  200,  201,  202,  203,  204,  205,  206,  207,  208,  209,  210,  211,  212,  213,  214,  215,  216,  217,  218,  220,  221,  222,  223,  224,  225,  226,  227,  228,  229,  230,  231,  232,  233,  234,  235,  236,  237,  238,  239,  240,  241,  242,  243,  244,  245,  246,  247,  248,  249,  250,  251,  252,  253,  254,  255,  256,  257,  258,  259,  260,  261,  262,  263,  264,  265,  266,  267,  268,  269,  270,  271,  272,  273,  274,  275,  276,  277,  278,  279,  280,  281,  282,  283,  284,  285,  286,  287,  288,  289,  290,  291,  292,  293,  294,  295,  296,  297,  298,  299,  300,  301,  302,  303,  304,  305,  306,  307,  308,  309,  310,  311,  312,  313,  314,  315,  316,  317,  318,  319,  320,  321,  322,  323,  324,  325,  326,  327,  328,  330,  331,  332,  333,  334,  335,  336,  337,  338,  339,  340,  341,  342,  343,  344,  346,  347,  348,  349,  350,  351,  352,  354,  355,  356,  357,  358,  359,  360,  361,  362,  363,  364,  365,  366,  368,  369,  370,  371,  372,  373,  374,  375,  376,  377,  378,  379,  380,  381,  382,  383,  384,  385,  386,  387,  389,  390,  391,  392,  393,  394,  395,  396,  397,  398,  399,  400,  401,  402,  403,  404,  406,  407,  408,  409,  410,  411,  412,  413,  414,  415,  416,  417,  418,  419,  420,  421,  422,  423,  424,  425,  426,  427,  428,  429,  430,  431,  432,  433,  434,  435,  436,  437,  438,  439,  440,  441,  442,  443,  444,  445,  446,  447,  448,  449,  450,  451,  452,  453,  454,  455,  456,  457,  458,  459,  460,  461,  462,  463,  464,  465,  466,  467,  468,  469,  470,  471,  472,  473,  474,  475,  476,  477,  478,  479,  481,  482,  483,  484,  485,  486,  487,  488,  489,  490,  491,  492,  493,  494,  495,  496,  497,  498,  499,  500,  501,  502,  503,  504,  505,  506,  507,  508,  509,  510,  511,  512,  513,  514,  515,  516,  517,  518,  519,  520,  521,  522,  523,  524,  525,  526,  527,  528,  529,  532,  533,  534,  535,  536,  537,  538,  539,  540,  541,  542,  543,  544,  545,  546,  547,  548,  549,  550,  551,  552,  553,  554,  555,  556,  557,  558,  559,  560,  561,  562,  563,  564,  565,  567,  568,  569,  570,  571,  572,  573,  574,  575,  576,  577,  578,  579,  580,  581,  582,  583,  584,  585,  586,  587,  588,  589,  590,  591,  592,  593,  594,  595,  596,  597,  598,  599,  600,  601,  602,  603,  604,  605,  606,  607,  608,  609,  610,  611,  612,  613,  614,  615,  616,  617,  618,  619,  620,  621,  622,  623,  624,  625,  626,  627,  628,  629,  630,  631,  632,  633,  634,  635,  636,  637,  638,  639,  640,  641,  642,  643,  644,  645,  646,  647,  648,  649,  650,  651,  652,  653,  654,  655,  656,  657,  658,  659,  660,  661,  662,  663,  664,  665,  666,  667,  668,  669,  670,  671,  672,  673,  674,  675,  676,  677,  678,  679,  680,  681,  682,  683,  684,  685,  686,  687,  688,  689,  690,  691,  692,  693,  694,  695,  696,  697,  698,  699,  700,  701,  702,  703,  704,  705,  706,  707,  708,  709,  710,  711,  712,  713,  714,  715,  716,  717,  718,  719,  720,  721,  722,  723,  724,  725,  726,  727,  728,  729,  730,  731,  732,  733,  734,  735,  736,  737,  738,  739,  740,  741,  742,  743,  744,  745,  746,  747,  748,  749,  750,  751,  752,  753,  754,  755,  756,  757,  758,  759,  760,  761,  762,  763,  764,  765,  766,  767,  768,  769,  770,  771,  772,  773,  774,  775,  776,  777,  778,  779,  780,  781,  782,  783,  784,  785,  786,  787,  788,  789,  790,  791,  792,  793,  794,  795,  796,  797,  798,  799,  800,  801,  802,  803,  804,  805,  806,  807,  808,  809,  810,  812,  813,  814,  815,  816,  817,  818,  820,  821,  822,  823,  824,  825,  826,  827,  828,  829,  830,  831,  832,  833,  834,  835,  836,  837,  838,  839,  840,  841,  842,  843,  844,  845,  846,  847,  848,  849,  850,  851,  852,  853,  854,  855,  856,  857,  858,  859,  860,  861,  862,  863,  864,  865,  866,  867,  868,  869,  870,  871,  872,  873,  874,  875,  876,  877,  878,  879,  880,  881,  882,  883,  884,  885,  886,  887,  888,  889,  890,  891,  892,  893,  894,  895,  896,  897,  898,  899,  900,  901,  902,  903,  904,  905,  906,  907,  908,  909,  910,  911,  912,  913,  914,  915,  917,  918,  919,  920,  921,  922,  923,  924,  925,  926,  927,  928,  929,  931,  932,  933,  934,  935,  936,  937,  938,  939,  940,  941,  942,  943,  944,  945,  946,  947,  948,  949,  950,  951,  952,  953,  954,  955,  956,  957,  958,  959,  960,  961,  962,  963,  964,  965,  966,  967,  968,  969,  970,  971,  972,  973,  974,  975,  976,  977,  978,  979,  980,  981,  982,  983,  985,  986,  987,  988,  989,  990,  991,  992,  993,  994,  995,  996,  997,  998,  999,  1000,  1001,  1002,  1003,  1004,  1005,  1006,  1007,  1008,  1009,  1010,  1011,  1012,  1013,  1014,  1015,  1016,  1017,  1019,  1020,  1021,  1022,  1023,  1025,  1026,  1027,  1028,  1029,1030,  1031, 1032, 1033, 1034, 1035, 1036, 1037, 1038, 1039, 1040, 1041, 1042, 1043, 1044, 1045, 1046, 1048, 1049, 1050, 1051, 1052, 1053, 1054, 1055, 1056, 1057, 1058, 1059, 1060, 1061, 1062, 1063,1064, 1065, 1066, 1067, 1068, 1069, 1070, 1071, 1072, 1073, 1074, 1075, 1076, 1077, 1078, 1080, 1081, 1082, 1083, 1084, 1085, 1086, 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109, 1110, 1111, 1112, 1113, 1114, 1115, 1116, 1117, 1118, 1119, 1120, 1121, 1122, 1123, 1124, 1125, 1126, 1127, 1128, 1129, 1130, 1131,1132, 1133, 1134, 1135, 1136, 1137, 1138, 1139, 1140, 1142, 1143, 1144, 1145, 1146, 1147, 1148, 1149, 1150, 1151, 1152, 1153, 1154, 1155, 1156, 1157, 1158, 1159, 1160, 1161, 1162, 1163, 1164, 1165,1166, 1167, 1168, 1169, 1170, 1171, 1172, 1173, 1174, 1175, 1176, 1177, 1178, 1179, 1180, 1181, 1182, 1183, 1184, 1185, 1186, 1187, 1188, 1189, 1190, 1191, 1192, 1193, 1194, 1195, 1196, 1197, 1198,1199, 1200, 1201, 1202, 1203, 1204, 1205, 1206, 1207, 1208, 1209, 1210, 1211, 1212, 1213, 1214, 1215, 1216, 1217, 1218, 1219, 1220, 1221, 1222, 1223, 1224, 1225, 1226, 1227, 1228, 1229, 1230, 1231,1232, 1233, 1234, 1235, 1236, 1237, 1238, 1239, 1240, 1241, 1242, 1243, 1244, 1245, 1246, 1247, 1248, 1249, 1250, 1251, 1252, 1253, 1254, 1255, 1256, 1257, 1258, 1259, 1260, 1261, 1262, 1263, 1264,1265, 1266, 1267, 1268, 1269, 1270, 1271, 1272, 1273, 1275, 1276, 1277, 1278, 1279, 1280, 1281, 1282, 1283, 1284, 1285, 1286, 1287, 1288, 1289, 1290, 1291, 1292, 1293, 1294, 1295, 1296, 1297, 1298,1300, 1301, 1302, 1303, 1304, 1305, 1306, 1307, 1308, 1309, 1310, 1311, 1312, 1313, 1314, 1315, 1316, 1317, 1318, 1319, 1320, 1321, 1322, 1323, 1324, 1325, 1326, 1327, 1328, 1329, 1330, 1331, 1332,1333, 1334, 1335, 1336, 1337, 1338, 1339, 1340, 1341, 1342, 1343, 1344, 1345, 1346, 1347, 1348, 1349, 1350, 1351, 1352, 1353, 1354, 1355, 1356, 1357, 1358, 1359, 1360, 1361, 1362, 1363, 1364, 1365,1366, 1367, 1368, 1369, 1370, 1371, 1372, 1373, 1374, 1375, 1376, 1377, 1378, 1379, 1380, 1381, 1382, 1383, 1384, 1385, 1386, 1387, 1388, 1389, 1390, 1391, 1392, 1393, 1394, 1395, 1396, 1398, 1399,1400, 1401, 1402, 1403, 1404, 1405, 1406, 1407, 1408, 1409, 1410, 1411, 1412, 1413, 1414, 1416, 1417, 1418, 1419, 1420, 1421, 1422, 1423, 1424, 1425, 1426, 1427, 1428, 1429, 1430, 1431, 1432, 1433,1434, 1435, 1437, 1438, 1439, 1440, 1441, 1442, 1443, 1444, 1445, 1446, 1447, 1448, 1449, 1450, 1451, 1452, 1453, 1454, 1455, 1456, 1457, 1458, 1459, 1460, 1461, 1462, 1463, 1464, 1465, 1466, 1467,1468, 1469, 1470, 1471, 1472, 1473, 1474, 1475, 1476, 1477, 1478, 1479, 1480, 1481, 1482, 1483, 1484, 1485, 1486, 1487, 1488, 1489, 1490, 1491, 1492, 1493, 1494, 1495, 1496, 1497, 1498, 1499, 1500,1501, 1502, 1503, 1504, 1505, 1506, 1507, 1508, 1509, 1510, 1511, 1512, 1513, 1514, 1515, 1516, 1517, 1518, 1519, 1520, 1521, 1522, 1523, 1524, 1525, 1526, 1527, 1528, 1529, 1530, 1531, 1532, 1533,1534, 1536, 1537, 1538, 1539, 1540, 1541, 1542, 1543, 1544, 1545, 1546, 1547, 1548, 1549, 1550, 1552, 1553, 1554, 1555, 1556, 1557, 1558, 1559, 1560, 1561, 1562, 1563, 1564, 1565, 1566, 1567, 1568,1569, 1570, 1571, 1572, 1573, 1574, 1575, 1576, 1577, 1578, 1579, 1580, 1581, 1582, 1583, 1584, 1585, 1586, 1587, 1588, 1589, 1590, 1591, 1592, 1593, 1594, 1595, 1596, 1597, 1598, 1599, 1600, 1601,1602, 1603, 1604, 1605, 1606, 1607, 1608, 1609, 1610, 1611, 1612, 1613, 1614, 1615, 1616, 1617, 1618, 1619, 1620, 1621, 1622, 1623, 1624, 1625, 1626, 1627, 1628, 1629, 1630, 1631, 1632, 1633, 1634,1635, 1636, 1637, 1638, 1639, 1640, 1641, 1642, 1643, 1644, 1645, 1646, 1647, 1648, 1649, 1650], \n",
    "                        [2,  7,  9,  13,  22,  42,  54,  63,  68,  80,  91,  112,  116,  119,  129,  139,  143,  153,  166,  174,  183,  188,  194,  196,  220,  228,  257,  284,  286,  288,  330,  332,  367,  383,  392,  413,  416,  426,  436,  439,  442,  454,  481,  487,  492,  494,  528,  533,  537,  550,  557,  570,  583,  586,  595,  597,  613,  630,  653,  675,  690,  704,  707,  709,  721,  722,  751,  753,  769,  779,  780,  787,  803,  813,  817,  818,  825,  840,  847,  857,  876,  878,  881,  886,  899,  922,  934,  938,  940,  941,  963,  969,  985,  997,  1001,  1003,  1005,  1013,  1027,  1030,  1038,  1043,  1044,  1056,  1062,  1069,  1096,  1129,  1133,  1149,  1152,  1155,  1166,  1173,  1178,  1180,  1181,  1187,  1192,  1194,  1216,  1220,  1231,  1234,  1258,  1284,  1290,  1319,  1324,  1335,  1342,  1345,  1363,  1374,  1381,  1385,  1387,  1391,  1394,  1418,  1421,  1423,  1424,  1429,  1430,  1432,  1434,  1466,  1481,  1504,  1511,  1523,  1553,  1558,  1578,  1590,  1592,  1601,  1603,  1612,  1618,  1620,  1628,  1629], \n",
    "                        [72, 772, 816, 1259, 1377, 1499, 1520, 1524, 1538, 1639]]\n",
    "        self.f_txt = [['1882',  '1895',  '19',  '24000',  '30',  '30yearold',  '6.1',  '600000',  '74th',  'abilities',  'able',  'account',  'actually.1',  'administrator',  'adventure.1',  'affair',  'affecting',  'affections',  \t\t'afraid',  'africa',  'age.1',  'agency',  'agent.2',  'ahead',  'ahead.1',  'aided',  'alaska',  'alberts',  'alexei',  'alice',  'alien',  'allows',  'altered',  'alzheimers',  'amelia.1',  'amid',  'amidst',  \t\t'angeles.1',  'animals',  'anna',  'announce',  'anymore',  'apprentice',  'archer',  'archie',  'area',  'arrest',  'asia',  'ask',  'asks',  'assigning',  'attitudes',  'attorney',  'available',  'awakens',  \t\t'aware',  'away',  'awesome',  'babadook.1',  'badges',  'bajaj',  'ball.1',  'band',  'bartender',  'batman',  'battle',  'beach.1',  'beautiful',  'begin',  'big',  'bigfoot',  'biggest',  'billionaire',  \t\t'birth',  'blair',  'blamed',  'blinded',  'blue',  'blurs',  'body.1',  'boggarts',  'boot',  'booth',  'born',  'bourne',  'boys.2',  'bradley',  'bringing',  'bros',  'bumpy',  'bury',  'busan',  'busboy',  \t\t'buying',  'cade',  'california',  'called',  'came',  'capitol',  'captain',  'captures',  'car',  'care',  'career',  'career.1',  'cares',  'carry',  'case',  'celebrated',  'celebrity',  'center',  'certainly',  \t\t'chairman',  'challenge',  'change',  'changes',  'chaos',  'charge',  'charity',  'charley',  'chicago',  'childhood.1',  'china',  'choice.1',  'chris',  'chungsu',  'cia',  'city',  'club',  'club.1',  'clubs',  \t\t'cocaine',  'cofounder',  'coincidentally',  'collapses',  'collect',  'comes',  'comic',  'command',  'common',  'community',  'company',  'concert.1',  'conquer',  'consecutive',  'constant',  'contacted',  \t\t'continuum',  'control',  'conventional',  'conversations',  'cool.1',  'cope',  'corroding',  'corrupt',  'course.1',  'courtship',  'coveted',  'crevice',  'crime',  'cross',  'crown',  'cryptanalyst',  'cuba',  \t\t'cuca',  'culture',  'cypher',  'dad',  'dancer',  'dancer.1',  'dancers',  'daring',  'darkest',  'darko',  'daughterinlaw',  'death',  'death.1',  'decided',  'decides',  'decision',  'deepens',  'defeats',  \t\t'defense',  'del',  'devastating',  'devoted',  'die.1',  'different',  'different.1',  'difficult',  'directed',  'dirty',  'disappears.1',  'disease',  'disguise',  'dishonorably',  'dismantle',  'divorced',  \t\t'dj.1',  'doctor.1',  'dollars',  'donnies',  'doubts',  'dracula',  'drag',  'drawing',  'drawn',  'dread',  'drought',  'dumbest',  'dysfunctional',  'earlier',  'earth.1',  'einstein',  'elections',  'elsa',  \t\t'eluded',  'embarks',  'endurance',  'enforcement',  'engaged',  'enraged',  'entertaining',  'entertainment',  'entertainment.1',  'entertainment.2',  'environment',  'epic',  'epidemic',  'escape',  'escape.1',  \t\t'essential',  'ethan',  'evening',  'events.1',  'eventually',  'everdeen',  'expand',  'expect',  'extinction',  'extremely',  'exwife',  'faces',  'falling',  'falls',  'fame.1',  'famine',  'father',  'father.1',  \t\t'fault',  'favorites',  'fellow',  'feuding',  'fight.1',  'finally',  'flamboyant',  'flames',  'fling',  'florida',  'foil',  'followed',  'follower',  'food.1',  'forestwithout',  'forget',  'formality',  \t\t'fortyfive',  'forward',  'frances',  'frankenstein',  'fransokyo',  'frayed',  'freedom',  'french',  'friends',  'friendship',  'funeral',  'gal',  'games',  'games.2',  'garage',  'gatsby',  'generational',  \t\t'georgina',  'ghostcommunity',  'ghosts',  'girl',  'girls',  'giving',  'good.1',  'government',  'graduated',  'grimm',  'ground',  'group.2',  'grown',  'guard',  'guests',  'gutierrez',  'guy',  'guy.1',  'haars',  \t\t'hacker',  'halfstriped',  'hammond',  'hanna.1',  'happens',  'harrowing',  'harsh',  'having',  'hazel',  'heads',  'heart.1',  'heartland',  'heirloom',  'help',  'helping',  'herd.1',  'hero',  'hide',  'history',  \t\t'hivaids',  'hoeks',  'home.1',  'homerun',  'honey',  'hopelessly',  'hoping',  'hotel',  'hotheaded',  'human.1',  'hunted',  'hunter',  'husband',  'hysterical',  'ill',  'imagination',  'immediately',  'immensely',  \t\t'impaired',  'implanted',  'importer',  'importing',  'impresses',  'impromptu',  'incidents',  'incredible',  'inept',  'ineptness',  'infatuated',  'inner',  'innocent',  'inspire',  'instead',  'insults',  \t\t'interaction',  'intricate',  'intrigued',  'isabella.1',  'islandset',  'italian',  'jack.1',  'james.1',  'japanese',  'jedediah',  'jim',  'jobs',  'john.1',  'johnnys',  'jorge',  'josephine',  'journey',  \t\t'jules',  'jungle.1',  'junkie',  'kane',  'karenina',  'kate',  'katniss',  'katniss.1',  'kenai.1',  'khoobsurat',  'kidnapped',  'killed',  'killing',  'know',  'kodas',  'kunis',  'la.1',  'lame',  'land',  \t\t'law',  'leach',  'lead',  'leading',  'lease',  'legendary',  'lemon',  'leopard',  'life.1',  'life.2',  'like',  'likes',  'lineaments',  'linglings',  'living',  'lizzie.2',  'locksley',  'loggerheads',  'london',  \t\t'look',  'looks',  'lottery.1',  'love.2',  'madness',  'magnate',  'maiden',  'make',  'mammoths',  'man',  'manage',  'manages',  'manolo',  'mars.1',  'mars.2',  'marylouis',  'mastering',  'materialistic',  \t\t'mathis',  'matters',  'mayhem.1',  'mechanic',  'medical',  'meet',  'meeting',  'merger',  'met.1',  'middleclass',  'military',  'mindset',  'mirkwood',  'mismatched',  'mobsters',  'modern',  'monroe',  \t\t'montenegro',  'months',  'morpheus.1',  'movie.1',  'mrs',  'mushu',  'musicians',  'myriad',  'mysteriously',  'mystery',  'nailbiting',  'national',  'native',  'natural',  'needs',  'nefarious',  'neglects',  \t\t'neighborhood',  'nerdy',  'newfound',  'news',  'nice',  'nina.1',  'northern',  'note',  'nouveau',  'novel.1',  'novels',  'november',  'obsession',  'odds',  'offensive',  'office',  'ones',  'ones.1',  'open',  \t\t'opposition',  'opting',  'ordered',  'outcast',  'outcome.1',  'overcame',  'owner',  'owner.1',  'paid',  'pair',  'pals',  'pamela',  'pan',  'panem.1',  'paris',  'parker',  'partner',  'partner.1',  'past',  \t\t'perfectionist',  'personification',  'peterson',  'pharaoh',  'physiotherapist',  'picture',  'places',  'play.3',  'plot',  'poet',  'policeman',  'politics',  'politics.1',  'pose',  'post',  'powerful',  'powers',  \t\t'president',  'prey',  'prince',  'problem',  'production',  'project',  'promised',  'proper',  'provide',  'puts',  'quite',  'race.1',  'ralph.1',  'ralphs',  'ram.1',  'range',  'rationalizes',  'raucous',  \t\t'ravaged',  'reacher.1',  'reader',  'realize',  'recklessness',  'recording',  'recovers',  'recovers.1',  'recurring',  'refugees',  'regina',  'regina.1',  'relationship',  'relief',  'remember',  'remembering',  \t\t'remembers',  'remind',  'rene',  'renee',  'reporter',  'rescued',  'resolves',  'responsibilities',  'restaurant.1',  'retelling',  'retired.1',  'retribution.1',  'richard',  'riche',  'riding',  'rises',  'rivals',  \t\t'riverine',  'robert',  'roberts',  'robotics',  'role',  'romance.1',  'romantic',  'rose',  'roxanne.1',  'rumors',  'ryder',  'safe',  'safely',  'salmon',  'samuels.1',  'sangha',  'sarah',  'school.1',  'scott',  \t\t'scouted',  'scouting',  'secrets',  'seeks',  'seemingly',  'selfconfessed',  'send',  'sensation',  'sensuality',  'seriously',  'set',  'sexual',  'shape',  'shearing',  'shocked',  'sister',  'situation',  'slaves',  \t\t'slum',  'sneaks',  'solar',  'solution.1',  'son',  'son.1',  'souls',  'spare',  'spies',  'spirit.1',  'spook',  'sportello',  'starina',  'stars',  'statement',  'states',  'status',  'stealing',  'stephen',  \t\t'stephens',  'stir',  'stolen',  'stripes.1',  'stripped',  'student',  'students.1',  'studying',  'successful',  'suddenly',  'suffers',  'sulley',  'summary',  'sung',  'superiors',  'superstar',  'surprised',  \t\t'survival',  'survive',  'sustain',  'sweat',  'sweeps',  'sweet',  'switch',  'sylvia',  'sympathetic',  'tablets',  'talent',  'teaching',  'team',  'televised',  'temperamental',  'thriller',  'thrown',  'time.2',  \t\t'told',  'topsecret',  'torn',  'toro',  'tour',  'towns',  'trade.1',  'traitors',  'travels',  'triangle.1',  'trip',  'trouble',  'truest',  'truths',  'turned',  'twisted',  'ultimately',  'ultimatum',  \t\t'unbeknownst',  'unit',  'universal',  'vain',  'valentines',  'various',  'villain',  'viralvideoed',  'wallace',  'ward',  'way',  'werewolves',  'wessex',  'west.1',  'wicked',  'widen',  'wife.1',  'wild',  \t\t'wilde.1',  'windfall',  'winning',  'witch.2',  'woods.1',  'wooed',  'world',  'world.2',  'writer',  'writing.1',  'yeager',  'yeager.1',  'years.1',  'young.1',  'youngja',  'zoologists'], \n",
    "                        ['118th',  '15',  '15year',  '1895',  '1950',  '24000',  '30',  '3d',  '6',  '70s',  'abandon',  'academic',  'accident.1',  'accused',  'accuseds',  'actionpacked',  'activists',  'actorgigolo',  'adolescent',  \t\t'advancement',  'adventure.1',  'affairs',  'affections',  'agent',  'ago',  'agree',  'ahkmenrah',  'albert',  'alberts',  'alice',  'ally',  'alongside',  'amelia',  'america',  'ancestors',  'anderson',  'appearing',  \t\t'armed',  'army',  'arrest',  'assembly',  'astounding',  'attention',  'attila',  'attitudes',  'attorney',  'auditions',  'avoid',  'awesome',  'baby.1',  'background',  'bajaj',  'ball.1',  'baronial',  'baseball.1',  \t\t'batman',  'battling',  'baymax',  'beach.1',  'befalls',  'befriend',  'belief',  'believes',  'best.1',  'bhoot',  'bhoothnath',  'bhoothnaths',  'big.1',  'bigfoot.1',  'birthday',  'black',  'blessings',  'bloom',  \t\t'blue',  'blurs',  'bobby',  'body.1',  'bohemian',  'boil',  'bond.1',  'book',  'boot',  'born',  'bourne',  'boxer',  'boy.1',  'boyfriend.2',  'boys',  'boys.2',  'brass',  'break',  'british',  'brittle',  \t\t'brutally',  'bumpy',  'burglar',  'busboy',  'business',  'buying',  'california',  'camp',  'capitol',  'captured',  'captures',  'career.1',  'carl',  'cassel',  'castle',  'catapults',  'celebrate',  'challenge',  \t\t'chaos.1',  'charge',  'charity',  'cheer',  'chiffre',  'chiffre.1',  'chinese',  'choice.1',  'chose',  'city.2',  'citys',  'claims',  'clark',  'classmates',  'climate',  'coal',  'codebreakers',  'cold',  \t\t'collapsed',  'collapses',  'colleagues',  'collection',  'colombo',  'comatose',  'comedyadventure',  'comes',  'committee',  'compassionate',  'comprising',  'conclusion',  'conservative',  'considerable',  \t\t'consumed',  'control',  'converted',  'coolidge',  'corporation',  'corps',  'county',  'court',  'creating',  'crevice',  'crime.1',  'crossing',  'crude',  'cruise',  'cuba',  'curse',  'cutthroat.1',  'dads',  \t\t'daley',  'dancer',  'day.2',  'days',  'days.1',  'deal',  'death.1',  'declared',  'defeats',  'defend',  'defense',  'deliberately',  'demands',  'descend',  'desperate',  'destroys',  'detective',  'develops',  \t\t'devotes',  'diamond',  'died.1',  'difference',  'different',  'discovered.1',  'discovery',  'disenchanted',  'disgraced',  'dishonorably',  'dissuaded',  'divorce',  'divorced',  'dj',  'dj.1',  'doctor',  'doesnt',  \t\t'doing',  'donnie',  'donnies',  'doubts',  'dracula',  'drag',  'dragon.1',  'dread',  'dream',  'dress',  'drives',  'drug',  'duksoo.1',  'dwarves',  'eager',  'earth',  'earthquake',  'east',  'egypt',  'elections',  \t\t'elevator',  'eliminate',  'elite',  'emory',  'employer.1',  'encourages',  'ends',  'endurance',  'enjoying',  'entertaining',  'entertainment.1',  'entertainment.2',  'enthusiasm',  'equally',  'especially',  \t\t'establishes',  'eternity',  'ethan',  'event',  'events.1',  'excavation',  'excited',  'exerts',  'exes',  'exhibit',  'expert.1',  'fabled',  'facing',  'fairy',  'fake',  'fame',  'family',  'family.1',  \t\t'familyfriend',  'familys',  'fantasies',  'fantasies.1',  'fathers',  'favorites',  'fawkes',  'fears',  'ferocious',  'fighting',  'finn',  'fires',  'flames',  'fleeting',  'flight',  'flip',  'foil',  'follows',  \t\t'footage',  'frank',  'french.1',  'friends',  'friends.1',  'friendship.1',  'fueling',  'fun',  'fund',  'funded',  'future',  'futuristic',  'gandalf',  'gangland',  'gangster',  'gastronomic',  'gatsby',  'gave',  \t\t'gentle',  'georges',  'germany.1',  'getaway',  'gets',  'girls.3',  'gives',  'giving',  'glory',  'goal',  'goldman',  'good',  'gooding',  'grab',  'grace',  'grievers',  'group.1',  'group.2',  'grows',  'guard',  \t\t'guile',  'gutierrez',  'gyllenhaal',  'haars',  'hand',  'hank',  'hannas',  'hard',  'harrowing',  'harsh',  'hazels',  'head',  'heading',  'heart',  'heartland',  'hed',  'helicopter',  'helps',  'henslow',  \t\t'heroes.1',  'hit',  'hivaids',  'hoeks',  'holiday',  'honey',  'hoping',  'hotel',  'hotel.1',  'hotheaded',  'houghton',  'huge',  'humans.2',  'humor',  'humorist',  'huns',  'hunted',  'husband.1',  'ill',  \t\t'illness.1',  'imperial',  'imprison',  'indictment',  'industrialist',  'ineptness',  'inflatable',  'innocent',  'inspire',  'instead',  'intrigue.1',  'invades',  'ironically',  'island',  'japanese',  'jay',  \t\t'jill',  'jims',  'john',  'john.1',  'join',  'jokes',  'jonathan',  'josephine',  'jr',  'judge',  'juliet',  'juliette',  'karen',  'kates',  'katniss',  'katniss.1',  'kenny',  'kidnapped',  'kids.1',  'kills',  \t\t'king',  'kitty.1',  'klieg',  'knee',  'know',  'known',  'koda.1',  'lack',  'lake',  'lands',  'large',  'laurence',  'learn',  'leave',  'leaves',  'leg',  'leopard',  'let',  'leveled',  'lifeand',  'lifeanddeath',  \t\t'lights.1',  'like',  'likes',  'line.1',  'lineaments',  'lingling.2',  'linglings',  'linguistics',  'live',  'lived',  'lizzie.1',  'lok',  'london',  'long',  'long.1',  'looking',  'lookout',  'love',  \t\t'lynd',  'machines.1',  'madness',  'maguire',  'manage',  'manages',  'manolo',  'mans',  'mantis',  'marian',  'marilyn',  'marines',  'marry',  'martins',  'mask',  'massive',  'masterpiece',  'mathis',  \t\t'mayhem',  'mayhem.1',  'mccandless',  'mechanic',  'medications',  'meet',  'meet.1',  'mellark',  'memory',  'men',  'mia',  'miami.1',  'mighty',  'mines',  'mismatched',  'misunderstanding',  'mobsters',  \t\t'money',  'monroe',  'monster',  'montenegro',  'moral',  'mothers',  'moulin',  'mr.1',  'mrgo',  'mummy',  'muscles',  'museum.1',  'music',  'mystery',  'named',  'narcotics',  'nation',  'native',  'near',  \t\t'neighbor',  'neighborhood',  'nerdy',  'nina',  'nondescript',  'novels',  'november',  'oblonsky',  'odds',  'officers',  'operations',  'opposite',  'opposition',  'ops',  'optimus',  'orders',  'outcome.1',  \t\t'outrace',  'overcome',  'owen',  'oxygen',  'pamela',  'parents.1',  'participants',  'partner.2',  'passionate',  'peace',  'peeta.2',  'pension',  'people',  'percys',  'perfectionist',  'perilous',  'perspective',  \t\t'petersburg',  'petersen',  'pharoah',  'physiotherapist',  'pilot',  'place.1',  'plan',  'play.1',  'plot',  'politician',  'portrays',  'pose',  'possible',  'powers',  'prem.1',  'prep',  'prepares',  'present',  \t\t'prey',  'prince',  'professional',  'program',  'promised',  'proper',  'prosthetic',  'protect',  'protected',  'providedriving',  'psychiatrist',  'pursue',  'puts',  'quality',  'questions',  'quickly.1',  'quits',  \t\t'race.1',  'rachel',  'raises',  'rataxes',  'raucous',  'ravaged',  'read',  'readymade',  'realizes',  'recruits.1',  'redeem',  'regina',  'regina.2',  'rekall',  'releases',  'religion',  'renowned',  \t\t'representative',  'resolves',  'result',  'retired',  'retrieve',  'rhinos',  'riding',  'rigours',  'ringmaster',  'rivalry',  'road',  'robbery',  'robert',  'roberts.1',  'robot.1',  'romantic',  'romp',  \t\t'routine',  'royals',  'rules',  'runs',  'russia',  'sabha',  'samuels',  'satisfactorily',  'saved',  'saves',  'saxophone',  'says',  'scott',  'scraggily',  'secrecy',  'secret',  'secrets.1',  'seeing',  \t\t'seeking',  'selfconfessed',  'senator.1',  'sends',  'sergeant',  'set',  'settlement',  'settling',  'sex',  'sexual',  'sexy',  'shakespeare',  'shakespeares',  'shanty',  'shape',  'short',  'simply',  'singing',  \t\t'sings',  'sink',  'sixteen',  'skills',  'slamdunk',  'snow',  'soared',  'sole',  'sort',  'south',  'space',  'spacetime',  'spanish',  'spears',  'spent',  'spontaneity',  'spook.1',  'sports',  'spy',  'stage.1',  \t\t'stand',  'starina',  'start.1',  'states',  'station',  'stepping',  'stooges',  'store',  'stories',  'story',  'strange',  'strangely',  'striking',  'stripes',  'struggle',  'students.1',  'subsequent',  'sudden',  \t\t'suddenly',  'sulley',  'sums',  'sung',  'sung.1',  'superiors',  'sure',  'surrounds',  'survival.1',  'swan.2',  'sweet',  'swim',  'swing',  'sword',  'sympathetic',  'takeover',  'takes',  'talk',  'talks',  \t\t'taunts',  'teacher',  'teachers',  'teams.1',  'teddy',  'televised',  'tell',  'temporal',  'terrorizes',  'tessa',  'tet',  'theres',  'things',  'thorin',  'tiny',  'told',  'toll',  'toro',  'tour.1',  'trade',  \t\t'training',  'traitors',  'travels',  'trials',  'trip.1',  'truck',  'trusty',  'truth',  'twist',  'twisted',  'ultimatum',  'unbeknownst',  'unique',  'unit',  'universal',  'unlocks',  'unrelated',  'uprising',  \t\t'usual',  'various',  'victims',  'victors',  'villain',  'voyage',  'vulnerable',  'wallace',  'wallace.1',  'walt',  'wanted',  'wants',  'war',  'ward',  'wasnt',  'watch',  'water',  'waterhole',  'ways',  \t\t'wazowski',  'wealthy',  'web',  'west.1',  'whats',  'wife.1',  'wildebeest',  'winners',  'wit',  'witches',  'woods',  'words',  'words.1',  'world',  'world.1',  'worlds.1',  'writer',  'writes',  'wrong.1',  \t\t'year',  'years',  'years.1',  'york',  'zebra',  'zellweger',  'zing'], \n",
    "                        ['states']]\n",
    "        self.f_meta = [['country_UK',  'genre_Crime',  'language_Croatian',  'language_Hungarian',  'language_Italian'], \n",
    "                        ['country_China', 'country_Egypt', 'country_France', 'genre_Crime'], \n",
    "                        ['country_Egypt',  'country_France',  'country_Spain',  'genre_Crime',  'genre_Romance',  'language_Greek',  'language_Japanese',  'language_Latin',  'language_Urdu',  'rated_N/A'], \n",
    "                        ['country_Argentina',  'country_France',  'genre_Crime',  'genre_Mystery',  'language_Filipino',  'language_Japanese',  'language_Navajo',  'language_Yiddish'], \n",
    "                        ['genre_Crime'], \n",
    "                        ['country_Argentina',  'country_Ireland',  'country_Italy',  'country_Netherlands',  'genre_Crime',  'genre_Music',  'language_Danish',  'language_Greek',  'language_Mandarin',  'rated_NOT RATED'], \n",
    "                        ['country_France', 'genre_Crime', 'rated_TV-MA'], \n",
    "                        ['country_Bahamas',  'country_France',  'country_SouthAfrica',  'country_SouthKorea',  'genre_Animation',  'genre_Biography',  'genre_Crime',  'genre_Sci-Fi',  'genre_Sport',  'language_Croatian',  'language_Greek'], \n",
    "                        ['country_Argentina',  'country_Egypt',  'genre_Crime',  'genre_Horror',  'language_Hungarian']]\n",
    "        \n",
    "    \n",
    "    def randomSet(self, S,n=0):\n",
    "        Set = []\n",
    "        if (n==0):\n",
    "            n = len(S)\n",
    "        elif (n==1):\n",
    "            n = 2\n",
    "        rand = random.randint(1,n-1)\n",
    "        #print(\"randomset size: \",rand)\n",
    "        return np.random.choice(S, rand ,replace=False)\n",
    "    \n",
    "    def getMBestFeatures(self, D,target, clfs, minf1=0):\n",
    "        scoring = ['precision', 'recall', 'f1']\n",
    "        f_arr=[]\n",
    "        df_scores = pd.DataFrame()\n",
    "        \n",
    "        for clf in clfs:\n",
    "            selected_features = []\n",
    "            n = len(D)\n",
    "            max_tries = 77*n^5\n",
    "            S_best = D.columns.values\n",
    "            C_best = len(S_best)\n",
    "            old_f1 = minf1\n",
    "            \n",
    "            for i in range(1,max_tries):\n",
    "                S = sorted(self.randomSet(D.columns.values, len(D.columns.values)))\n",
    "                C = len(S)\n",
    "                r = cross_validate(clf.clf,D[S],target, cv=10, scoring = scoring)\n",
    "                new_f1 = r['test_f1'].mean()\n",
    "                \n",
    "                if (new_f1 > old_f1):\n",
    "                    print(new_f1)\n",
    "                    S_best = S\n",
    "                    C_best = C\n",
    "                    selected_features = S_best\n",
    "                    old_f1 = new_f1\n",
    "                    print(\"best features: \",S_best)\n",
    "                elif (new_f1==old_f1 and C<C_best):\n",
    "                     S_best = S\n",
    "                     C_best = C\n",
    "                     selected_features = S_best\n",
    "                     print(\"better features: \", S_best)\n",
    "            \n",
    "            f_arr.append(selected_features)\n",
    "            \n",
    "        return f_arr\n",
    "        \n",
    "class OutputWrapper:\n",
    "    \n",
    "    def __init__(self, df_res_test, df_score_test, df_score_stats, df_res_all):\n",
    "        self.df_res_test = df_res_test\n",
    "        self.df_score_test = df_score_test\n",
    "        self.df_score_stats = df_score_stats\n",
    "        self.df_res_all = df_res_all\n",
    "        \n",
    "class Evaluator:\n",
    "    # TODO add clf stacking method\n",
    "    \n",
    "    def __init__(self, data_wrapper, use_audio=True, use_visual=True, use_text=True, use_meta=True):\n",
    "        self.data_wrapper = data_wrapper\n",
    "        self.use_audio = use_audio\n",
    "        self.use_visual = use_visual\n",
    "        self.use_text = use_text\n",
    "        self.use_meta = use_meta            \n",
    "            \n",
    "    def cv_modality(self, df_features, df_targets, clf, cv=10, verbose=True, predict_all = False):\n",
    "        kf = StratifiedKFold(n_splits=cv, random_state=9832432)\n",
    "        df_cvs = []\n",
    "        df_all_cvs = []\n",
    "        i=1\n",
    "        for train_index, test_index in kf.split(df_features, df_targets):\n",
    "            if verbose:\n",
    "                print(\"Performing CV fold {i}..\")\n",
    "            i += 1\n",
    "            X_train, X_test = df_features.iloc[train_index,:], df_features.iloc[test_index,:]\n",
    "            y_train, y_test = df_targets[train_index], df_targets[test_index]\n",
    "            \n",
    "            df_res = pd.DataFrame(y_test)\n",
    "            df_res.columns = ['TARGET']\n",
    "            \n",
    "            # init df for prediction of all entries\n",
    "            df_res_all = pd.DataFrame(df_targets)\n",
    "            df_res_all.columns = ['TARGET']\n",
    "            \n",
    "            #for clf in clfs:\n",
    "            clf.fit(X_train, y_train, verbose)\n",
    "            y_pred = clf.predict(X_test, verbose)\n",
    "            df_res[clf.clf_name+\"_\"+clf.modality] = y_pred\n",
    "\n",
    "            if predict_all:\n",
    "                y_pred_all = clf.predict(df_features, verbose)\n",
    "                df_res_all[clf.clf_name+\"_\"+clf.modality] = y_pred_all\n",
    "                \n",
    "            df_cvs.append(df_res)\n",
    "            df_all_cvs.append(df_res_all)\n",
    "            \n",
    "        return df_cvs, df_all_cvs\n",
    "        \n",
    "        return df_cvs, df_all_cvs\n",
    "            \n",
    "    def cv(self, cv=10, verbose=True, predict_all=False):\n",
    "        \"\"\"\n",
    "            set predict_all to True to also include predictions for all data\n",
    "        \"\"\"\n",
    "        df_cvs = []\n",
    "        df_all_cvs = []\n",
    "        df_targets = self.data_wrapper.df_targets\n",
    "        if (self.use_audio):\n",
    "            part_audio = []\n",
    "            clfs = ClassifierFactory.get_audio_classifiers()\n",
    "            for i, clf in enumerate(clfs):\n",
    "                df_test, df_all = self.cv_modality(\n",
    "                    self.data_wrapper.df_audio[self.data_wrapper.f_audio[i]],\n",
    "                    df_targets,\n",
    "                    clf,\n",
    "                    cv,\n",
    "                    verbose,\n",
    "                    predict_all)\n",
    "                df_cvs.append(df_test)\n",
    "                df_all_cvs.append(df_all)\n",
    "        if (self.use_visual):\n",
    "            clfs = ClassifierFactory.get_visual_classifiers()\n",
    "            for i, clf in enumerate(clfs):\n",
    "                df_test, df_all = self.cv_modality(\n",
    "                    self.data_wrapper.df_vis[self.data_wrapper.f_vis[i]],\n",
    "                    df_targets,\n",
    "                    clf,\n",
    "                    cv,\n",
    "                    verbose,\n",
    "                    predict_all)\n",
    "                df_cvs.append(df_test)\n",
    "                df_all_cvs.append(df_all)\n",
    "        if (self.use_text):\n",
    "            clfs = ClassifierFactory.get_text_classifiers()\n",
    "            for i, clf in enumerate(clfs):\n",
    "                df_test, df_all = self.cv_modality(\n",
    "                    self.data_wrapper.df_txt[self.data_wrapper.f_txt[i]],\n",
    "                    df_targets,\n",
    "                    clf,\n",
    "                    cv,\n",
    "                    verbose,\n",
    "                    predict_all)\n",
    "                df_cvs.append(df_test)\n",
    "                df_all_cvs.append(df_all)\n",
    "        if (self.use_meta):\n",
    "            clfs = ClassifierFactory.get_metadata_classifiers()\n",
    "            for i, clf in enumerate(clfs):\n",
    "                df_test, df_all = self.cv_modality(\n",
    "                    self.data_wrapper.df_meta[self.data_wrapper.f_meta[i]],\n",
    "                    df_targets,\n",
    "                    clf,\n",
    "                    cv,\n",
    "                    verbose,\n",
    "                    predict_all)\n",
    "                df_cvs.append(df_test)\n",
    "                df_all_cvs.append(df_all)\n",
    "        \n",
    "        for i in range(0,10):\n",
    "            for j in range(1,21):\n",
    "                df_cvs[j][i]=df_cvs[j][i].drop(['TARGET'],axis = 1)\n",
    "                df_cvs[0][i]=pd.merge(df_cvs[0][i], df_cvs[j][i], left_index=True, right_index=True)\n",
    "                df_all_cvs[j][i]=df_all_cvs[j][i].drop(['TARGET'],axis = 1)\n",
    "                df_all_cvs[0][i]=pd.merge(df_all_cvs[0][i], df_all_cvs[j][i], left_index=True, right_index=True)\n",
    "        \n",
    "        df_c = df_cvs[0]\n",
    "        df_all_c = df_all_cvs[0]\n",
    "            \n",
    "        df_res = self.evaluate(df_c)\n",
    "        \n",
    "        return OutputWrapper(df_c, df_res, self.overall_eval(df_res), df_all_c)\n",
    "\n",
    "    def evaluate(self, df_res):\n",
    "        \"\"\"\n",
    "        returns precision, recall and F1 in a DF\n",
    "        returns list of DFs if df_res is list of DFs\n",
    "        \"\"\"\n",
    "        if type(df_res) == type([]):\n",
    "            return [self.evaluate_single(df_x) for df_x in df_res]\n",
    "        else:\n",
    "            return self.evaluate_single(df_res)\n",
    "\n",
    "    def evaluate_single(self, df_res):\n",
    "        cols = list(df_res.columns)\n",
    "        cols.remove('TARGET')\n",
    "        df_ev = pd.DataFrame(columns=['clf', 'precision', 'recall', 'f1'])\n",
    "        for col in cols:\n",
    "            prec = precision_score(df_res['TARGET'], df_res[col])\n",
    "            recall = recall_score(df_res['TARGET'], df_res[col])\n",
    "            f1 = f1_score(df_res['TARGET'], df_res[col])\n",
    "            df_ev = df_ev.append({'clf':col, 'precision': prec, 'recall':recall, 'f1':f1}, ignore_index=True)\n",
    "        return df_ev\n",
    "    \n",
    "    def overall_eval(self, df_results):\n",
    "        df_score = pd.DataFrame(columns =['clf', 'mean_precision', 'mean_recall', 'mean_f1', 'var_precision', 'var_recall', 'var_f1', 'std_precision', 'std_recall', 'std_f1'])\n",
    "        df_score['clf'] = df_results[0]['clf']\n",
    "        # mean \n",
    "        df_res = reduce(lambda x, y: x.add(y, fill_value=0), df_results)\n",
    "        df_score[['mean_precision', 'mean_recall', 'mean_f1']] = df_res[['precision', 'recall', 'f1']].div(len(df_results))\n",
    "\n",
    "        # var\n",
    "        df_mean = df_score[['mean_precision', 'mean_recall', 'mean_f1']]\n",
    "        df_mean.columns = ['precision', 'recall', 'f1']\n",
    "        df_dev = []\n",
    "        for df in df_results:\n",
    "            df_dev.append(np.square(df[['precision', 'recall', 'f1']].subtract(df_mean)))\n",
    "        df_dev = reduce(lambda x, y: x.add(y, fill_value=0), df_dev)\n",
    "        df_score[['var_precision', 'var_recall', 'var_f1']] = df_dev[['precision', 'recall', 'f1']].div(len(df_results))\n",
    "\n",
    "        # std \n",
    "        df_score[['std_precision', 'std_recall', 'std_f1']] = np.power(df_score[['var_precision', 'var_recall', 'var_f1']], 0.5)\n",
    "        \n",
    "        # 95-ci\n",
    "        for metric in ['precision', 'recall', 'f1']:\n",
    "            l_ci = df_score['mean_'+metric] - df_score['std_'+metric]*CI_95_FACTOR/len(df_results)\n",
    "            u_ci = df_score['mean_'+metric] + df_score['std_'+metric]*CI_95_FACTOR/len(df_results)\n",
    "            df_score['l_95ci_'+metric] = l_ci\n",
    "            df_score['u_95ci_'+metric] = u_ci\n",
    "\n",
    "        return df_score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Call Example"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "warnings.filterwarnings(action='ignore')\n",
    "dw = DataWrapper(df_audio_train, df_vis_train, df_txt_train, df_meta_train, df_targets_train)\n",
    "dw.generate_subspace()\n",
    "ev = Evaluator(dw) # TODO preprocessing for meta needs to be done\n",
    "ow = ev.cv(verbose=False, predict_all=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>TARGET</th>\n",
       "      <th>Logistic regression_Audio</th>\n",
       "      <th>Gradient Boosting Tree_Audio</th>\n",
       "      <th>k-Nearest neighbors_Visual</th>\n",
       "      <th>Decision tree_Visual</th>\n",
       "      <th>Logistic regression_Visual</th>\n",
       "      <th>SVM (Gaussian Kernel)_Visual</th>\n",
       "      <th>Random Forest_Visual</th>\n",
       "      <th>AdaBoost_Visual</th>\n",
       "      <th>Gradient Boosting Tree_Visual</th>\n",
       "      <th>...</th>\n",
       "      <th>SVM (Gaussian Kernel)_Textual</th>\n",
       "      <th>k-Nearest neighbors_Metadata</th>\n",
       "      <th>Nearest mean classifier_Metadata</th>\n",
       "      <th>Decision tree_Metadata</th>\n",
       "      <th>Logistic regression_Metadata</th>\n",
       "      <th>SVM (Gaussian Kernel)_Metadata</th>\n",
       "      <th>Bagging_Metadata</th>\n",
       "      <th>Random Forest_Metadata</th>\n",
       "      <th>AdaBoost_Metadata</th>\n",
       "      <th>Gradient Boosting Tree_Metadata</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    TARGET  Logistic regression_Audio  Gradient Boosting Tree_Audio  \\\n",
       "0        1                          1                             0   \n",
       "1        0                          1                             0   \n",
       "2        1                          1                             1   \n",
       "3        1                          1                             1   \n",
       "4        1                          1                             1   \n",
       "5        1                          1                             0   \n",
       "6        1                          1                             1   \n",
       "8        0                          1                             0   \n",
       "9        0                          1                             1   \n",
       "13       0                          1                             1   \n",
       "14       0                          1                             1   \n",
       "\n",
       "    k-Nearest neighbors_Visual  Decision tree_Visual  \\\n",
       "0                            1                     0   \n",
       "1                            0                     0   \n",
       "2                            1                     1   \n",
       "3                            1                     0   \n",
       "4                            1                     0   \n",
       "5                            1                     1   \n",
       "6                            1                     1   \n",
       "8                            0                     0   \n",
       "9                            0                     1   \n",
       "13                           1                     0   \n",
       "14                           0                     1   \n",
       "\n",
       "    Logistic regression_Visual  SVM (Gaussian Kernel)_Visual  \\\n",
       "0                            1                             1   \n",
       "1                            0                             1   \n",
       "2                            1                             1   \n",
       "3                            1                             0   \n",
       "4                            0                             1   \n",
       "5                            1                             1   \n",
       "6                            1                             1   \n",
       "8                            1                             0   \n",
       "9                            1                             0   \n",
       "13                           1                             0   \n",
       "14                           1                             0   \n",
       "\n",
       "    Random Forest_Visual  AdaBoost_Visual  Gradient Boosting Tree_Visual  \\\n",
       "0                      1                1                              1   \n",
       "1                      1                0                              0   \n",
       "2                      1                1                              1   \n",
       "3                      0                1                              1   \n",
       "4                      0                1                              1   \n",
       "5                      1                1                              1   \n",
       "6                      1                1                              0   \n",
       "8                      0                0                              0   \n",
       "9                      0                1                              0   \n",
       "13                     1                1                              0   \n",
       "14                     0                0                              0   \n",
       "\n",
       "                 ...                 SVM (Gaussian Kernel)_Textual  \\\n",
       "0                ...                                             1   \n",
       "1                ...                                             1   \n",
       "2                ...                                             1   \n",
       "3                ...                                             1   \n",
       "4                ...                                             1   \n",
       "5                ...                                             1   \n",
       "6                ...                                             1   \n",
       "8                ...                                             1   \n",
       "9                ...                                             1   \n",
       "13               ...                                             1   \n",
       "14               ...                                             1   \n",
       "\n",
       "    k-Nearest neighbors_Metadata  Nearest mean classifier_Metadata  \\\n",
       "0                              1                                 0   \n",
       "1                              1                                 1   \n",
       "2                              1                                 1   \n",
       "3                              1                                 1   \n",
       "4                              1                                 0   \n",
       "5                              1                                 1   \n",
       "6                              1                                 1   \n",
       "8                              1                                 0   \n",
       "9                              1                                 1   \n",
       "13                             1                                 1   \n",
       "14                             1                                 0   \n",
       "\n",
       "    Decision tree_Metadata  Logistic regression_Metadata  \\\n",
       "0                        0                             0   \n",
       "1                        1                             1   \n",
       "2                        1                             1   \n",
       "3                        1                             1   \n",
       "4                        0                             0   \n",
       "5                        1                             1   \n",
       "6                        1                             1   \n",
       "8                        0                             0   \n",
       "9                        1                             1   \n",
       "13                       1                             1   \n",
       "14                       0                             0   \n",
       "\n",
       "    SVM (Gaussian Kernel)_Metadata  Bagging_Metadata  Random Forest_Metadata  \\\n",
       "0                                0                 0                       0   \n",
       "1                                1                 1                       1   \n",
       "2                                1                 1                       1   \n",
       "3                                1                 1                       1   \n",
       "4                                0                 0                       0   \n",
       "5                                1                 1                       1   \n",
       "6                                1                 1                       1   \n",
       "8                                1                 1                       0   \n",
       "9                                1                 1                       1   \n",
       "13                               1                 1                       1   \n",
       "14                               0                 0                       0   \n",
       "\n",
       "    AdaBoost_Metadata  Gradient Boosting Tree_Metadata  \n",
       "0                   0                                0  \n",
       "1                   1                                1  \n",
       "2                   1                                1  \n",
       "3                   1                                1  \n",
       "4                   0                                0  \n",
       "5                   1                                1  \n",
       "6                   1                                1  \n",
       "8                   0                                1  \n",
       "9                   1                                1  \n",
       "13                  1                                1  \n",
       "14                  0                                0  \n",
       "\n",
       "[11 rows x 22 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ow.df_res_test[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[                                 clf  precision    recall        f1\n",
       " 0          Logistic regression_Audio   0.545455  1.000000  0.705882\n",
       " 1       Gradient Boosting Tree_Audio   0.571429  0.666667  0.615385\n",
       " 2         k-Nearest neighbors_Visual   0.857143  1.000000  0.923077\n",
       " 3               Decision tree_Visual   0.600000  0.500000  0.545455\n",
       " 4         Logistic regression_Visual   0.555556  0.833333  0.666667\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.833333  0.833333  0.833333\n",
       " 6               Random Forest_Visual   0.666667  0.666667  0.666667\n",
       " 7                    AdaBoost_Visual   0.750000  1.000000  0.857143\n",
       " 8      Gradient Boosting Tree_Visual   1.000000  0.833333  0.909091\n",
       " 9                Naive Bayes_Textual   0.750000  0.500000  0.600000\n",
       " 10       k-Nearest neighbors_Textual   0.545455  1.000000  0.705882\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.545455  1.000000  0.705882\n",
       " 12      k-Nearest neighbors_Metadata   0.545455  1.000000  0.705882\n",
       " 13  Nearest mean classifier_Metadata   0.571429  0.666667  0.615385\n",
       " 14            Decision tree_Metadata   0.571429  0.666667  0.615385\n",
       " 15      Logistic regression_Metadata   0.571429  0.666667  0.615385\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.500000  0.666667  0.571429\n",
       " 17                  Bagging_Metadata   0.500000  0.666667  0.571429\n",
       " 18            Random Forest_Metadata   0.571429  0.666667  0.615385\n",
       " 19                 AdaBoost_Metadata   0.571429  0.666667  0.615385\n",
       " 20   Gradient Boosting Tree_Metadata   0.500000  0.666667  0.571429,\n",
       "                                  clf  precision    recall        f1\n",
       " 0          Logistic regression_Audio   0.545455  1.000000  0.705882\n",
       " 1       Gradient Boosting Tree_Audio   1.000000  0.666667  0.800000\n",
       " 2         k-Nearest neighbors_Visual   0.666667  0.666667  0.666667\n",
       " 3               Decision tree_Visual   0.571429  0.666667  0.615385\n",
       " 4         Logistic regression_Visual   0.833333  0.833333  0.833333\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.500000  0.833333  0.625000\n",
       " 6               Random Forest_Visual   0.666667  0.666667  0.666667\n",
       " 7                    AdaBoost_Visual   0.428571  0.500000  0.461538\n",
       " 8      Gradient Boosting Tree_Visual   0.666667  0.666667  0.666667\n",
       " 9                Naive Bayes_Textual   0.555556  0.833333  0.666667\n",
       " 10       k-Nearest neighbors_Textual   0.750000  1.000000  0.857143\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.545455  1.000000  0.705882\n",
       " 12      k-Nearest neighbors_Metadata   0.600000  1.000000  0.750000\n",
       " 13  Nearest mean classifier_Metadata   0.600000  1.000000  0.750000\n",
       " 14            Decision tree_Metadata   0.666667  1.000000  0.800000\n",
       " 15      Logistic regression_Metadata   0.666667  1.000000  0.800000\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.600000  1.000000  0.750000\n",
       " 17                  Bagging_Metadata   0.666667  1.000000  0.800000\n",
       " 18            Random Forest_Metadata   0.600000  1.000000  0.750000\n",
       " 19                 AdaBoost_Metadata   0.600000  1.000000  0.750000\n",
       " 20   Gradient Boosting Tree_Metadata   0.666667  1.000000  0.800000,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.500000     1.0  0.666667\n",
       " 1       Gradient Boosting Tree_Audio   0.500000     0.6  0.545455\n",
       " 2         k-Nearest neighbors_Visual   0.571429     0.8  0.666667\n",
       " 3               Decision tree_Visual   0.666667     0.8  0.727273\n",
       " 4         Logistic regression_Visual   0.625000     1.0  0.769231\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.625000     1.0  0.769231\n",
       " 6               Random Forest_Visual   0.500000     0.6  0.545455\n",
       " 7                    AdaBoost_Visual   0.600000     0.6  0.600000\n",
       " 8      Gradient Boosting Tree_Visual   0.625000     1.0  0.769231\n",
       " 9                Naive Bayes_Textual   0.666667     0.4  0.500000\n",
       " 10       k-Nearest neighbors_Textual   0.500000     1.0  0.666667\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.500000     1.0  0.666667\n",
       " 12      k-Nearest neighbors_Metadata   0.555556     1.0  0.714286\n",
       " 13  Nearest mean classifier_Metadata   0.555556     1.0  0.714286\n",
       " 14            Decision tree_Metadata   0.571429     0.8  0.666667\n",
       " 15      Logistic regression_Metadata   0.555556     1.0  0.714286\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.555556     1.0  0.714286\n",
       " 17                  Bagging_Metadata   0.555556     1.0  0.714286\n",
       " 18            Random Forest_Metadata   0.555556     1.0  0.714286\n",
       " 19                 AdaBoost_Metadata   0.555556     1.0  0.714286\n",
       " 20   Gradient Boosting Tree_Metadata   0.555556     1.0  0.714286,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.714286     1.0  0.833333\n",
       " 2         k-Nearest neighbors_Visual   0.625000     1.0  0.769231\n",
       " 3               Decision tree_Visual   0.500000     0.6  0.545455\n",
       " 4         Logistic regression_Visual   0.666667     0.8  0.727273\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.571429     0.8  0.666667\n",
       " 6               Random Forest_Visual   0.500000     0.6  0.545455\n",
       " 7                    AdaBoost_Visual   0.800000     0.8  0.800000\n",
       " 8      Gradient Boosting Tree_Visual   0.800000     0.8  0.800000\n",
       " 9                Naive Bayes_Textual   0.800000     0.8  0.800000\n",
       " 10       k-Nearest neighbors_Textual   0.625000     1.0  0.769231\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.625000     1.0  0.769231\n",
       " 13  Nearest mean classifier_Metadata   0.625000     1.0  0.769231\n",
       " 14            Decision tree_Metadata   0.714286     1.0  0.833333\n",
       " 15      Logistic regression_Metadata   0.625000     1.0  0.769231\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.625000     1.0  0.769231\n",
       " 17                  Bagging_Metadata   0.625000     1.0  0.769231\n",
       " 18            Random Forest_Metadata   0.625000     1.0  0.769231\n",
       " 19                 AdaBoost_Metadata   0.625000     1.0  0.769231\n",
       " 20   Gradient Boosting Tree_Metadata   0.625000     1.0  0.769231,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.666667     0.8  0.727273\n",
       " 2         k-Nearest neighbors_Visual   0.600000     0.6  0.600000\n",
       " 3               Decision tree_Visual   0.600000     0.6  0.600000\n",
       " 4         Logistic regression_Visual   0.600000     0.6  0.600000\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.833333     1.0  0.909091\n",
       " 6               Random Forest_Visual   0.500000     0.4  0.444444\n",
       " 7                    AdaBoost_Visual   0.833333     1.0  0.909091\n",
       " 8      Gradient Boosting Tree_Visual   1.000000     1.0  1.000000\n",
       " 9                Naive Bayes_Textual   0.833333     1.0  0.909091\n",
       " 10       k-Nearest neighbors_Textual   0.500000     0.6  0.545455\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.555556     1.0  0.714286\n",
       " 13  Nearest mean classifier_Metadata   0.555556     1.0  0.714286\n",
       " 14            Decision tree_Metadata   0.625000     1.0  0.769231\n",
       " 15      Logistic regression_Metadata   0.555556     1.0  0.714286\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.555556     1.0  0.714286\n",
       " 17                  Bagging_Metadata   0.555556     1.0  0.714286\n",
       " 18            Random Forest_Metadata   0.555556     1.0  0.714286\n",
       " 19                 AdaBoost_Metadata   0.555556     1.0  0.714286\n",
       " 20   Gradient Boosting Tree_Metadata   0.555556     1.0  0.714286,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.666667     0.8  0.727273\n",
       " 2         k-Nearest neighbors_Visual   1.000000     0.6  0.750000\n",
       " 3               Decision tree_Visual   0.750000     0.6  0.666667\n",
       " 4         Logistic regression_Visual   0.666667     0.8  0.727273\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.555556     1.0  0.714286\n",
       " 6               Random Forest_Visual   0.500000     0.2  0.285714\n",
       " 7                    AdaBoost_Visual   0.800000     0.8  0.800000\n",
       " 8      Gradient Boosting Tree_Visual   0.666667     0.8  0.727273\n",
       " 9                Naive Bayes_Textual   0.428571     0.6  0.500000\n",
       " 10       k-Nearest neighbors_Textual   0.625000     1.0  0.769231\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.833333     1.0  0.909091\n",
       " 13  Nearest mean classifier_Metadata   0.833333     1.0  0.909091\n",
       " 14            Decision tree_Metadata   0.833333     1.0  0.909091\n",
       " 15      Logistic regression_Metadata   0.833333     1.0  0.909091\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.833333     1.0  0.909091\n",
       " 17                  Bagging_Metadata   0.833333     1.0  0.909091\n",
       " 18            Random Forest_Metadata   0.833333     1.0  0.909091\n",
       " 19                 AdaBoost_Metadata   0.833333     1.0  0.909091\n",
       " 20   Gradient Boosting Tree_Metadata   0.833333     1.0  0.909091,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.800000     0.8  0.800000\n",
       " 2         k-Nearest neighbors_Visual   1.000000     1.0  1.000000\n",
       " 3               Decision tree_Visual   0.333333     0.2  0.250000\n",
       " 4         Logistic regression_Visual   1.000000     0.8  0.888889\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.625000     1.0  0.769231\n",
       " 6               Random Forest_Visual   0.500000     0.6  0.545455\n",
       " 7                    AdaBoost_Visual   0.750000     0.6  0.666667\n",
       " 8      Gradient Boosting Tree_Visual   0.800000     0.8  0.800000\n",
       " 9                Naive Bayes_Textual   1.000000     0.8  0.888889\n",
       " 10       k-Nearest neighbors_Textual   0.555556     1.0  0.714286\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.625000     1.0  0.769231\n",
       " 13  Nearest mean classifier_Metadata   0.714286     1.0  0.833333\n",
       " 14            Decision tree_Metadata   0.714286     1.0  0.833333\n",
       " 15      Logistic regression_Metadata   0.714286     1.0  0.833333\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.625000     1.0  0.769231\n",
       " 17                  Bagging_Metadata   0.625000     1.0  0.769231\n",
       " 18            Random Forest_Metadata   0.714286     1.0  0.833333\n",
       " 19                 AdaBoost_Metadata   0.714286     1.0  0.833333\n",
       " 20   Gradient Boosting Tree_Metadata   0.625000     1.0  0.769231,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.750000     0.6  0.666667\n",
       " 2         k-Nearest neighbors_Visual   0.833333     1.0  0.909091\n",
       " 3               Decision tree_Visual   0.833333     1.0  0.909091\n",
       " 4         Logistic regression_Visual   0.714286     1.0  0.833333\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.625000     1.0  0.769231\n",
       " 6               Random Forest_Visual   0.666667     0.8  0.727273\n",
       " 7                    AdaBoost_Visual   1.000000     1.0  1.000000\n",
       " 8      Gradient Boosting Tree_Visual   0.625000     1.0  0.769231\n",
       " 9                Naive Bayes_Textual   0.600000     0.6  0.600000\n",
       " 10       k-Nearest neighbors_Textual   0.625000     1.0  0.769231\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.625000     1.0  0.769231\n",
       " 13  Nearest mean classifier_Metadata   0.625000     1.0  0.769231\n",
       " 14            Decision tree_Metadata   0.625000     1.0  0.769231\n",
       " 15      Logistic regression_Metadata   0.625000     1.0  0.769231\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.625000     1.0  0.769231\n",
       " 17                  Bagging_Metadata   0.625000     1.0  0.769231\n",
       " 18            Random Forest_Metadata   0.625000     1.0  0.769231\n",
       " 19                 AdaBoost_Metadata   0.625000     1.0  0.769231\n",
       " 20   Gradient Boosting Tree_Metadata   0.625000     1.0  0.769231,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.500000     0.4  0.444444\n",
       " 2         k-Nearest neighbors_Visual   0.571429     0.8  0.666667\n",
       " 3               Decision tree_Visual   0.666667     0.8  0.727273\n",
       " 4         Logistic regression_Visual   0.666667     0.8  0.727273\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.500000     0.8  0.615385\n",
       " 6               Random Forest_Visual   0.500000     0.4  0.444444\n",
       " 7                    AdaBoost_Visual   0.666667     0.8  0.727273\n",
       " 8      Gradient Boosting Tree_Visual   0.600000     0.6  0.600000\n",
       " 9                Naive Bayes_Textual   0.800000     0.8  0.800000\n",
       " 10       k-Nearest neighbors_Textual   0.800000     0.8  0.800000\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.555556     1.0  0.714286\n",
       " 13  Nearest mean classifier_Metadata   0.555556     1.0  0.714286\n",
       " 14            Decision tree_Metadata   0.555556     1.0  0.714286\n",
       " 15      Logistic regression_Metadata   0.555556     1.0  0.714286\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.555556     1.0  0.714286\n",
       " 17                  Bagging_Metadata   0.555556     1.0  0.714286\n",
       " 18            Random Forest_Metadata   0.555556     1.0  0.714286\n",
       " 19                 AdaBoost_Metadata   0.555556     1.0  0.714286\n",
       " 20   Gradient Boosting Tree_Metadata   0.555556     1.0  0.714286,\n",
       "                                  clf  precision  recall        f1\n",
       " 0          Logistic regression_Audio   0.555556     1.0  0.714286\n",
       " 1       Gradient Boosting Tree_Audio   0.666667     0.8  0.727273\n",
       " 2         k-Nearest neighbors_Visual   0.800000     0.8  0.800000\n",
       " 3               Decision tree_Visual   0.600000     0.6  0.600000\n",
       " 4         Logistic regression_Visual   0.750000     0.6  0.666667\n",
       " 5       SVM (Gaussian Kernel)_Visual   0.555556     1.0  0.714286\n",
       " 6               Random Forest_Visual   0.333333     0.2  0.250000\n",
       " 7                    AdaBoost_Visual   0.750000     0.6  0.666667\n",
       " 8      Gradient Boosting Tree_Visual   1.000000     0.6  0.750000\n",
       " 9                Naive Bayes_Textual   1.000000     0.8  0.888889\n",
       " 10       k-Nearest neighbors_Textual   0.833333     1.0  0.909091\n",
       " 11     SVM (Gaussian Kernel)_Textual   0.555556     1.0  0.714286\n",
       " 12      k-Nearest neighbors_Metadata   0.555556     1.0  0.714286\n",
       " 13  Nearest mean classifier_Metadata   0.555556     1.0  0.714286\n",
       " 14            Decision tree_Metadata   0.625000     1.0  0.769231\n",
       " 15      Logistic regression_Metadata   0.625000     1.0  0.769231\n",
       " 16    SVM (Gaussian Kernel)_Metadata   0.555556     1.0  0.714286\n",
       " 17                  Bagging_Metadata   0.625000     1.0  0.769231\n",
       " 18            Random Forest_Metadata   0.555556     1.0  0.714286\n",
       " 19                 AdaBoost_Metadata   0.555556     1.0  0.714286\n",
       " 20   Gradient Boosting Tree_Metadata   0.625000     1.0  0.769231]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ow.df_score_test        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>clf</th>\n",
       "      <th>mean_precision</th>\n",
       "      <th>mean_recall</th>\n",
       "      <th>mean_f1</th>\n",
       "      <th>var_precision</th>\n",
       "      <th>var_recall</th>\n",
       "      <th>var_f1</th>\n",
       "      <th>std_precision</th>\n",
       "      <th>std_recall</th>\n",
       "      <th>std_f1</th>\n",
       "      <th>l_95ci_precision</th>\n",
       "      <th>u_95ci_precision</th>\n",
       "      <th>l_95ci_recall</th>\n",
       "      <th>u_95ci_recall</th>\n",
       "      <th>l_95ci_f1</th>\n",
       "      <th>u_95ci_f1</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Logistic regression_Audio</td>\n",
       "      <td>0.547980</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.707843</td>\n",
       "      <td>0.000272</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000199</td>\n",
       "      <td>0.016482</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014120</td>\n",
       "      <td>0.544749</td>\n",
       "      <td>0.551210</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.705076</td>\n",
       "      <td>0.710611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Gradient Boosting Tree_Audio</td>\n",
       "      <td>0.683571</td>\n",
       "      <td>0.713333</td>\n",
       "      <td>0.688710</td>\n",
       "      <td>0.019987</td>\n",
       "      <td>0.024044</td>\n",
       "      <td>0.013620</td>\n",
       "      <td>0.141375</td>\n",
       "      <td>0.155063</td>\n",
       "      <td>0.116704</td>\n",
       "      <td>0.655862</td>\n",
       "      <td>0.711281</td>\n",
       "      <td>0.682941</td>\n",
       "      <td>0.743726</td>\n",
       "      <td>0.665836</td>\n",
       "      <td>0.711584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>k-Nearest neighbors_Visual</td>\n",
       "      <td>0.752500</td>\n",
       "      <td>0.826667</td>\n",
       "      <td>0.775140</td>\n",
       "      <td>0.025471</td>\n",
       "      <td>0.025067</td>\n",
       "      <td>0.015765</td>\n",
       "      <td>0.159595</td>\n",
       "      <td>0.158325</td>\n",
       "      <td>0.125558</td>\n",
       "      <td>0.721219</td>\n",
       "      <td>0.783781</td>\n",
       "      <td>0.795635</td>\n",
       "      <td>0.857698</td>\n",
       "      <td>0.750530</td>\n",
       "      <td>0.799749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Decision tree_Visual</td>\n",
       "      <td>0.612143</td>\n",
       "      <td>0.636667</td>\n",
       "      <td>0.618660</td>\n",
       "      <td>0.016629</td>\n",
       "      <td>0.040100</td>\n",
       "      <td>0.025758</td>\n",
       "      <td>0.128952</td>\n",
       "      <td>0.200250</td>\n",
       "      <td>0.160494</td>\n",
       "      <td>0.586868</td>\n",
       "      <td>0.637417</td>\n",
       "      <td>0.597418</td>\n",
       "      <td>0.675916</td>\n",
       "      <td>0.587203</td>\n",
       "      <td>0.650117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Logistic regression_Visual</td>\n",
       "      <td>0.707817</td>\n",
       "      <td>0.806667</td>\n",
       "      <td>0.743994</td>\n",
       "      <td>0.014969</td>\n",
       "      <td>0.016178</td>\n",
       "      <td>0.007113</td>\n",
       "      <td>0.122349</td>\n",
       "      <td>0.127192</td>\n",
       "      <td>0.084337</td>\n",
       "      <td>0.683837</td>\n",
       "      <td>0.731798</td>\n",
       "      <td>0.781737</td>\n",
       "      <td>0.831596</td>\n",
       "      <td>0.727464</td>\n",
       "      <td>0.760524</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>SVM (Gaussian Kernel)_Visual</td>\n",
       "      <td>0.622421</td>\n",
       "      <td>0.926667</td>\n",
       "      <td>0.738574</td>\n",
       "      <td>0.013050</td>\n",
       "      <td>0.008178</td>\n",
       "      <td>0.007530</td>\n",
       "      <td>0.114238</td>\n",
       "      <td>0.090431</td>\n",
       "      <td>0.086776</td>\n",
       "      <td>0.600030</td>\n",
       "      <td>0.644811</td>\n",
       "      <td>0.908942</td>\n",
       "      <td>0.944391</td>\n",
       "      <td>0.721566</td>\n",
       "      <td>0.755582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Random Forest_Visual</td>\n",
       "      <td>0.533333</td>\n",
       "      <td>0.513333</td>\n",
       "      <td>0.512157</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.037378</td>\n",
       "      <td>0.022652</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.193333</td>\n",
       "      <td>0.150506</td>\n",
       "      <td>0.513733</td>\n",
       "      <td>0.552933</td>\n",
       "      <td>0.475440</td>\n",
       "      <td>0.551227</td>\n",
       "      <td>0.482658</td>\n",
       "      <td>0.541656</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>AdaBoost_Visual</td>\n",
       "      <td>0.737857</td>\n",
       "      <td>0.770000</td>\n",
       "      <td>0.748838</td>\n",
       "      <td>0.020573</td>\n",
       "      <td>0.032100</td>\n",
       "      <td>0.022439</td>\n",
       "      <td>0.143433</td>\n",
       "      <td>0.179165</td>\n",
       "      <td>0.149797</td>\n",
       "      <td>0.709744</td>\n",
       "      <td>0.765970</td>\n",
       "      <td>0.734884</td>\n",
       "      <td>0.805116</td>\n",
       "      <td>0.719478</td>\n",
       "      <td>0.778198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Gradient Boosting Tree_Visual</td>\n",
       "      <td>0.778333</td>\n",
       "      <td>0.810000</td>\n",
       "      <td>0.779149</td>\n",
       "      <td>0.025211</td>\n",
       "      <td>0.021789</td>\n",
       "      <td>0.011501</td>\n",
       "      <td>0.158780</td>\n",
       "      <td>0.147611</td>\n",
       "      <td>0.107244</td>\n",
       "      <td>0.747212</td>\n",
       "      <td>0.809454</td>\n",
       "      <td>0.781068</td>\n",
       "      <td>0.838932</td>\n",
       "      <td>0.758129</td>\n",
       "      <td>0.800169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Naive Bayes_Textual</td>\n",
       "      <td>0.743413</td>\n",
       "      <td>0.713333</td>\n",
       "      <td>0.715354</td>\n",
       "      <td>0.030708</td>\n",
       "      <td>0.029600</td>\n",
       "      <td>0.023383</td>\n",
       "      <td>0.175237</td>\n",
       "      <td>0.172047</td>\n",
       "      <td>0.152915</td>\n",
       "      <td>0.709066</td>\n",
       "      <td>0.777759</td>\n",
       "      <td>0.679612</td>\n",
       "      <td>0.747054</td>\n",
       "      <td>0.685382</td>\n",
       "      <td>0.745325</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>k-Nearest neighbors_Textual</td>\n",
       "      <td>0.635934</td>\n",
       "      <td>0.940000</td>\n",
       "      <td>0.750622</td>\n",
       "      <td>0.013086</td>\n",
       "      <td>0.016400</td>\n",
       "      <td>0.009240</td>\n",
       "      <td>0.114393</td>\n",
       "      <td>0.128062</td>\n",
       "      <td>0.096125</td>\n",
       "      <td>0.613513</td>\n",
       "      <td>0.658355</td>\n",
       "      <td>0.914900</td>\n",
       "      <td>0.965100</td>\n",
       "      <td>0.731781</td>\n",
       "      <td>0.769462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>SVM (Gaussian Kernel)_Textual</td>\n",
       "      <td>0.547980</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.707843</td>\n",
       "      <td>0.000272</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000199</td>\n",
       "      <td>0.016482</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.014120</td>\n",
       "      <td>0.544749</td>\n",
       "      <td>0.551210</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.705076</td>\n",
       "      <td>0.710611</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>k-Nearest neighbors_Metadata</td>\n",
       "      <td>0.607601</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.752981</td>\n",
       "      <td>0.006662</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.003338</td>\n",
       "      <td>0.081620</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.057775</td>\n",
       "      <td>0.591604</td>\n",
       "      <td>0.623599</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.741657</td>\n",
       "      <td>0.764305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Nearest mean classifier_Metadata</td>\n",
       "      <td>0.619127</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.750341</td>\n",
       "      <td>0.007381</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.005622</td>\n",
       "      <td>0.085916</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.074977</td>\n",
       "      <td>0.602288</td>\n",
       "      <td>0.635966</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.735646</td>\n",
       "      <td>0.765037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Decision tree_Metadata</td>\n",
       "      <td>0.650198</td>\n",
       "      <td>0.946667</td>\n",
       "      <td>0.767979</td>\n",
       "      <td>0.006530</td>\n",
       "      <td>0.012267</td>\n",
       "      <td>0.006592</td>\n",
       "      <td>0.080806</td>\n",
       "      <td>0.110755</td>\n",
       "      <td>0.081189</td>\n",
       "      <td>0.634361</td>\n",
       "      <td>0.666036</td>\n",
       "      <td>0.924959</td>\n",
       "      <td>0.968375</td>\n",
       "      <td>0.752066</td>\n",
       "      <td>0.783892</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Logistic regression_Metadata</td>\n",
       "      <td>0.632738</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.760836</td>\n",
       "      <td>0.006985</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.005664</td>\n",
       "      <td>0.083576</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.075258</td>\n",
       "      <td>0.616357</td>\n",
       "      <td>0.649119</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.746085</td>\n",
       "      <td>0.775586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>SVM (Gaussian Kernel)_Metadata</td>\n",
       "      <td>0.603056</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.739535</td>\n",
       "      <td>0.007413</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.006231</td>\n",
       "      <td>0.086097</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.078939</td>\n",
       "      <td>0.586181</td>\n",
       "      <td>0.619931</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.724063</td>\n",
       "      <td>0.755008</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Bagging_Metadata</td>\n",
       "      <td>0.616667</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.750030</td>\n",
       "      <td>0.007454</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.006500</td>\n",
       "      <td>0.086335</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.080625</td>\n",
       "      <td>0.599745</td>\n",
       "      <td>0.633588</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.734228</td>\n",
       "      <td>0.765832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Random Forest_Metadata</td>\n",
       "      <td>0.619127</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.750341</td>\n",
       "      <td>0.007381</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.005622</td>\n",
       "      <td>0.085916</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.074977</td>\n",
       "      <td>0.602288</td>\n",
       "      <td>0.635966</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.735646</td>\n",
       "      <td>0.765037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>AdaBoost_Metadata</td>\n",
       "      <td>0.619127</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.750341</td>\n",
       "      <td>0.007381</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.005622</td>\n",
       "      <td>0.085916</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.074977</td>\n",
       "      <td>0.602288</td>\n",
       "      <td>0.635966</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.735646</td>\n",
       "      <td>0.765037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Gradient Boosting Tree_Metadata</td>\n",
       "      <td>0.616667</td>\n",
       "      <td>0.966667</td>\n",
       "      <td>0.750030</td>\n",
       "      <td>0.007454</td>\n",
       "      <td>0.010000</td>\n",
       "      <td>0.006500</td>\n",
       "      <td>0.086335</td>\n",
       "      <td>0.100000</td>\n",
       "      <td>0.080625</td>\n",
       "      <td>0.599745</td>\n",
       "      <td>0.633588</td>\n",
       "      <td>0.947067</td>\n",
       "      <td>0.986267</td>\n",
       "      <td>0.734228</td>\n",
       "      <td>0.765832</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                 clf  mean_precision  mean_recall   mean_f1  \\\n",
       "0          Logistic regression_Audio        0.547980     1.000000  0.707843   \n",
       "1       Gradient Boosting Tree_Audio        0.683571     0.713333  0.688710   \n",
       "2         k-Nearest neighbors_Visual        0.752500     0.826667  0.775140   \n",
       "3               Decision tree_Visual        0.612143     0.636667  0.618660   \n",
       "4         Logistic regression_Visual        0.707817     0.806667  0.743994   \n",
       "5       SVM (Gaussian Kernel)_Visual        0.622421     0.926667  0.738574   \n",
       "6               Random Forest_Visual        0.533333     0.513333  0.512157   \n",
       "7                    AdaBoost_Visual        0.737857     0.770000  0.748838   \n",
       "8      Gradient Boosting Tree_Visual        0.778333     0.810000  0.779149   \n",
       "9                Naive Bayes_Textual        0.743413     0.713333  0.715354   \n",
       "10       k-Nearest neighbors_Textual        0.635934     0.940000  0.750622   \n",
       "11     SVM (Gaussian Kernel)_Textual        0.547980     1.000000  0.707843   \n",
       "12      k-Nearest neighbors_Metadata        0.607601     1.000000  0.752981   \n",
       "13  Nearest mean classifier_Metadata        0.619127     0.966667  0.750341   \n",
       "14            Decision tree_Metadata        0.650198     0.946667  0.767979   \n",
       "15      Logistic regression_Metadata        0.632738     0.966667  0.760836   \n",
       "16    SVM (Gaussian Kernel)_Metadata        0.603056     0.966667  0.739535   \n",
       "17                  Bagging_Metadata        0.616667     0.966667  0.750030   \n",
       "18            Random Forest_Metadata        0.619127     0.966667  0.750341   \n",
       "19                 AdaBoost_Metadata        0.619127     0.966667  0.750341   \n",
       "20   Gradient Boosting Tree_Metadata        0.616667     0.966667  0.750030   \n",
       "\n",
       "    var_precision  var_recall    var_f1  std_precision  std_recall    std_f1  \\\n",
       "0        0.000272    0.000000  0.000199       0.016482    0.000000  0.014120   \n",
       "1        0.019987    0.024044  0.013620       0.141375    0.155063  0.116704   \n",
       "2        0.025471    0.025067  0.015765       0.159595    0.158325  0.125558   \n",
       "3        0.016629    0.040100  0.025758       0.128952    0.200250  0.160494   \n",
       "4        0.014969    0.016178  0.007113       0.122349    0.127192  0.084337   \n",
       "5        0.013050    0.008178  0.007530       0.114238    0.090431  0.086776   \n",
       "6        0.010000    0.037378  0.022652       0.100000    0.193333  0.150506   \n",
       "7        0.020573    0.032100  0.022439       0.143433    0.179165  0.149797   \n",
       "8        0.025211    0.021789  0.011501       0.158780    0.147611  0.107244   \n",
       "9        0.030708    0.029600  0.023383       0.175237    0.172047  0.152915   \n",
       "10       0.013086    0.016400  0.009240       0.114393    0.128062  0.096125   \n",
       "11       0.000272    0.000000  0.000199       0.016482    0.000000  0.014120   \n",
       "12       0.006662    0.000000  0.003338       0.081620    0.000000  0.057775   \n",
       "13       0.007381    0.010000  0.005622       0.085916    0.100000  0.074977   \n",
       "14       0.006530    0.012267  0.006592       0.080806    0.110755  0.081189   \n",
       "15       0.006985    0.010000  0.005664       0.083576    0.100000  0.075258   \n",
       "16       0.007413    0.010000  0.006231       0.086097    0.100000  0.078939   \n",
       "17       0.007454    0.010000  0.006500       0.086335    0.100000  0.080625   \n",
       "18       0.007381    0.010000  0.005622       0.085916    0.100000  0.074977   \n",
       "19       0.007381    0.010000  0.005622       0.085916    0.100000  0.074977   \n",
       "20       0.007454    0.010000  0.006500       0.086335    0.100000  0.080625   \n",
       "\n",
       "    l_95ci_precision  u_95ci_precision  l_95ci_recall  u_95ci_recall  \\\n",
       "0           0.544749          0.551210       1.000000       1.000000   \n",
       "1           0.655862          0.711281       0.682941       0.743726   \n",
       "2           0.721219          0.783781       0.795635       0.857698   \n",
       "3           0.586868          0.637417       0.597418       0.675916   \n",
       "4           0.683837          0.731798       0.781737       0.831596   \n",
       "5           0.600030          0.644811       0.908942       0.944391   \n",
       "6           0.513733          0.552933       0.475440       0.551227   \n",
       "7           0.709744          0.765970       0.734884       0.805116   \n",
       "8           0.747212          0.809454       0.781068       0.838932   \n",
       "9           0.709066          0.777759       0.679612       0.747054   \n",
       "10          0.613513          0.658355       0.914900       0.965100   \n",
       "11          0.544749          0.551210       1.000000       1.000000   \n",
       "12          0.591604          0.623599       1.000000       1.000000   \n",
       "13          0.602288          0.635966       0.947067       0.986267   \n",
       "14          0.634361          0.666036       0.924959       0.968375   \n",
       "15          0.616357          0.649119       0.947067       0.986267   \n",
       "16          0.586181          0.619931       0.947067       0.986267   \n",
       "17          0.599745          0.633588       0.947067       0.986267   \n",
       "18          0.602288          0.635966       0.947067       0.986267   \n",
       "19          0.602288          0.635966       0.947067       0.986267   \n",
       "20          0.599745          0.633588       0.947067       0.986267   \n",
       "\n",
       "    l_95ci_f1  u_95ci_f1  \n",
       "0    0.705076   0.710611  \n",
       "1    0.665836   0.711584  \n",
       "2    0.750530   0.799749  \n",
       "3    0.587203   0.650117  \n",
       "4    0.727464   0.760524  \n",
       "5    0.721566   0.755582  \n",
       "6    0.482658   0.541656  \n",
       "7    0.719478   0.778198  \n",
       "8    0.758129   0.800169  \n",
       "9    0.685382   0.745325  \n",
       "10   0.731781   0.769462  \n",
       "11   0.705076   0.710611  \n",
       "12   0.741657   0.764305  \n",
       "13   0.735646   0.765037  \n",
       "14   0.752066   0.783892  \n",
       "15   0.746085   0.775586  \n",
       "16   0.724063   0.755008  \n",
       "17   0.734228   0.765832  \n",
       "18   0.735646   0.765037  \n",
       "19   0.735646   0.765037  \n",
       "20   0.734228   0.765832  "
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from itertools import chain\n",
    "ow.df_score_stats\n",
    "#ow.df_score_stats.loc[:,['clf', 'mean_precision', 'mean_recall', 'mean_f1']]\n",
    "#a = ow.df_score_stats.loc[:1,['clf', 'mean_recall']].values #audio\n",
    "#a = ow.df_score_stats.loc[2:8,['clf', 'mean_recall']] #visual\n",
    "#a = ow.df_score_stats.loc[9:11,['mean_recall']].values #text\n",
    "# a = ow.df_score_stats.loc[12:,['mean_recall']].values #meta\n",
    "#a = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['mean_recall']].values]))\n",
    "#a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>TARGET</th>\n",
       "      <th>Logistic regression_Audio</th>\n",
       "      <th>Gradient Boosting Tree_Audio</th>\n",
       "      <th>k-Nearest neighbors_Visual</th>\n",
       "      <th>Decision tree_Visual</th>\n",
       "      <th>Logistic regression_Visual</th>\n",
       "      <th>SVM (Gaussian Kernel)_Visual</th>\n",
       "      <th>Random Forest_Visual</th>\n",
       "      <th>AdaBoost_Visual</th>\n",
       "      <th>Gradient Boosting Tree_Visual</th>\n",
       "      <th>...</th>\n",
       "      <th>SVM (Gaussian Kernel)_Textual</th>\n",
       "      <th>k-Nearest neighbors_Metadata</th>\n",
       "      <th>Nearest mean classifier_Metadata</th>\n",
       "      <th>Decision tree_Metadata</th>\n",
       "      <th>Logistic regression_Metadata</th>\n",
       "      <th>SVM (Gaussian Kernel)_Metadata</th>\n",
       "      <th>Bagging_Metadata</th>\n",
       "      <th>Random Forest_Metadata</th>\n",
       "      <th>AdaBoost_Metadata</th>\n",
       "      <th>Gradient Boosting Tree_Metadata</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>85</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>86</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>87</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>88</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>89</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>90</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>91</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>94</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>95 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    TARGET  Logistic regression_Audio  Gradient Boosting Tree_Audio  \\\n",
       "0        1                          1                             0   \n",
       "1        0                          1                             0   \n",
       "2        1                          1                             1   \n",
       "3        1                          1                             1   \n",
       "4        1                          1                             1   \n",
       "5        1                          1                             0   \n",
       "6        1                          1                             1   \n",
       "7        1                          1                             1   \n",
       "8        0                          1                             0   \n",
       "9        0                          1                             1   \n",
       "10       1                          1                             1   \n",
       "11       1                          1                             1   \n",
       "12       1                          1                             1   \n",
       "13       0                          1                             1   \n",
       "14       0                          1                             1   \n",
       "15       1                          1                             1   \n",
       "16       0                          1                             0   \n",
       "17       1                          1                             1   \n",
       "18       0                          1                             0   \n",
       "19       1                          1                             1   \n",
       "20       1                          1                             1   \n",
       "21       0                          1                             0   \n",
       "22       0                          1                             0   \n",
       "23       0                          1                             0   \n",
       "24       0                          1                             0   \n",
       "25       1                          1                             1   \n",
       "26       1                          1                             1   \n",
       "27       1                          1                             1   \n",
       "28       0                          1                             0   \n",
       "29       1                          1                             1   \n",
       "..     ...                        ...                           ...   \n",
       "65       1                          1                             1   \n",
       "66       1                          1                             1   \n",
       "67       1                          1                             1   \n",
       "68       0                          1                             0   \n",
       "69       0                          1                             0   \n",
       "70       0                          1                             0   \n",
       "71       0                          1                             0   \n",
       "72       0                          1                             0   \n",
       "73       1                          1                             1   \n",
       "74       1                          1                             1   \n",
       "75       0                          1                             0   \n",
       "76       1                          1                             1   \n",
       "77       0                          1                             0   \n",
       "78       0                          1                             0   \n",
       "79       1                          1                             1   \n",
       "80       1                          1                             1   \n",
       "81       1                          1                             1   \n",
       "82       1                          1                             1   \n",
       "83       0                          1                             0   \n",
       "84       1                          1                             1   \n",
       "85       1                          1                             1   \n",
       "86       0                          1                             0   \n",
       "87       1                          1                             1   \n",
       "88       1                          1                             1   \n",
       "89       1                          1                             1   \n",
       "90       0                          1                             0   \n",
       "91       0                          1                             0   \n",
       "92       0                          1                             0   \n",
       "93       0                          1                             0   \n",
       "94       1                          1                             1   \n",
       "\n",
       "    k-Nearest neighbors_Visual  Decision tree_Visual  \\\n",
       "0                            1                     0   \n",
       "1                            0                     0   \n",
       "2                            1                     1   \n",
       "3                            1                     0   \n",
       "4                            1                     0   \n",
       "5                            1                     1   \n",
       "6                            1                     1   \n",
       "7                            1                     1   \n",
       "8                            0                     0   \n",
       "9                            0                     1   \n",
       "10                           0                     1   \n",
       "11                           1                     1   \n",
       "12                           1                     1   \n",
       "13                           1                     0   \n",
       "14                           0                     1   \n",
       "15                           1                     1   \n",
       "16                           0                     0   \n",
       "17                           0                     1   \n",
       "18                           1                     0   \n",
       "19                           1                     1   \n",
       "20                           1                     1   \n",
       "21                           1                     1   \n",
       "22                           0                     0   \n",
       "23                           0                     0   \n",
       "24                           0                     0   \n",
       "25                           1                     1   \n",
       "26                           1                     1   \n",
       "27                           0                     1   \n",
       "28                           0                     0   \n",
       "29                           1                     1   \n",
       "..                         ...                   ...   \n",
       "65                           1                     1   \n",
       "66                           1                     1   \n",
       "67                           1                     1   \n",
       "68                           0                     0   \n",
       "69                           0                     0   \n",
       "70                           0                     0   \n",
       "71                           0                     0   \n",
       "72                           0                     0   \n",
       "73                           1                     1   \n",
       "74                           1                     1   \n",
       "75                           0                     0   \n",
       "76                           1                     1   \n",
       "77                           1                     1   \n",
       "78                           0                     0   \n",
       "79                           1                     1   \n",
       "80                           1                     1   \n",
       "81                           1                     1   \n",
       "82                           0                     1   \n",
       "83                           1                     0   \n",
       "84                           1                     1   \n",
       "85                           0                     1   \n",
       "86                           1                     0   \n",
       "87                           0                     1   \n",
       "88                           1                     1   \n",
       "89                           1                     1   \n",
       "90                           0                     0   \n",
       "91                           0                     0   \n",
       "92                           0                     0   \n",
       "93                           0                     0   \n",
       "94                           1                     1   \n",
       "\n",
       "    Logistic regression_Visual  SVM (Gaussian Kernel)_Visual  \\\n",
       "0                            1                             1   \n",
       "1                            0                             1   \n",
       "2                            1                             1   \n",
       "3                            1                             0   \n",
       "4                            0                             1   \n",
       "5                            1                             1   \n",
       "6                            1                             1   \n",
       "7                            1                             1   \n",
       "8                            1                             0   \n",
       "9                            1                             0   \n",
       "10                           0                             1   \n",
       "11                           1                             1   \n",
       "12                           1                             1   \n",
       "13                           1                             0   \n",
       "14                           1                             0   \n",
       "15                           1                             1   \n",
       "16                           0                             0   \n",
       "17                           1                             1   \n",
       "18                           0                             0   \n",
       "19                           1                             1   \n",
       "20                           0                             1   \n",
       "21                           1                             1   \n",
       "22                           0                             0   \n",
       "23                           0                             0   \n",
       "24                           0                             0   \n",
       "25                           1                             1   \n",
       "26                           1                             1   \n",
       "27                           1                             1   \n",
       "28                           0                             0   \n",
       "29                           1                             1   \n",
       "..                         ...                           ...   \n",
       "65                           1                             1   \n",
       "66                           1                             1   \n",
       "67                           1                             1   \n",
       "68                           0                             0   \n",
       "69                           0                             0   \n",
       "70                           1                             1   \n",
       "71                           0                             0   \n",
       "72                           0                             0   \n",
       "73                           1                             1   \n",
       "74                           1                             1   \n",
       "75                           0                             0   \n",
       "76                           1                             1   \n",
       "77                           1                             1   \n",
       "78                           0                             0   \n",
       "79                           1                             1   \n",
       "80                           1                             1   \n",
       "81                           1                             1   \n",
       "82                           1                             1   \n",
       "83                           1                             0   \n",
       "84                           0                             1   \n",
       "85                           0                             1   \n",
       "86                           0                             0   \n",
       "87                           0                             1   \n",
       "88                           1                             1   \n",
       "89                           1                             1   \n",
       "90                           0                             0   \n",
       "91                           1                             0   \n",
       "92                           0                             0   \n",
       "93                           0                             0   \n",
       "94                           0                             1   \n",
       "\n",
       "    Random Forest_Visual  AdaBoost_Visual  Gradient Boosting Tree_Visual  \\\n",
       "0                      1                1                              1   \n",
       "1                      1                0                              0   \n",
       "2                      1                1                              1   \n",
       "3                      0                1                              1   \n",
       "4                      0                1                              1   \n",
       "5                      1                1                              1   \n",
       "6                      1                1                              0   \n",
       "7                      1                1                              1   \n",
       "8                      0                0                              0   \n",
       "9                      0                1                              0   \n",
       "10                     1                1                              1   \n",
       "11                     1                1                              1   \n",
       "12                     1                1                              1   \n",
       "13                     1                1                              0   \n",
       "14                     0                0                              0   \n",
       "15                     1                1                              1   \n",
       "16                     0                0                              0   \n",
       "17                     1                1                              1   \n",
       "18                     0                0                              0   \n",
       "19                     1                1                              1   \n",
       "20                     1                1                              1   \n",
       "21                     1                1                              1   \n",
       "22                     0                0                              0   \n",
       "23                     0                0                              0   \n",
       "24                     0                0                              0   \n",
       "25                     0                1                              1   \n",
       "26                     1                1                              1   \n",
       "27                     1                1                              1   \n",
       "28                     0                0                              0   \n",
       "29                     1                1                              1   \n",
       "..                   ...              ...                            ...   \n",
       "65                     1                1                              1   \n",
       "66                     0                1                              1   \n",
       "67                     1                1                              1   \n",
       "68                     0                0                              0   \n",
       "69                     0                0                              0   \n",
       "70                     0                0                              0   \n",
       "71                     0                0                              0   \n",
       "72                     0                0                              0   \n",
       "73                     1                1                              1   \n",
       "74                     1                1                              1   \n",
       "75                     0                0                              0   \n",
       "76                     1                1                              1   \n",
       "77                     1                1                              1   \n",
       "78                     0                0                              0   \n",
       "79                     1                1                              1   \n",
       "80                     1                1                              1   \n",
       "81                     1                1                              1   \n",
       "82                     1                1                              1   \n",
       "83                     0                0                              0   \n",
       "84                     1                1                              1   \n",
       "85                     1                1                              1   \n",
       "86                     0                0                              0   \n",
       "87                     1                1                              1   \n",
       "88                     1                1                              1   \n",
       "89                     0                1                              1   \n",
       "90                     0                0                              0   \n",
       "91                     0                0                              0   \n",
       "92                     0                0                              0   \n",
       "93                     0                0                              0   \n",
       "94                     1                1                              1   \n",
       "\n",
       "                 ...                 SVM (Gaussian Kernel)_Textual  \\\n",
       "0                ...                                             1   \n",
       "1                ...                                             1   \n",
       "2                ...                                             1   \n",
       "3                ...                                             1   \n",
       "4                ...                                             1   \n",
       "5                ...                                             1   \n",
       "6                ...                                             1   \n",
       "7                ...                                             1   \n",
       "8                ...                                             1   \n",
       "9                ...                                             1   \n",
       "10               ...                                             1   \n",
       "11               ...                                             1   \n",
       "12               ...                                             1   \n",
       "13               ...                                             1   \n",
       "14               ...                                             1   \n",
       "15               ...                                             1   \n",
       "16               ...                                             1   \n",
       "17               ...                                             1   \n",
       "18               ...                                             1   \n",
       "19               ...                                             1   \n",
       "20               ...                                             1   \n",
       "21               ...                                             1   \n",
       "22               ...                                             1   \n",
       "23               ...                                             1   \n",
       "24               ...                                             1   \n",
       "25               ...                                             1   \n",
       "26               ...                                             1   \n",
       "27               ...                                             1   \n",
       "28               ...                                             1   \n",
       "29               ...                                             1   \n",
       "..               ...                                           ...   \n",
       "65               ...                                             1   \n",
       "66               ...                                             1   \n",
       "67               ...                                             1   \n",
       "68               ...                                             1   \n",
       "69               ...                                             1   \n",
       "70               ...                                             1   \n",
       "71               ...                                             1   \n",
       "72               ...                                             1   \n",
       "73               ...                                             1   \n",
       "74               ...                                             1   \n",
       "75               ...                                             1   \n",
       "76               ...                                             1   \n",
       "77               ...                                             1   \n",
       "78               ...                                             1   \n",
       "79               ...                                             1   \n",
       "80               ...                                             1   \n",
       "81               ...                                             1   \n",
       "82               ...                                             1   \n",
       "83               ...                                             1   \n",
       "84               ...                                             1   \n",
       "85               ...                                             1   \n",
       "86               ...                                             1   \n",
       "87               ...                                             1   \n",
       "88               ...                                             1   \n",
       "89               ...                                             1   \n",
       "90               ...                                             1   \n",
       "91               ...                                             1   \n",
       "92               ...                                             1   \n",
       "93               ...                                             1   \n",
       "94               ...                                             1   \n",
       "\n",
       "    k-Nearest neighbors_Metadata  Nearest mean classifier_Metadata  \\\n",
       "0                              1                                 0   \n",
       "1                              1                                 1   \n",
       "2                              1                                 1   \n",
       "3                              1                                 1   \n",
       "4                              1                                 0   \n",
       "5                              1                                 1   \n",
       "6                              1                                 1   \n",
       "7                              1                                 1   \n",
       "8                              1                                 0   \n",
       "9                              1                                 1   \n",
       "10                             1                                 1   \n",
       "11                             1                                 1   \n",
       "12                             1                                 1   \n",
       "13                             1                                 1   \n",
       "14                             1                                 0   \n",
       "15                             1                                 1   \n",
       "16                             1                                 1   \n",
       "17                             1                                 1   \n",
       "18                             1                                 1   \n",
       "19                             1                                 1   \n",
       "20                             1                                 1   \n",
       "21                             0                                 0   \n",
       "22                             1                                 1   \n",
       "23                             1                                 0   \n",
       "24                             1                                 1   \n",
       "25                             1                                 1   \n",
       "26                             1                                 1   \n",
       "27                             1                                 1   \n",
       "28                             1                                 1   \n",
       "29                             1                                 1   \n",
       "..                           ...                               ...   \n",
       "65                             1                                 1   \n",
       "66                             1                                 1   \n",
       "67                             1                                 1   \n",
       "68                             1                                 1   \n",
       "69                             1                                 1   \n",
       "70                             1                                 0   \n",
       "71                             1                                 1   \n",
       "72                             1                                 1   \n",
       "73                             1                                 1   \n",
       "74                             1                                 1   \n",
       "75                             1                                 1   \n",
       "76                             1                                 1   \n",
       "77                             1                                 1   \n",
       "78                             1                                 1   \n",
       "79                             1                                 1   \n",
       "80                             1                                 1   \n",
       "81                             1                                 1   \n",
       "82                             1                                 1   \n",
       "83                             1                                 1   \n",
       "84                             1                                 1   \n",
       "85                             1                                 1   \n",
       "86                             1                                 1   \n",
       "87                             1                                 1   \n",
       "88                             1                                 1   \n",
       "89                             1                                 1   \n",
       "90                             1                                 1   \n",
       "91                             1                                 1   \n",
       "92                             1                                 1   \n",
       "93                             1                                 1   \n",
       "94                             1                                 1   \n",
       "\n",
       "    Decision tree_Metadata  Logistic regression_Metadata  \\\n",
       "0                        0                             0   \n",
       "1                        1                             1   \n",
       "2                        1                             1   \n",
       "3                        1                             1   \n",
       "4                        0                             0   \n",
       "5                        1                             1   \n",
       "6                        1                             1   \n",
       "7                        1                             1   \n",
       "8                        0                             0   \n",
       "9                        1                             1   \n",
       "10                       1                             1   \n",
       "11                       1                             1   \n",
       "12                       1                             1   \n",
       "13                       1                             1   \n",
       "14                       0                             0   \n",
       "15                       1                             1   \n",
       "16                       1                             1   \n",
       "17                       1                             1   \n",
       "18                       1                             1   \n",
       "19                       1                             1   \n",
       "20                       1                             1   \n",
       "21                       0                             0   \n",
       "22                       0                             0   \n",
       "23                       0                             1   \n",
       "24                       0                             1   \n",
       "25                       1                             1   \n",
       "26                       0                             1   \n",
       "27                       1                             1   \n",
       "28                       1                             1   \n",
       "29                       1                             1   \n",
       "..                     ...                           ...   \n",
       "65                       1                             1   \n",
       "66                       1                             1   \n",
       "67                       1                             1   \n",
       "68                       1                             0   \n",
       "69                       1                             0   \n",
       "70                       0                             0   \n",
       "71                       1                             1   \n",
       "72                       1                             1   \n",
       "73                       1                             1   \n",
       "74                       1                             1   \n",
       "75                       1                             1   \n",
       "76                       1                             1   \n",
       "77                       1                             1   \n",
       "78                       1                             1   \n",
       "79                       1                             1   \n",
       "80                       1                             1   \n",
       "81                       1                             1   \n",
       "82                       1                             1   \n",
       "83                       1                             1   \n",
       "84                       1                             1   \n",
       "85                       1                             1   \n",
       "86                       1                             1   \n",
       "87                       1                             1   \n",
       "88                       1                             1   \n",
       "89                       1                             1   \n",
       "90                       1                             1   \n",
       "91                       1                             1   \n",
       "92                       1                             1   \n",
       "93                       0                             0   \n",
       "94                       1                             1   \n",
       "\n",
       "    SVM (Gaussian Kernel)_Metadata  Bagging_Metadata  Random Forest_Metadata  \\\n",
       "0                                0                 0                       0   \n",
       "1                                1                 1                       1   \n",
       "2                                1                 1                       1   \n",
       "3                                1                 1                       1   \n",
       "4                                0                 0                       0   \n",
       "5                                1                 1                       1   \n",
       "6                                1                 1                       1   \n",
       "7                                1                 1                       1   \n",
       "8                                1                 1                       0   \n",
       "9                                1                 1                       1   \n",
       "10                               1                 1                       1   \n",
       "11                               1                 1                       1   \n",
       "12                               1                 1                       1   \n",
       "13                               1                 1                       1   \n",
       "14                               0                 0                       0   \n",
       "15                               1                 1                       1   \n",
       "16                               1                 1                       1   \n",
       "17                               1                 1                       1   \n",
       "18                               1                 1                       1   \n",
       "19                               1                 1                       1   \n",
       "20                               1                 1                       1   \n",
       "21                               0                 0                       0   \n",
       "22                               1                 0                       1   \n",
       "23                               1                 1                       1   \n",
       "24                               1                 1                       1   \n",
       "25                               1                 1                       1   \n",
       "26                               1                 1                       1   \n",
       "27                               1                 1                       1   \n",
       "28                               1                 1                       1   \n",
       "29                               1                 1                       1   \n",
       "..                             ...               ...                     ...   \n",
       "65                               1                 1                       1   \n",
       "66                               1                 1                       1   \n",
       "67                               1                 1                       1   \n",
       "68                               1                 1                       1   \n",
       "69                               1                 1                       1   \n",
       "70                               0                 0                       0   \n",
       "71                               1                 1                       1   \n",
       "72                               1                 1                       1   \n",
       "73                               1                 1                       1   \n",
       "74                               1                 1                       1   \n",
       "75                               1                 1                       1   \n",
       "76                               1                 1                       1   \n",
       "77                               1                 1                       1   \n",
       "78                               1                 0                       1   \n",
       "79                               1                 1                       1   \n",
       "80                               1                 1                       1   \n",
       "81                               1                 1                       1   \n",
       "82                               1                 1                       1   \n",
       "83                               1                 1                       1   \n",
       "84                               1                 1                       1   \n",
       "85                               1                 1                       1   \n",
       "86                               1                 1                       1   \n",
       "87                               1                 1                       1   \n",
       "88                               1                 1                       1   \n",
       "89                               1                 1                       1   \n",
       "90                               1                 1                       1   \n",
       "91                               1                 1                       1   \n",
       "92                               1                 1                       1   \n",
       "93                               1                 0                       1   \n",
       "94                               1                 1                       1   \n",
       "\n",
       "    AdaBoost_Metadata  Gradient Boosting Tree_Metadata  \n",
       "0                   0                                0  \n",
       "1                   1                                1  \n",
       "2                   1                                1  \n",
       "3                   1                                1  \n",
       "4                   0                                0  \n",
       "5                   1                                1  \n",
       "6                   1                                1  \n",
       "7                   1                                1  \n",
       "8                   0                                1  \n",
       "9                   1                                1  \n",
       "10                  1                                1  \n",
       "11                  1                                1  \n",
       "12                  1                                1  \n",
       "13                  1                                1  \n",
       "14                  0                                0  \n",
       "15                  1                                1  \n",
       "16                  0                                1  \n",
       "17                  1                                1  \n",
       "18                  1                                1  \n",
       "19                  1                                1  \n",
       "20                  1                                1  \n",
       "21                  0                                0  \n",
       "22                  1                                0  \n",
       "23                  1                                0  \n",
       "24                  1                                1  \n",
       "25                  1                                1  \n",
       "26                  1                                1  \n",
       "27                  1                                1  \n",
       "28                  1                                1  \n",
       "29                  1                                1  \n",
       "..                ...                              ...  \n",
       "65                  1                                1  \n",
       "66                  1                                1  \n",
       "67                  1                                1  \n",
       "68                  1                                1  \n",
       "69                  1                                1  \n",
       "70                  0                                0  \n",
       "71                  1                                1  \n",
       "72                  1                                1  \n",
       "73                  1                                1  \n",
       "74                  1                                1  \n",
       "75                  1                                1  \n",
       "76                  1                                1  \n",
       "77                  1                                1  \n",
       "78                  1                                1  \n",
       "79                  1                                1  \n",
       "80                  1                                1  \n",
       "81                  1                                1  \n",
       "82                  1                                1  \n",
       "83                  1                                1  \n",
       "84                  1                                1  \n",
       "85                  1                                1  \n",
       "86                  1                                1  \n",
       "87                  1                                1  \n",
       "88                  1                                1  \n",
       "89                  1                                1  \n",
       "90                  1                                1  \n",
       "91                  1                                1  \n",
       "92                  1                                1  \n",
       "93                  1                                0  \n",
       "94                  1                                1  \n",
       "\n",
       "[95 rows x 22 columns]"
      ]
     },
     "execution_count": 109,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ow.df_res_all[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Classifier_Modality</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "      <th>paper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>k-Nearest neighbors</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.608</td>\n",
       "      <td>0.607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Nearest mean classifier</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.619</td>\n",
       "      <td>0.603</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Decision tree</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.650</td>\n",
       "      <td>0.538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.633</td>\n",
       "      <td>0.548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SVM (Gaussian Kernel)</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.603</td>\n",
       "      <td>0.501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Bagging</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.617</td>\n",
       "      <td>0.604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.619</td>\n",
       "      <td>0.559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>AdaBoost</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.619</td>\n",
       "      <td>0.511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.617</td>\n",
       "      <td>0.544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>k-Nearest neighbors</td>\n",
       "      <td>Recall</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Nearest mean classifier</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Decision tree</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.947</td>\n",
       "      <td>0.591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>SVM (Gaussian Kernel)</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Bagging</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>AdaBoost</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.967</td>\n",
       "      <td>0.596</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>k-Nearest neighbors</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.753</td>\n",
       "      <td>0.630</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>Nearest mean classifier</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.591</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Decision tree</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.768</td>\n",
       "      <td>0.563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.761</td>\n",
       "      <td>0.578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>SVM (Gaussian Kernel)</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.740</td>\n",
       "      <td>0.574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>Bagging</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.576</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>AdaBoost</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.750</td>\n",
       "      <td>0.569</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        Classifier_Modality   variable  value  paper\n",
       "0       k-Nearest neighbors  Precision  0.608  0.607\n",
       "1   Nearest mean classifier  Precision  0.619  0.603\n",
       "2             Decision tree  Precision  0.650  0.538\n",
       "3       Logistic regression  Precision  0.633  0.548\n",
       "4     SVM (Gaussian Kernel)  Precision  0.603  0.501\n",
       "5                   Bagging  Precision  0.617  0.604\n",
       "6             Random Forest  Precision  0.619  0.559\n",
       "7                  AdaBoost  Precision  0.619  0.511\n",
       "8    Gradient Boosting Tree  Precision  0.617  0.544\n",
       "9       k-Nearest neighbors     Recall  1.000  0.654\n",
       "10  Nearest mean classifier     Recall  0.967  0.579\n",
       "11            Decision tree     Recall  0.947  0.591\n",
       "12      Logistic regression     Recall  0.967  0.609\n",
       "13    SVM (Gaussian Kernel)     Recall  0.967  0.672\n",
       "14                  Bagging     Recall  0.967  0.662\n",
       "15            Random Forest     Recall  0.967  0.593\n",
       "16                 AdaBoost     Recall  0.967  0.563\n",
       "17   Gradient Boosting Tree     Recall  0.967  0.596\n",
       "18      k-Nearest neighbors         F1  0.753  0.630\n",
       "19  Nearest mean classifier         F1  0.750  0.591\n",
       "20            Decision tree         F1  0.768  0.563\n",
       "21      Logistic regression         F1  0.761  0.578\n",
       "22    SVM (Gaussian Kernel)         F1  0.740  0.574\n",
       "23                  Bagging         F1  0.750  0.631\n",
       "24            Random Forest         F1  0.750  0.576\n",
       "25                 AdaBoost         F1  0.750  0.536\n",
       "26   Gradient Boosting Tree         F1  0.750  0.569"
      ]
     },
     "execution_count": 113,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_meta = pd.DataFrame()\n",
    "df_visual = pd.DataFrame()\n",
    "df_text = pd.DataFrame()\n",
    "df_audio = pd.DataFrame()\n",
    "\n",
    "clfs = ClassifierFactory.get_metadata_classifiers()\n",
    "classifier = [clf.clf_name for clf in clfs]\n",
    "#classifier = [clf.clf_name+\" \"+clf.modality for clf in clfs]\n",
    "df_meta['Classifier_Modality'] = classifier\n",
    "\n",
    "clfs = ClassifierFactory.get_visual_classifiers()\n",
    "classifier = [clf.clf_name for clf in clfs]\n",
    "#classifier = [clf.clf_name+\" \"+clf.modality for clf in clfs]\n",
    "df_visual['Classifier_Modality'] = classifier\n",
    "\n",
    "clfs = ClassifierFactory.get_text_classifiers()\n",
    "classifier = [clf.clf_name for clf in clfs]\n",
    "#classifier = [clf.clf_name+\" \"+clf.modality for clf in clfs]\n",
    "df_text['Classifier_Modality'] = classifier\n",
    "\n",
    "clfs = ClassifierFactory.get_audio_classifiers()\n",
    "classifier = [clf.clf_name for clf in clfs]\n",
    "#classifier = [clf.clf_name+\" \"+clf.modality for clf in clfs]\n",
    "df_audio['Classifier_Modality'] = classifier\n",
    "\n",
    "df_meta['Precision'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[12:,['mean_precision']].values]))\n",
    "df_text['Precision'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[9:11,['mean_precision']].values]))\n",
    "df_visual['Precision'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['mean_precision']].values]))\n",
    "df_audio['Precision'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[:1,['mean_precision']].values]))\n",
    "\n",
    "df_meta['Recall'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[12:,['mean_recall']].values]))\n",
    "df_text['Recall'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[9:11,['mean_recall']].values]))\n",
    "df_visual['Recall'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['mean_recall']].values]))\n",
    "df_audio['Recall'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[:1,['mean_recall']].values]))\n",
    "\n",
    "df_meta['F1'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[12:,['mean_f1']].values]))\n",
    "df_text['F1'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[9:11,['mean_f1']].values]))\n",
    "df_visual['F1'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['mean_f1']].values]))\n",
    "df_audio['F1'] = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[:1,['mean_f1']].values]))\n",
    "\n",
    "df_meta = df_meta.melt(id_vars=['Classifier_Modality'])\n",
    "df_visual = df_visual.melt(id_vars=['Classifier_Modality'])\n",
    "df_text = df_text.melt(id_vars=['Classifier_Modality'])\n",
    "df_audio = df_audio.melt(id_vars=['Classifier_Modality'])\n",
    "\n",
    "\n",
    "df_meta['paper'] = [0.607,0.603,0.538,0.548,0.501,0.604,0.559,0.511,0.544,\n",
    "                    0.654,0.579,0.591,0.609,0.672,0.662,0.593,0.563,0.596,\n",
    "                    0.630,0.591,0.563,0.578,0.574,0.631,0.576,0.536,0.569]\n",
    "df_visual['paper'] = [0.582,0.521,0.616,0.511,0.614,0.601,0.561,\n",
    "                      0.636,0.550,0.600,0.670,0.664,0.717,0.616,\n",
    "                      0.608,0.535,0.608,0.580,0.638,0.654,0.587\n",
    "                      ]\n",
    "df_text['paper'] = [0.545,0.549,0.547,\n",
    "                    0.987,0.844,1.000,\n",
    "                    0.702,0.666,0.707]\n",
    "df_audio['paper'] = [0.507,0.560,\n",
    "                     0.597,0.617,\n",
    "                     0.546,0.587]\n",
    "df_meta"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.016, 0.141]\n",
      "[0.0, 0.155]\n",
      "[0.014, 0.117]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Classifier_Modality</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "      <th>paper</th>\n",
       "      <th>std</th>\n",
       "      <th>sig</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.548</td>\n",
       "      <td>0.507</td>\n",
       "      <td>0.016</td>\n",
       "      <td>7.687500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.684</td>\n",
       "      <td>0.560</td>\n",
       "      <td>0.141</td>\n",
       "      <td>2.638298</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>Recall</td>\n",
       "      <td>1.000</td>\n",
       "      <td>0.597</td>\n",
       "      <td>0.000</td>\n",
       "      <td>inf</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.713</td>\n",
       "      <td>0.617</td>\n",
       "      <td>0.155</td>\n",
       "      <td>1.858065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.708</td>\n",
       "      <td>0.546</td>\n",
       "      <td>0.014</td>\n",
       "      <td>34.714286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.689</td>\n",
       "      <td>0.587</td>\n",
       "      <td>0.117</td>\n",
       "      <td>2.615385</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Classifier_Modality   variable  value  paper    std        sig\n",
       "0     Logistic regression  Precision  0.548  0.507  0.016   7.687500\n",
       "1  Gradient Boosting Tree  Precision  0.684  0.560  0.141   2.638298\n",
       "2     Logistic regression     Recall  1.000  0.597  0.000        inf\n",
       "3  Gradient Boosting Tree     Recall  0.713  0.617  0.155   1.858065\n",
       "4     Logistic regression         F1  0.708  0.546  0.014  34.714286\n",
       "5  Gradient Boosting Tree         F1  0.689  0.587  0.117   2.615385"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "s_p = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[:1,['std_precision']].values]))\n",
    "s_r = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[:1,['std_recall']].values]))\n",
    "s_f = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[:1,['std_f1']].values]))\n",
    "print(s_p)\n",
    "print(s_r)\n",
    "print(s_f)\n",
    "df_std = pd.DataFrame(s_p)\n",
    "df_std.columns=['std']\n",
    "df_std = df_std.append(pd.DataFrame(s_r, columns=['std']), ignore_index=True)\n",
    "df_std = df_std.append(pd.DataFrame(s_f, columns=['std']), ignore_index=True)\n",
    "df_std\n",
    "df_audio['std'] = df_std.loc[:,['std']]\n",
    "# (µ_self - µ_paper) / (std/3)\n",
    "df_audio['sig'] = [(row.value-row.paper)*3 for i,row in df_audio.iterrows()]\n",
    "df_audio['sig'] = df_audio['sig']/df_audio['std']\n",
    "df_audio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.16, 0.129, 0.122, 0.114, 0.1, 0.143, 0.159]\n",
      "[0.158, 0.2, 0.127, 0.09, 0.193, 0.179, 0.148]\n",
      "[0.126, 0.16, 0.084, 0.087, 0.151, 0.15, 0.107]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Classifier_Modality</th>\n",
       "      <th>variable</th>\n",
       "      <th>value</th>\n",
       "      <th>paper</th>\n",
       "      <th>std</th>\n",
       "      <th>sig</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>k-Nearest neighbors</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.752</td>\n",
       "      <td>0.582</td>\n",
       "      <td>0.160</td>\n",
       "      <td>3.187500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Decision tree</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.612</td>\n",
       "      <td>0.521</td>\n",
       "      <td>0.129</td>\n",
       "      <td>2.116279</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.708</td>\n",
       "      <td>0.616</td>\n",
       "      <td>0.122</td>\n",
       "      <td>2.262295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVM (Gaussian Kernel)</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.622</td>\n",
       "      <td>0.511</td>\n",
       "      <td>0.114</td>\n",
       "      <td>2.921053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.533</td>\n",
       "      <td>0.614</td>\n",
       "      <td>0.100</td>\n",
       "      <td>-2.430000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>AdaBoost</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.738</td>\n",
       "      <td>0.601</td>\n",
       "      <td>0.143</td>\n",
       "      <td>2.874126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>Precision</td>\n",
       "      <td>0.778</td>\n",
       "      <td>0.561</td>\n",
       "      <td>0.159</td>\n",
       "      <td>4.094340</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>k-Nearest neighbors</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.827</td>\n",
       "      <td>0.636</td>\n",
       "      <td>0.158</td>\n",
       "      <td>3.626582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Decision tree</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.637</td>\n",
       "      <td>0.550</td>\n",
       "      <td>0.200</td>\n",
       "      <td>1.305000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.807</td>\n",
       "      <td>0.600</td>\n",
       "      <td>0.127</td>\n",
       "      <td>4.889764</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>SVM (Gaussian Kernel)</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.927</td>\n",
       "      <td>0.670</td>\n",
       "      <td>0.090</td>\n",
       "      <td>8.566667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.513</td>\n",
       "      <td>0.664</td>\n",
       "      <td>0.193</td>\n",
       "      <td>-2.347150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>AdaBoost</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.770</td>\n",
       "      <td>0.717</td>\n",
       "      <td>0.179</td>\n",
       "      <td>0.888268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>Recall</td>\n",
       "      <td>0.810</td>\n",
       "      <td>0.616</td>\n",
       "      <td>0.148</td>\n",
       "      <td>3.932432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>k-Nearest neighbors</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.775</td>\n",
       "      <td>0.608</td>\n",
       "      <td>0.126</td>\n",
       "      <td>3.976190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>Decision tree</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.619</td>\n",
       "      <td>0.535</td>\n",
       "      <td>0.160</td>\n",
       "      <td>1.575000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>Logistic regression</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.744</td>\n",
       "      <td>0.608</td>\n",
       "      <td>0.084</td>\n",
       "      <td>4.857143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>SVM (Gaussian Kernel)</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.739</td>\n",
       "      <td>0.580</td>\n",
       "      <td>0.087</td>\n",
       "      <td>5.482759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Random Forest</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.512</td>\n",
       "      <td>0.638</td>\n",
       "      <td>0.151</td>\n",
       "      <td>-2.503311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>AdaBoost</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.749</td>\n",
       "      <td>0.654</td>\n",
       "      <td>0.150</td>\n",
       "      <td>1.900000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>Gradient Boosting Tree</td>\n",
       "      <td>F1</td>\n",
       "      <td>0.779</td>\n",
       "      <td>0.587</td>\n",
       "      <td>0.107</td>\n",
       "      <td>5.383178</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Classifier_Modality   variable  value  paper    std       sig\n",
       "0      k-Nearest neighbors  Precision  0.752  0.582  0.160  3.187500\n",
       "1            Decision tree  Precision  0.612  0.521  0.129  2.116279\n",
       "2      Logistic regression  Precision  0.708  0.616  0.122  2.262295\n",
       "3    SVM (Gaussian Kernel)  Precision  0.622  0.511  0.114  2.921053\n",
       "4            Random Forest  Precision  0.533  0.614  0.100 -2.430000\n",
       "5                 AdaBoost  Precision  0.738  0.601  0.143  2.874126\n",
       "6   Gradient Boosting Tree  Precision  0.778  0.561  0.159  4.094340\n",
       "7      k-Nearest neighbors     Recall  0.827  0.636  0.158  3.626582\n",
       "8            Decision tree     Recall  0.637  0.550  0.200  1.305000\n",
       "9      Logistic regression     Recall  0.807  0.600  0.127  4.889764\n",
       "10   SVM (Gaussian Kernel)     Recall  0.927  0.670  0.090  8.566667\n",
       "11           Random Forest     Recall  0.513  0.664  0.193 -2.347150\n",
       "12                AdaBoost     Recall  0.770  0.717  0.179  0.888268\n",
       "13  Gradient Boosting Tree     Recall  0.810  0.616  0.148  3.932432\n",
       "14     k-Nearest neighbors         F1  0.775  0.608  0.126  3.976190\n",
       "15           Decision tree         F1  0.619  0.535  0.160  1.575000\n",
       "16     Logistic regression         F1  0.744  0.608  0.084  4.857143\n",
       "17   SVM (Gaussian Kernel)         F1  0.739  0.580  0.087  5.482759\n",
       "18           Random Forest         F1  0.512  0.638  0.151 -2.503311\n",
       "19                AdaBoost         F1  0.749  0.654  0.150  1.900000\n",
       "20  Gradient Boosting Tree         F1  0.779  0.587  0.107  5.383178"
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df_visual\n",
    "s_p = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['std_precision']].values]))\n",
    "s_r = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['std_recall']].values]))\n",
    "s_f = list(chain.from_iterable([np.round(x,3) for x in ow.df_score_stats.loc[2:8,['std_f1']].values]))\n",
    "print(s_p)\n",
    "print(s_r)\n",
    "print(s_f)\n",
    "df_std = pd.DataFrame(s_p)\n",
    "df_std.columns=['std']\n",
    "df_std = df_std.append(pd.DataFrame(s_r, columns=['std']), ignore_index=True)\n",
    "df_std = df_std.append(pd.DataFrame(s_f, columns=['std']), ignore_index=True)\n",
    "df_std\n",
    "df['std'] = df_std.loc[:,['std']]\n",
    "# (µ_self - µ_paper) / (std/3)\n",
    "df['sig'] = [(row.value-row.paper)*3 for i,row in df.iterrows()]\n",
    "df['sig'] = df['sig']/df['std']\n",
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3EAAAKSCAYAAAB801LUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3Xt8z/X///H7zsPmfAprGJvDDpbEirAyQ1MtFTKRXzmVToqpRPRxKDRn8nXKMSWJDSH56EB08NFHKGNTdJo5DNvee79+f/ju/fVuGxt7773X3K6Xi8vFnq/X6/l8vF97e9t9z9fr+XIxDMMQAAAAAMAUXJ1dAAAAAACg8AhxAAAAAGAihDgAAAAAMBFCHAAAAACYCCEOAAAAAEyEEAcAAAAAJkKIA8qg8+fPa+HChYqNjVXLli3VokUL9ejRQ6tXr5bVarXbNzIyUnFxcU6qVJoxY4aCgoJ04sQJW9uXX36pLl26KDg4WL1799batWsVFBSk3bt3O63OsqY0n9Nff/1VcXFxCg0NVevWrZWWluaUOk6cOKGgoCDNmDHDKePnCgoK0siRI0v82NTU1Os6DgDgeO7OLgBA8Tp69KgGDx6sX3/9VTExMXrooYeUmZmpbdu2afTo0frmm2/01ltvycXFxdmlSpI6deqkW2+9VVWrVpUkWa1Wvfjii3Jzc1N8fLxq166twMBATZ48WQEBAU6utuxo1apVqT2nkyZN0t69e/X000+rRo0atvdGSatataomT56soKAgp4zvTB9++KHGjh2r/fv3O7sUAEA+CHFAGZKZmakhQ4YoPT1dH3zwgZo0aWLb9sQTT2js2LFasWKFQkND1bdvXydW+n+aNGliV+eff/6ptLQ09e/fX4899pit3c/PzxnllVl+fn6l9pweOnRITZs21dChQ51aR/ny5XX//fc7tQZn+eabb5SZmensMgAABeBySqAMWbFihZKTkxUfH28XjHKNGDFClSpV0qpVq5xQXeFkZ2dLkipUqODkSuAs2dnZfP8BALgKQhxQhmzcuFHly5dXt27d8t3u7e2t999/X+vWrSuwD8MwtHLlSvXo0UPh4eEKCQlRdHS05s+fL8MwbPudOXNGI0eOVIcOHRQcHKx7771XU6ZMsfvtfVZWlt58803dc889Cg4OVvv27TV27FidOXPGts+V98TNmDFD99xzjyRp5syZtnu28rt/KzMzU9OmTVNkZKSCg4N1zz33KCEhQVlZWbZ9co/bvHmzIiMjFRYWdtX7m7KysjRjxgxFRUUpNDRUnTt31vz585WTk2Pb5+LFi5oyZYpt3MjISL399tu6ePFinnF/+uknDRs2TOHh4WrTpo0mTZqknJwcffTRR+rcubNatGihnj176qeffrI7H82aNdPRo0cVFxensLAwRUZGavbs2XZ1SNKPP/6oZ555RnfeeaeaN2+uiIgIvfjiizp16pRdfyEhIfr000911113KTw8XGvWrMn3nG7evFkPPfSQwsPD1bJlS/Xv31/79u2zGzMnJ0cLFixQ586dFRwcrLZt2+r111+3u29t9+7dCgoK0hdffKGxY8cqIiJCYWFhevzxx+1e6z/lHvfrr79qz549dvejFWXcjz76SDExMQoJCVF8fHy+3+dWrVpp0KBBebblnpdvvvkm33viDh06pAEDBqhNmzYKCwvTgw8+qA8++MCuj4LuQ/tne3Z2tubNm6fu3burRYsWCg0NVffu3fP0V1jLly9X586dFRoaqh49eujQoUN59inMmHFxcfroo4/y1FzYzwYAgONxOSVQRhiGoYMHD+q2226Th4dHgfvVr1//qv288847mjt3rh588EE98sgjysjI0Lp16zRlyhTVqFFDDz74oCTpueee03//+1/17dtXNWvW1Hfffaf58+crPT1d48aNkyS98cYb2rBhg/r27Ss/Pz8dOXJEy5cv1/Hjx7Vw4cI8Y3fq1Em+vr6aMGGCOnXqpE6dOikgIEC//vqr3X45OTkaOHCgvv32Wz3yyCMKCAjQgQMHNHfuXB08eFBz5syxu+cvPj5ecXFx8vX1VYsWLQp87UOHDtXOnTsVExOj/v37a//+/ZoyZYr+/vtvxcfHKysrS/3799f333+v2NhYBQcHa//+/Xr33Xe1b98+LV261O7cP/XUU2rZsqVGjhypLVu2aOHChTp8+LAOHTqkxx9/XIZhaM6cORo2bJgSExPl7u5u+172799fjRs31ksvvaTdu3crISFBp06d0htvvCHpcpjo3bu3/P399dRTT6lcuXL69ttv9fHHH+uPP/7Qe++9Z6vDYrHo1Vdf1YABA5SVlaWWLVvq+++/t3vte/bs0fPPP6+7775bDz/8sC5evKhly5apf//+2rhxo+3Sy+eff16bN29WVFSU+vbtq+TkZK1cuVJff/211qxZo4oVK9r6fPXVV1WzZk0NGTJEZ86c0YIFC/Tkk0/qs88+s73WKwUEBGjy5MmaMGGCqlSpokGDBtnuRyvKuG+88Ybuv/9+Pfzww6pTp06ecTw9PRUVFaX169fr3Llz8vX1tW1LTEzULbfcottvvz3P+y4tLU0DBgxQlSpVNHjwYHl5eWnjxo165ZVX5OXlpZiYmALfW/mJj49XUlKSevXqpbi4OJ0+fVrvv/++XnnlFd1666264447Ct3XjBkzNHPmTLVr1059+/bV/v377S5HLsqYgwYNktVq1d69ezV58mTdeuutkgr/2QAAKAEGgDLh77//NgIDA43nn3++SMd17NjR6NOnj2EYhpGVlWXcdtttefo4d+6cERwcbAwcONAwDMP466+/jMDAQGPBggV2+40cOdJ4/PHHbV+HhoYaY8eOtdtn2rRpRmxsrHH+/HnDMAxj+vTpRmBgoJGammoYhmGkpqYagYGBxvTp023HfPjhh0ZgYKDx9ddf2329c+dOu75XrVplBAYGGp9++qndfiNGjLjmedixY4cRGBhozJkzx679xRdfNJo3b26kp6cbK1asMAIDA41FixbZ7fPuu+8agYGBxvLly+3GfeaZZ2z7nD171mjevLnRpEkT4/Dhw7b2qVOnGoGBgUZycrLd+RgyZIhhtVrt6ggKCjJ+/vlnwzAMY/To0UZYWJhx+vRpu1qef/55IzAw0Nae29+V5zO/c/r6668b4eHhdmP+9NNPRlRUlJGUlGQYhmF8/vnnRmBgoDF+/Hi7vhITE43AwEBj8uTJhmEYxtdff20EBgYaDz30kGGxWGz7zZs3zwgMDDR27dqV5/xf6cr35PWMe+WxBfnqq6+MwMBA46OPPrK1paWlGc2aNTPeeustwzDyvhc3btxoBAYGGvv377cdk5mZaTz44IPG22+/bWsr6D13Zfsff/xhBAUF2R1nGIbxyy+/GIGBgca4ceOu2V+uv//+2wgODs7znsn93l/PmCNGjDACAwNtXxf2swEAUDK4nBIoI1xdL/9z/ucld0Xh4eGhL7/80jbbk+v06dPy8fHRhQsXJEm+vr4qX768VqxYoc2bN9vaJ0yYoMWLF9uOq127thITE7V27VqdPXtW0uUZvA8//PCG7nnasmWLqlatqubNmystLc32p3379nJzc9OOHTvs9m/btu01+9yxY4dcXV3Vp08fu/YRI0bo448/lo+Pj7Zv3y4fH588Mxx9+/aVj4+Ptm3bZtd+77332v7u6+urqlWrqn79+mrcuLGtvV69epIuL+hypaeeespuNrF///4yDEOfffaZJGnMmDHavn27KleubNvn/Pnz8vLykiTb96Sw56B27drKyMjQ+PHj9csvv0iS7VLU6OhoSdL27dslSQMHDrQ7tkuXLmrQoIG2bt1q1x4VFSU3Nzfb102bNs33tV5LUcctzPf7jjvuUK1atbRp0yZb25YtW2SxWAqcUatdu7YkacqUKdq7d69ycnLk6emptWvX6sUXXyzSa6pRo4b27dunIUOG2NoMw5DFYpEkZWRkFLqv3bt3KysrS4888ojde+afjw65kTEL+9kAACgZXE4JlBGVKlWSh4fHDT9Ty8PDQzt27NC2bduUnJys48eP2+5hM/73vhdPT0+98cYbeu211zRs2DB5enrqjjvuUFRUlB544AFbkBgzZoyee+45xcfH67XXXlOLFi3UqVMnPfTQQ3aXsBVVSkqK0tLSFBERke/2kydP2n1drVq1a/b566+/qlq1avLx8bFrr1GjhmrUqCHp8nPD/Pz88lyu6unpKT8/vzyX31WvXt3ua3d39zy15Iacfz6/759L//v7+9vqlCQXFxedPn1a8+bN06FDh5SSkqLffvvN9j36Z3/XOgd9+vTRrl27tGzZMi1btkz16tVTx44d1aNHD9siOSdOnFDFihXzvK7cenfu3GnX9s9HA3h6euZb27Xc6Lj5cXV1Vbdu3fTee+/ZLqlMTExUYGBggY8UuO222xQXF6dly5bpq6++UuXKldW2bVvFxMSoQ4cORXpN0uXzsX79eu3atUvHjh3T8ePHbUHKKMI9ZrnvidzLHnNVrlw5z/f9RsYszGcDAKBkEOKAMsLFxUXh4eE6cOCALBZLvvccSdK0adOUmpqq+Ph4WzjJZRiGXnrpJW3YsEEtW7ZUeHi4Hn30UbVq1UqPP/643b4xMTFq166dtm7dqs8//1xffvmldu3apRUrVmjNmjXy9PRURESEPvvsM9ufL774wjZbt3bt2ut+/ldOTo7q16+v119/Pd/tV94fJf3fLOW1+rzWs/Ou9oOq1WrNE+6unIXKVdjn8/2zr9zgk9vnjh07NGTIENWsWVNt2rTR3XffreDgYO3atUvz5s3L09+1zoGPj4+WLVum77//Xlu3btXOnTv13nvvafny5Zo8ebJiYmKK/PoLc94LozjOe35iYmK0cOFCbdu2TW3bttU333yj55577qrHvPrqq+rbt682b96snTt3avPmzdqwYYMeffTRPLNUV/rnDHlWVpYGDBigffv2qXXr1oqIiFC/fv10xx13FDkQ5r6n8nskwJWB+UbGLMpnAwDA8QhxQBnSqVMn7dmzR4mJierevXue7ZcuXdIHH3ygnJwcu8vwcu3du1cbNmzQkCFD9Oyzz9raLRaL0tPTbYtbZGRk6ODBg2rcuLF69OihHj16KCsrS2+99ZaWLl2qXbt2qW3btjp48KBq166tbt26qVu3brJarVq0aJEmT56sjRs35rncq7Dq1aunAwcOqE2bNnZBITs7W59++qntsreiqFOnjr788ktlZGTYXer5448/auHChRo8eLDq1q2r77//XtnZ2XbBISsrSydOnNDtt99+Xa8nP6mpqWrUqJHt62PHjkn6vxm5cePGyd/fXx9++KHKly9v2++TTz65rvGSk5N17tw5tWjRQi1atNDw4cP1888/67HHHtOiRYsUExOjunXrateuXfrrr7/yzIolJyfrlltuua6xr8VR4zZr1kwBAQHaunWrMjIyZLVaC1zZVZL++usvHTlyRBEREXryySf15JNP6vTp0xo6dKjef/99vfTSS/L19ZWrq6vdKqm5x14pMTFRe/bs0ZtvvqkePXrY2n///fciv47cf5fHjh2ze7TI+fPndfr06WIZs7CfDQCAksE9cUAZ8uijj6pu3bqaNGmSDh8+bLctJydHY8aM0V9//aUnn3wy3xUs09PTJckuPEjS+++/r4sXL9runTly5Igee+wxu2XJPT091axZM0mXZ0LS09P16KOP2s0Kubq6KiQkxPb36xUZGan09HStXLnSrn3VqlV6/vnn9dVXXxW5z/bt28tqtWrNmjV27StXrlRSUpKqV6+uyMhInT9/XsuXL7fbZ8WKFcrIyLiuS+oKcuXqkpK0aNEiubu7KzIyUtLl71WdOnXsAtzJkye1ZcsWSUW/N3L8+PEaMmSI3X1RDRs2VMWKFW3fq9yx/znTt3XrViUnJxfr67+SI8eNiYnRF198oU2bNqlly5aqW7dugfuuXbtW/fr103/+8x9bW5UqVeTv7y8XFxfbeapevbp++uknuxnExMREu74K+re2dOlSSbL9WyuMO++8U+XLl9eSJUvsjvvn+7QoY+a+ltyZvMJ+NgAASgYzcUAZ4uXlpZkzZ+qJJ55Qjx49bM/KSk9P16ZNm3Tw4EFFR0erf//++R4fHh4uHx8fTZgwQb/99psqVqyo3bt3KzExUV5eXrYf8MPCwnT77bdr2rRpOnnypIKCgnTy5EktW7ZMDRs2VEREhDw9PRUTE6MVK1bo4sWLCg8PV3p6upYtW6bq1aurS5cu1/06H374YX300UcaN26cfvzxR4WGhurw4cNavXq1mjdvrtjY2CL3GRkZqbvuuksTJ07UkSNHFBISou+++07r1q3T0KFDVblyZdu4EydO1OHDhxUcHKwDBw5o7dq1CgsL08MPP3zdr+mfPvroI50/f1633Xab/v3vf+uzzz7T0KFDbSHj7rvvVmJiokaPHq2QkBCdOHHC9gO1VLSFMaTLC6c8+eSTeuyxx2z3NW7dulUpKSmaNGmSpMtB95577tHSpUv1+++/q3Xr1jp27JhWrlwpPz+/PAuPFBdHjnvffffpnXfe0Z49ezR27Nir7vvAAw9o0aJFGjRokHr16qVatWrpwIEDWrdunR588EHbDO59992nhQsX6umnn1aHDh30448/Kikpye7y4TvvvFPu7u56+eWX9dhjj8nd3V2fffaZdu3aJQ8PjyJ9/3x8fPTSSy9p7Nixevzxx9WlSxcdOXJE69evV7ly5a5rzNxap0+frtatWxf6swEAUDIIcUAZ06xZM3388cdavHixdu7cqcTERBmGoaCgIP3rX/9SbGxsgfdlVa9eXfPnz9fbb7+t2bNny9PTUw0aNNDUqVO1f/9+LV261HZJ26xZszRz5kx99tlnWr16tSpVqqSoqCg9++yztgUsxo0bJz8/P23cuFEbN25UuXLlFBERoeeff/6674eTLs/6LV68WLNmzdLmzZu1fv161axZU7169dLQoUPtfnAtLFdXV82ePVuzZ8/WJ598ovXr1+vWW2/V6NGj1atXrzzjJiUlaf369apdu7YGDhyowYMHX/X5fEU1c+ZMzZo1S1u2bJGfn5/GjRunRx55xLZ9zJgxKl++vLZv366PP/5YtWvX1gMPPKBOnTqpV69e+vrrr20zo4XRtm1bzZkzR/PmzdPs2bOVmZmpxo0ba+rUqbZLDF1cXJSQkKB3331X69at0/bt21WtWjU9+uijeuaZZ/Lci1hcHDmun5+f7V7S3FU4C1KzZk0tXbpU06dP16pVq5Senq66devq6aef1pNPPmnb79lnn5XFYtHGjRu1a9cuhYWFacmSJRo+fLhtn8DAQE2fPl0zZ87U1KlTVaFCBTVu3FiLFi3SihUrtGfPnjyX7V5N79695evrq/nz52vSpEmqX7++Zs+erREjRlzXmLnvoQULFug///mP/ud//qfQnw0AAMdzMVhSCgBKjdyHNm/bts32+AEAAIArcU8cAAAAAJgIIQ4AAAAATIQQBwAAAAAmwj1xAAAAAGAiN+XqlFarVRkZGfLw8ChwlT4AAADcvAzDUHZ2tipUqHBDzzYFHOGmDHEZGRl5HoQMAAAA/FNgYKB8fX2dWsPZs2f1xx9/KDs726l1oGS4u7vL29tbNWrUkLe3d/77lHBNpULuc3cCAwNtz7MCAAAAcmVlZenw4cPF+gzQ63H27Fn9/vvvqlu3rsqVK8dVZGWcYRiyWCw6f/68UlJSVKtWLVWqVCnPfjdliMt983t6esrLy8vJ1QAAAKC0cnZo+uOPP1S3bl2VL1/eqXWgZLi4uMjDw0NVqlSRl5eXTp06lW+I4wJfAAAAoJTKzs5WuXLlnF0GnKBcuXLKzMzMdxshDgAAACjFnD0bCOe42vedEAcAAAAAJkKIAwAAAAATuSkXNgEAAADMLjsnRx5ubqasIS4uTnv27LFr8/DwUK1atdSpUyc999xzBS6vX1zi4uLk5uamxYsXO2R/RyLEAQAAACbk4eamF5OWOrWGKV36XvexISEhevXVV21fZ2Zm6ptvvtGsWbP0+++/a9q0acVRYoFef/31It1vWNT9HYkQBwAAAKDE+fj4qEWLFnZtrVu31qlTp/TBBx8oPj5eNWvWdNj4jRo1cuj+jsQ9cQAAAABKjWbNmskwDJ08eVKRkZGaOHGi4uLidNttt2nChAmSpNOnT+vVV19VRESEQkND1atXL+3bt8+un6ysLL3zzjuKjIxUWFiYYmJilJiYaNseFxenfv362b7+4osv9Mgjjyg8PFytWrXSkCFD9MsvvxS4/6VLl5SQkKDOnTsrJCREXbt21erVq+1qiIyM1MyZMzVx4kTdeeedCgsL04ABA3T8+PEbOkeEOAAAAAClxrFjxyRJfn5+kqT33ntPwcHBSkhIULdu3ZSZmal+/fppx44deuGFFzR9+nRVqlRJ/fr10/79+239DB8+XIsXL1bPnj01d+5ctWrVSi+88II+++yzPGOmpqZqyJAhCg4O1pw5czR+/HgdPXpUAwcOlGEYefY3DENPPvmklixZol69emnOnDm688479frrr2vWrFl2+y5evFjJycmaMGGCxo0bpwMHDig+Pv6GzhGXUwIAAAAocYZhyGKx2L4+ffq0du7cqVWrVik6OlpVq1aVJNWuXVsvv/yy7X60999/X4cOHdKaNWsUEhIiSbr77rvVo0cPTZs2TYsWLdLhw4e1efNmjR49Wo899pgkKSIiQikpKdq9e7c6duxoV8v+/ft16dIlDRw4ULVq1ZIk3XLLLdq2bZsyMjLk4+Njt//nn3+uPXv2KCEhQdHR0ZKktm3bymKxaO7cuerdu7eqVKkiSapcubJmz54tt/9dACYlJUUzZszQuXPn5Ovre13njhAHAAAAoMR9/fXXat68uV2bm5ub7r33Xo0ZM8bW1rhxY7sFRb766ivVqlVLTZs2tQuBHTt21Lx585SVlWW7tLJTp052/S9YsCDfWsLCwuTl5aUePXooOjpad999t1q3bq3Q0NB89//mm2/k4eGhqKgou/aYmBitXLlSP/zwgzp06GDr2+2KFTxr164tSbpw4QIhDgAAAIB5hIaGavTo0ZIkFxcXeXt7q27duipXrpzdftWqVbP7Oj09XadOncoTAHOdPn1a6enp+R5bkHr16mnZsmWaP3++PvjgAy1dulQVK1ZU79699dxzz+VZlfLMmTOqVq2aXF3t706rXr26JOncuXO2tn8+KiH3mPwu0ywsQhwAAACAElehQgXb5ZBF4evrq4CAAE2aNCnf7VWqVLHNcKWlpalGjRq2bYcPH9bFixcVFhaW57jQ0FDNnDnTNpO3evVqzZ07V82aNVPnzp3t9q1YsaL+/vtvWa1WuyD3559/2mpwJBY2AQAAAGAarVq10m+//aaaNWsqJCTE9mfbtm1677335OHhoZYtW0pSnkVM3nzzTU2dOjVPn++9954iIyOVlZUlT09PRUREaNy4cZKkkydP5tn/jjvuUHZ2trZs2WLXvmHDBnl4eBR4GWZxYSYOAAAAgGnExsZq2bJl6t+/v20hkh07dmjRokV6+umn5eLioqZNmyoqKkoTJkzQhQsXFBQUpK1bt2rPnj36n//5nzx9tmnTRpMnT9bQoUPVp08fubm5adWqVfLy8sqzCIp0eSGVVq1a6ZVXXtGpU6fUuHFjff7551q1apUGDx6sihUrOvQcEOIAAAAAE8rOydGULn2dXoPHFYt2lIQKFSpo+fLlmjJliiZOnKiMjAz5+fnptddeU58+fWz7TZkyRQkJCVq4cKHOnDmjgIAA26MA/qlx48aaN2+eZsyYoRdeeEE5OTkKDg7WwoUL5e/vn2d/V1dXzZs3T++8844WLFigM2fOqH79+hozZox69uzp0NcvSS7GjdxRZ1KZmZk6cOCAgoOD5eXl5exyAAAAUMqUlp8XDx48qKZNmzptfDhXQd9/7okDAAAAABMhxAFAGWVYsk3dPwAAyB/3xAFAGeXi7qE/5rzssP5rDp7ssL4BAEDBmIkDAAAAABMhxAEAAACAiRDiAAAAAMBECHEAAAAAYCKEOAAAAAAwEUIcAAAAAJgIIQ4AAAAATIQQBwAAAJiQYcl2dgnXXUNcXJyCgoLs/jRp0kS33XabYmNj9fHHHxdzpde2du1aBQUF6dSpU7Ya+/XrV+J1FAYP+wYAAABMyMXdQ3/MedmpNdQcPPm6jw0JCdGrr75q+9pqterUqVNasmSJXn75ZVWuXFnt27cvjjLLHEIcAAAAgBLn4+OjFi1a5Gm/++67FRERobVr1xLiCsDllAAAAABKDU9PT3l4eMjFxUXS5Rm6uXPn6t5771VwcLCio6O1Zs2aPMetW7dODzzwgMLCwhQZGanp06crJyfHtn3z5s3q1auXwsPDFRwcrC5dumjFihUl9rqKEzNxAAAAAEqcYRiyWCy2r3NycnTixAnNnj1bGRkZuv/++yVJY8aM0dq1azV48GCFhYXpiy++0GuvvaZLly4pLi5OkrR8+XK98cYbevTRRzV8+HAdPXpUb731li5duqSXX35Z27Zt07Bhw9SvXz8NGzZMly5d0ooVKzR27FgFBwcrNDTUKefgehHiAAAAAJS4r7/+Ws2bN7drc3FxUVBQkBISEtSxY0clJyfr/fff18svv6wnnnhCktS2bVvl5OQoISFBPXr0kJeXl2bNmqXo6Gi98cYbtn3Onj2rL774QoZh6JdfflFsbKzi4+NtY4WHh6t169bas2cPIQ4AAAAAriU0NFSjR4+WJP3+++9KSEiQxWLRtGnT1LBhQ0mXg55hGOrYsaPdrF1kZKSWLFmi/fv3q3r16vr777/VqVMnu/6ffvppPf3005Kkp556SpKUkZGh5ORkpaSk6D//+Y8kKTvb+at8FhUhDgAAAECJq1ChgkJCQiRdXqmyRYsW6t69uwYMGKAPP/xQVatWVXp6uiQpOjo63z7++OMPubtfjjTVqlUrcKy0tDS9/vrr2rp1q1xcXOTv76+WLVtKunxZp9kQ4gAAAAA4XfXq1TV69Gg9++yzevPNNzVlyhT5+vpKkpYtWyZvb+88x9SrV09//vmnpMtSb2P5AAAgAElEQVRB7Up//fWXjhw5ottuu03Dhw9XcnKyFi9erPDwcHl6eurixYv5LpBiBqxOCQAAAKBUiI6OVrt27bRhwwbt2bNHt99+uyTpzJkzCgkJsf05efKkpk+frosXL6phw4aqXLmytm/fbtfX6tWrNWTIEEnSvn37FB0drdatW8vT01OStHPnTkmXV780G2biAAAAAJQao0aNUvfu3TV+/Hh99NFHuu+++zRq1CilpqaqadOm+vnnnzV16lQ1b95cderUkXT5/rc333xTVapUUWRkpA4fPqz58+drwIAB8vLyUmhoqNavX6+mTZuqVq1a+vbbbzV//ny5uLjo4sWLTn7FRUeIAwAAAFBqNGzYUHFxcVq4cKFWrlypiRMnau7cuVq2bJl+//13Va9eXT169NCwYcNsx8TFxalcuXJauHChVq1apTp16mjYsGHq37+/JGnixIkaN26cbfXK+vXra+zYsVq/fr327dvnlNd5I1wMM97Jd4MyMzN14MABBQcHy8vLy9nlAIDD/DHnZYf1XXPwZIf1DQDOVlp+Xjx48KCaNm2a7zbDki0Xd48Srqj01VCWFfT95544AAAAwIRKQ3gqDTXcjAhxAAAAAGAihDgAAAAAMBFCHAAAAACYCCEOAAAAAEyEEAcAAAAAJkKIAwAAAAATIcQBAAAAgIkQ4gAAAADARAhxAAAAAGAihDgAAADAhCw5VmeXcN01jBw5UkFBQQX++fLLL+32P3jwoJo3b65Tp04VR9mm5+7sAgAAAAAUnbubq2Yv2+XUGob0aXvdx9auXVsJCQn5bmvUqJHt70ePHtXAgQNlsViue6yyhhAHAAAAoMR5enqqRYsWBW63WCxavXq1pkyZIg8PjxKsrPTjckoAAAAApc6+ffv09ttv64knntDw4cOdXU6pwkwcAAAAAKfI7xJJNzc3ubi4KCAgQFu3blW1atW0du1aJ1RXehHiAAAAAJS4lJQUNW/ePE/7mDFj1KtXL1WvXt0JVZkDIQ4AAABAiatdu7ZmzpyZp71u3bpOqMZcCHEAAAAASpynp6dCQkKcXYYpsbAJAAAAAJgIIQ4AAAAATIQQBwAAAAAmQogDAAAAABNhYRMAAADAhCw5Vg3p09bpNbi7FX1eaOLEiUXaPzY2VrGxsUUep6xiJg4AAAAwoesJT2WxhpsRZ93JDEt2mRgDAAAAQMngckonc3H30B9zXnboGDUHT3Zo/wAAAABKDjNxAAAAAGAihDgAAAAAMBFCHAAAAFCKGYbh7BLgBFf7vhPiAAAAgFLK3d1dFovF2WXACbKzs+Xm5pbvNkIcAAAAUEp5e3vr/Pnzzi4DTnD27Fn5+vrmu83pIW7Dhg3q1q2bQkND1aVLF61bt+6q+6elpSk+Pl5t27bVHXfcoYEDB+rYsWMlUywAAABQgmrUqKE///xTFy5c4LLKm4BhGMrKytJff/2l06dPq2rVqvnu59RHDCQlJWn48OHq27ev2rVrp61bt2rEiBHy9vZWdHR0nv0Nw9DQoUOVkpKil156SZUrV9b06dPVt29fffLJJ6pUqZITXgUAAMDNw7Bky8Xdw/RjmIW3t7dq1aqlU6dOKTMz09nloAS4ubnJ19dXt956q7y8vPLdx6khburUqerSpYtGjRolSWrXrp3OnDmjhISEfEPcsWPH9O2332rSpEl64IEHJEkBAQG69957tX37dj344IMlWj8AAMDNhmfclrxKlSoxWQE7TrucMjU1VSkpKYqKirJr79y5s44eParU1NQ8x+T+9qFChQq2ttw3dHp6ugOrBXAjDEt2mRgDAACgNHDaTNzRo0clSQ0aNLBr9/f3lyQlJyfLz8/PbluTJk3UunVrzZo1Sw0bNlSVKlU0ceJElS9fXvfee2/JFA6gyPitLQAAQPFxWog7d+6cJMnHx8euPXeWraBVeMaMGaP/9//+n7p27SpJ8vT01KxZs/IEPgAAAAAoi5wW4nJX13Fxccm33dU175Wev/zyi3r27Klbb71Vo0aNkre3t95//30NGzZMCxYs0O23316kGg4cOHCd1Refli1blsg4+/btK5FxgPzwPneOkjjvnHPg5sNnOuB8Tgtxuc88+OeMW0ZGht32Ky1evFiStHDhQtu9cHfddZd69+6tf/3rX1q7dm2RaggODi5wxZeypqQ+cAFn4n1e8jjnABzF2Z8vmZmZpeIX/kB+nLawSe69cCkpKXbtx48ft9t+pd9++00BAQF2q/O4uLioZcuW+vnnnx1YLQAAAACUDk4Lcf7+/qpXr542bdpk175lyxbVr19fderUyXNMgwYNdOTIEZ05c8au/YcfflDdunUdWi8AAAAAlAZOfU7c0KFDFR8fr0qVKqlDhw7avn27kpKSNG3aNElSWlqaUlJS1KhRI/n4+Khfv35av369BgwYoKeeekre3t76+OOPtWfPHtsxAAAAAFCWOTXExcbGKisrSwsXLtSaNWvk5+enSZMm2Vae3LFjh+Lj47V06VK1bt1a9erV08qVK/XWW29p5MiRcnV1VWBgoBYtWqQ777zTmS8FAAAAAEqEU0OcJPXs2VM9e/bMd1tsbKxiY2Pt2gICAjR37tySKA0AAAAASh2n3RMHAAAAACg6QhwAAAAAmAghDgAAAABMhBAHAABQRmTn5Di7hGJhybGWiTEAR3H6wiZwPEuOVe5ujs3rJTEGAAC4Og83N72YtNShY0zp0teh/UuSu5urZi/b5dAxhvRp69D+AUcixN0E+CAEAAAAyg6mTgAAAADARAhxAADAtAxLdpkYAwCKgsspAQCAabm4e+iPOS87dIyagyc7tH8AKCpm4gAAAADARAhxAAAAAGAihDgAAAAAMBFCHAAAAACYCCEOAAAAAEyEEAcAAAAAJkKIu4rsnBxnlwAAAAAAdnhO3FV4uLnpxaSlDh1jSpe+Du0fAADcGEuOVe5ujv29d0mMAaDsIMQBAABchbubq2Yv2+XQMYb0aevQ/gGULfzKBwAAAABMhBAHoEyw5FjLxBgAAADXwuWUAMoELncCAAA3C2biAADXhdlPAACcg5k4AMB1YfYT15KdkyMPNzdnlwEAZQ4hDgAAOASP6gEAx+BySgAAAAAwEUIcAAAAAJgIIQ4AAAAATIQQBwAAAAAmQogDAAAAABMhxAEAAACAiRDiAAAAAMBEeE4cAAAwLcPIVuWHBjl8DAAoTQhxAADAtAwjR5npux06hnv1Og7tHwCKisspAQAAAMBECHEAAAAAYCKEOOAml52T4+wSAAAAUATcEwfc5Dzc3PRi0lKHjjGlS1+H9g8AAHAzYSYOAAAAAEyEEAcAAAAAJkKIAwAAAAATIcQBAAAAgIkQ4gAAAADARAhxAAAAAGAihDgAAAAAMBFCHAAAAACYCCEOAAAAAEyEEAcAAAAAJkKIAwAAAAATIcQBAAAAgIkQ4gAAAADARAhxAAAAAGAihDgAAAAAMBFCHAAAAACYiLuzCwAAOIZhZKvyQ4Mc2j8AACh5hDgAKKMMI0eZ6bsd1r979ToO6xsAABSMyykBAAAAwESYiQMAoJgYRrYMI8ehY7i4uMnFxcOhYwAASjdCHAAAxcQwcpSdfcihY3h4BBHiAOAmx+WUAAAAAGAihDgAAAAAMBEupwQAALgKi8VV3e9p7tAxsrMt8vDgxzIAhcOnBQAAwFVYrdL336U4dIxO9zRzaP8AyhYupwQAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAiPGMBNx7Bky8Xdw/RjALg5WSyuMoxsh47h6urCM8sAoBTjE9rJDCNblR8a5PAx8H9c3D30x5yXHTpGzcGTHdo/gJuX1Sr9fPiUQ8doHFTbof0DAG4MIc7JDCNHmem7HTqGe/U6Du0fAAAAQMnhnjgAAAAAMBFCHAAAAACYCCEOAAAAAEyEEAcAAAAAJkKIAwAnyM7JcXYJAADApFidEnAAS45V7m6O/R1JSYwBx/Fwc9OLSUsdOsZbnR9xaP8AAMA5CHGAA7i7uWr2sl0OHWNIn7YO7R8AAAClEyEOuMldsmTrqVadHD4GAKBsMIxsVX5okMPHAFAwQhxwk7NYrfrk0A8OHWNQq44O7R8ojEuWbFmsVoeOUc7dcGj/QGlgGDnKTN/t0DHcq9dxaP+A2Tk9xG3YsEFz5sxRamqq6tatq4EDB+qBBx4ocH+r1ap58+bpgw8+0J9//il/f38NGjRI3bp1K8Gq4SjZOTnycHNzdhkAyiCL1aqdx444dIyogACH9g8AgOTkEJeUlKThw4erb9++ateunbZu3aoRI0bI29tb0dHR+R7zr3/9S6tXr9YLL7ygJk2aaOPGjXrxxRfl4+Oj9u3bl/ArQHEricUepnTp69D+AQAAAEdyaoibOnWqunTpolGjRkmS2rVrpzNnzighISHfEJeSkqLly5frjTfe0MMPPyxJioiI0LFjx/Tvf/+bEFcAi8VV3e9p7tAxsrMt8vBw+sQubmLZ2Rbe5wAA4KbgtJ9GUlNTlZKSohdeeMGuvXPnzkpKSlJqaqr8/Pzstm3dulXe3t55LrdctmyZw+s1M6tV+v67FIeO0emeZg7tH7gWq9XgfQ4AAG4KTnvI1NGjRyVJDRo0sGv39/eXJCUnJ+c55tChQ2rQoIG+/PJLde/eXc2aNVNUVJQSExMdXzAAAAAAlAJOm4k7d+6cJMnHx8euvUKFCpKk8+fP5zkmLS1NJ0+e1KhRo/Tss8+qXr16WrNmjZ5//nlVrVpVbdq0KVINBw4cuOr2li1bFqm/m92+fftuuA/OedEUxzkPCgkuhkquzstN8nT4ctQlsypgcZxzifd6URTXOfcLaKgzZ84US18FsVqtOnMm3aFj+PrkOPx1XLhQUQcO/HzD/TQPC3X4I0wMlZ0VQfl/tOQV1+cLUNKcFuJyf+BycXHJt93VNe8kYXZ2ttLS0jR37lx17Hh5yfKIiAgdPXpUM2fOLHKICw4OlpeX1/WUj3zwH0fJK45zfj4rsxgquToXWXXJwctRe1Ss59D+c/E+L3nFdc7PZ2WqUqVKxdJXQVxdXVWpUmWHjmEYbg5/HeXLly+2zxdHP8LkuTZl5354Pl9K3tXOeWZm5jV/4Q84i9Mup/T19ZWUd8YtIyPDbvuVKlSoIDc3N9111122NhcXF9155506dOiQA6sFAAAAgNLBaSEu9164lBT7hQiOHz9ut/1K/v7+slqtslgsdu3Z2dl5ZvQAAAAAoCxyWojz9/dXvXr1tGnTJrv2LVu2qH79+qpTp06eY9q1ayfDMJSUlGRrs1gs+ve//80lCAAAAABuCk594NHQoUMVHx+vSpUqqUOHDtq+fbuSkpI0bdo0SZcXMklJSVGjRo3k4+OjiIgItW/fXuPHj9eFCxdUv359rVixQr/++qumTJnizJcCAAAAACXCqSEuNjZWWVlZWrhwodasWSM/Pz9NmjRJXbt2lSTt2LFD8fHxWrp0qVq3bi1Jmj59uhISEjR//nydOXNGzZo108KFCxUc7PgV9gAAAOB4Fourut/T3KFjZGdb5OHh1B+Fgevm9Hduz5491bNnz3y3xcbGKjY21q7N29tbI0aM0IgRI0qiPAAAAJQwq1X6/ruUa+94Azrd08yh/QOO5LR74gAAAAAARef0mbjS7JIlm4eUAkABuNwJAADn4H/Gq7BYrTykFAAKwOVOAAA4B5dTAgAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACbCc+IAAADKiEuWbD3VqpNDxzBkOLR/ANdGiAMAACgjLFarPjn0g0PHeK5Ne4f2D+DauJwSAAAAAEyEEAcAAAAAJkKIAwAAAAAT4Z443HQMI1uVHxrk8DEAAAAARyDE4aZjGDnKTN/t0DHcq9dxaP8AAAC4eXE5JQAAAACYCCEOAAAAAEyEEAcAAAAAJkKIAwAAAAATIcQBAAAAgImwOiXgABaLq7rf09yhY2RnW+ThwT9hAACAmw0/AQIOYLVK33+X4tAxOt3TzKH9A2WNYRjOLgEAgGJBiAMA3BRcXFz03z9POHSMqIAAh/YPAIB0nffE/f777/rhhx907tw5ZWVlyWq1FnddAAAAAIB8FCnE7du3T7GxserQoYN69uypAwcOaM+ePerQoYMSExMdVSMAAAAA4H8VOsTt379f/fv3V0ZGhh5//HFbe6VKleTu7q7hw4fr888/d0iRAAAAAIDLCh3iEhISVK9ePX388cd66qmnbDeIh4SEaP369QoICNC8efMcVigAAAAAoAgLm3z33XcaMmSIvL29dfHiRbttPj4+euSRRzR9+vRiLxA3l0uWbD3VqpNDxzDECnUAAAAwryKtTunp6VngtszMTBY4wQ2zWK365NAPDh3juTbtHdo/AAAA4EiFvpwyLCxMGzZsyHfbhQsXtGbNGoWEhBRbYQAAAACAvAod4oYNG6b//ve/6tOnj9atWycXFxft379fS5cu1f33368TJ05o0KBBjqwVAABwSTgA3PQKfTlleHi45s2bp9dff12TJk2SJE2bNk2SVKNGDU2bNk1t2rRxTJUAUMZw/2fZlXX8vw7t3+PWpg7tX5IM3joAUKoV6Z64u+66S59++ql+/PFHpaamymq1qm7dugoODpa7e5G6AoCbGvd/4kYcO5Hm0P4bB9V2aP8AgBtT5OTl4uKi4OBgBQcHO6IeAAAAAMBVFDrE9e3bt1D7LV269LqLAQAAAABcXaFD3IkTJ/K0Wa1WnT59WpmZmapbt64aN25crMUBAAAAAOwVOsRt37493/acnBxt27ZNr776qgYMGFBshQEAAAAA8ir0IwYK4ubmpqioKD388MN6++23i6MmAAAAAEABbjjE5apfv75++umn4uoOAAAAAJCPYglxWVlZWr9+vapVq1Yc3QEAAAAACnDDq1NmZWUpOTlZZ8+e1TPPPFNshQEAAAAA8rqh1Smly/fENWzYUPfdd5969+5dbIUBAAAAAPK64dUpAQAAAAAlp9gWNgEAAAAAOF6BM3EF3QN3NS4uLlqyZMkNFQQAAAAAKFiBIa6ge+AAAAAAAM5TYIjjHjgAAAAAKH2K9Z64tLS04uwOAAAAAPAPhV6dUpLWrVunLVu26MKFC7Jarbb2nJwcZWRk6Oeff9aBAweKvUgAAAAAwGWFDnHvvvuupk6dKg8PD/n4+Oj06dOqXbu20tPTdfHiRXl7eysuLs6RtQIAAADATa/Ql1OuXbtWTZo00ZdffqnVq1fLMAwtXbpUe/fu1ejRo5WZmamwsDBH1goAAAAAN71Ch7hff/1V999/v3x8fOTn56dKlSpp7969cnNzU+/evdW1a1ceLwAAAAAADlboEOfu7q4KFSrYvvb399ehQ4dsX7du3VrHjh0r1uIAAAAAAPYKHeICAgL03Xff2b5u0KCB3SImZ8+eVVZWVvFWBwAAAACwU+gQFxsbq7Vr12r48OG6cOGCIiMjtXfvXs2cOVOJiYlavHixmjRp4shaAQAAAOCmV+jVKXv16qVTp05p+fLlcnd3V1RUlLp166aZM2dKknx8fDR8+HCHFQoAAAAAKEKIe+WVVxQTE6Onn35aHh4ekqQpU6aoV69eSk9PV3h4uKpVq+awQgEAAAAARQhxGzdu1Nq1a1WjRg3dd999iomJUdOmTXX77bc7sj4AAAAAwBUKfU/c119/rSlTpig0NFTLly9XbGysunbtqrlz5yo1NdWRNQIAAAAA/lehZ+K8vb3VtWtXde3aVefPn9fWrVuVlJSkWbNmKSEhQWFhYYqJidFjjz3myHoBAAAA4KZW6Jm4K/n4+OiBBx7QvHnzlJiYqPbt2+v777/X+PHji7s+AAAAAMAVCj0Td6W0tDR9+umnSkpK0t69e5WTk6NWrVqpe/fuxV0fAAAAAOAKhQ5xp0+f1pYtW7Rp0yZ98803slgsCgoK0nPPPaf77rtPtWvXdmSdAAAAAAAVIcS1bdtWVqtVt9xyiwYMGKCYmBg1atTIkbUBAAAAAP6h0CHu4YcfVkxMjFq2bOnIegAAAAAAV1HoEDdmzBgHlgEAAAAAKIzrWp0SAAAAAOAchDgAAAAAMBFCHAAAAACYCCEOAAAAAEyEEAcAAAAAJkKIAwAAAAATIcQBAAAAgIkQ4gAAAADARAhxAAAAAGAihDgAAAAAMBFCHAAAAACYiNND3IYNG9StWzeFhoaqS5cuWrduXaGPPXnypFq2bKnZs2c7sEIAAAAAKD2cGuKSkpI0fPhw3XXXXZo1a5buuOMOjRgxQps2bbrmsYZhaNSoUTp//nwJVAoAAAAApYO7MwefOnWqunTpolGjRkmS2rVrpzNnzighIUHR0dFXPXbFihU6evRoSZQJAAAAAKWG02biUlNTlZKSoqioKLv2zp076+jRo0pNTb3qsW+//bbGjRvn6DIBAAAAoFRxWojLnUVr0KCBXbu/v78kKTk5Od/jrFarRo4cqS5duujuu+92bJEAAAAAUMo47XLKc+fOSZJ8fHzs2itUqCBJBd7rtmTJEqWmpmru3LmOLRAAAAAASiGnhTjDMCRJLi4u+ba7uuadJDx69KjeeecdTZ8+Xb6+vjdcw4EDB666PSgk+IbHuJns27fvhvvgnBcN57zkFcc5lzjvRcE5dw4+X0oe57zkFdfnC1DSnBbickPYP2fcMjIy7LbnysnJ0ciRIxUdHa277rpLFovFts1qtcpiscjdvWgvJzg4WF5eXgVuP5+VWaT+bnYtW7a84T4450XDOS95xXHOJc57UXDOnYPPl5LHOS95VzvnmZmZ1/yFP+AsTrsnLvdeuJSUFLv248eP223PdfLkSf3www9at26dmjdvbvsjSTNmzLD9HQAAAADKMqfNxPn7+6tevXratGmTOnXqZGvfsmWL6tevrzp16tjtX7NmTX3wwQd5+unRo4d69eqlhx56yOE1AwAAAICzOfU5cUOHDlV8fLwqVaqkDh06aPv27UpKStK0adMkSWlpaUpJSVGjRo3k4+OjkJCQfPupWbNmgdsAAAAAoCxx2uWUkhQbG6uxY8dq165dGjp0qPbs2aNJkyapa9eukqQdO3bo0Ucf1Y8//ujMMgEAAACg1HDqTJwk9ezZUz179sx3W2xsrGJjY696/KFDhxxRFgAAAACUSk6diQMAAAAAFA0hDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmAghDgAAAABMhBAHAAAAACZCiAMAAAAAEyHEAQAAAICJEOIAAAAAwEQIcQAAAABgIoQ4AAAAADARQhwAAAAAmMj/Z+/OA2O6+/7/P2cy2RNZEImICInYt1gaUULVXpRLtdqiqK201hJVyqVVbSlVSy211Y3auiCt4CJt0diKorY0xBKNNYvsmd8f/WXupHpf37bXVWPM6/EXZyaT9zkzOXNe57MpxImIiIiIiNgQhTgREREREREbohAnIiIiIiJiQxTiREREREREbIhCnIiIiIiIiA1RiBMREREREbEhCnEiIiIiIiI2RCFORERERETEhijEiYiIiIiI2BCFOBERERERERuiECciIiIiImJDFOJERERERERsiEKciIiIiIiIDVGIExERERERsSFWD3FbtmyhY8eO1KlTh/bt2/PZZ5/92+enpqYyceJEWrZsSf369enWrRuxsbH3qVoRERERERHrMlnzl8fGxjJmzBh69+7No48+yo4dOxg3bhwuLi60a9funufn5uYyYMAA0tPTefnll/Hz8+Prr79mxIgRFBQU0KlTJyvshYiIiIiIyP1j1RA3a9Ys2rdvz4QJEwB49NFHuXPnDnPmzPndEBcfH89PP/3E+vXrqVOnDgBRUVFcuXKFxYsXK8SJiIiIiMhDz2rdKZOTk7l48SJt2rQpsb1t27YkJiaSnJx8z8+4u7vTs2dPateuXWJ75cqVuXjx4t9ar4iIiIiIyIPAai1xiYmJAISEhJTYHhwcDMDPP/9MUFBQicciIyOJjIwssS0vL489e/YQFhb2N1YrIiIiIiLyYLBaS1x6ejoAHh4eJba7u7sDkJGR8Yde57333iMpKYmBAwf+dwsUERERERF5AFmtJc5sNgNgMBh+d7vR+O/zpdls5t1332X58uX079+f1q1b/+kafvzxx3/7eHjtWn/6Ne3ZoUOH/uPX0DH/c3TM77//xjEHHfc/Q8fcOnR+uf90zO+//9b5ReR+s1qI8/T0BO5tccvMzCzx+O/Jzc1l/PjxbN26lf79+/Pqq6/+pRpq1aqFs7Pz//l4Rm7OX3pdexUREfEfv4aO+Z+jY37//TeOOei4/xk65tah88v9p2N+//27Y56Tk/P/vOEvYi1WC3FFY+EuXrxIeBfRff4AACAASURBVHi4ZfuFCxdKPP5bGRkZDBo0iMOHDzNhwgT69Onz9xcrIiIiIiLygLDamLjg4GAqVKjAV199VWL79u3bqVSpEuXLl7/nZwoKChgyZAhHjx5l1qxZCnAiIiIiImJ3rLpO3EsvvURMTAxeXl5ER0eza9cuYmNjef/99wG4efMmFy9eJDQ0FA8PD9auXUtCQgI9e/YkICCAH374wfJaBoOBunXrWmtXRERERERE7gurhrhu3bqRm5vLxx9/zPr16wkKCmLGjBl06NABgN27dxMTE8PKlStp0qQJX3/9NQDr1q1j3bp1JV7LwcGBkydP3vd9EBERERERuZ+sGuIAnn76aZ5++unffaxbt25069bN8v+VK1fer7JEREREREQeSFYbEyciIiIiIiJ/nkKciIiIiIiIDVGIExERERERsSEKcSIiIiIiIjZEIU5ERERERMSGKMSJiIiIiIjYEIU4ERERERERG6IQJyIiIiIiYkMU4kRERERERGyIQpyIiIiIiIgNUYgTERERERGxIQpxIiIiIiIiNkQhTkRERERExIYoxImIiIiIiNgQhTgREREREREbohAnIiIiIiJiQxTiREREREREbIhCnIiIiIiIiA1RiBMREREREbEhCnEiIiIiIiI2RCFORERERETEhijEiYiIiIiI2BCFOBERERERERuiECciIiIiImJDFOJERERERERsiEKciIiIiIiIDVGIExERERERsSEKcSIiIiIiIjZEIU5ERERERMSGKMSJiIiIiIjYEIU4ERERERERG6IQJyIiIiIiYkMU4kRERERERGyIQpyIiIiIiIgNUYgTERERERGxIQpxIiIiIiIiNkQhTkRERERExIYoxImIiIiIiNgQhTgREREREREbohAnIiIiIiJiQxTiREREREREbIhCnIiIiIiIiA1RiBMREREREbEhCnEiIiIiIiI2RCFORERERETEhijEiYiIiIiI2BCFOBERERERERuiECciIiIiImJDFOJERERERERsiEKciIiIiIiIDVGIExERERERsSEKcSIiIiIiIjZEIU5ERERERMSGKMSJiIiIiIjYEIU4ERERERERG6IQJyIiIiIiYkMU4kRERERERGyIQpyIiIiIiIgNUYgTERERERGxIQpxIiIiIiIiNkQhTkRERERExIYoxImIiIiIiNgQhTgREREREREbohAnIiIiIiJiQxTiREREREREbIhCnIiIiIiIiA1RiBMREREREbEhCnEiIiIiIiI2RCFORERERETEhijEiYiIiIiI2BCFOBERERERERuiECciIiIiImJDFOJERERERERsiEKciIiIiIiIDVGIExERERERsSEKcSIiIiIiIjZEIU5ERERERMSGKMSJiIiIiIjYEIU4ERERERERG6IQJyIiIiIiYkOsHuK2bNlCx44dqVOnDu3bt+ezzz77t8/PzMxkypQpREVFUb9+fV588UWSkpLuT7EiIiIiIiJWZtUQFxsby5gxY4iKimLevHk0btyYcePG8dVXX/2fPzNy5Ei++uorxowZw4wZM7h27Rq9e/cmPT39PlYuIiIiIiJiHSZr/vJZs2bRvn17JkyYAMCjjz7KnTt3mDNnDu3atbvn+QcPHmTPnj0sXryY5s2bA9CwYUMee+wx1qxZw8CBA+9r/SIiIiIiIveb1VrikpOTuXjxIm3atCmxvW3btiQmJpKcnHzPz3z33Xe4u7sTFRVl2ebr60ujRo2Ij4//22sWERERERGxNqu1xCUmJgIQEhJSYntwcDAAP//8M0FBQff8THBwMA4ODiW2V6xYkdjY2D/8u81mMwC5ubn/9nl5ebm4GP7enJubk0te4d/7Nphzc3Ewmv/W3/HrsSz8j19Hx/yP0zEvyZaOOTwcx13H/F76rJekY/7H6ZiX9CAc86LrxKLrRpEHicFspU/mli1bGD16NDt37qRChQqW7RcuXKBNmza8//77dOjQocTP9O/fn9zcXFatWlVi+/vvv8/y5cs5evToH/rd6enpnDlz5j/fCRERERF5qFWtWhVPT09rlyFSgtVa4oqyo8Fg+N3tRuO9d5H+Xd78vef/X9zd3alatSqOjo73/H4REREREbPZTF5eHu7u7tYuReQeVgtxRXc0MjIySmzPzMws8XhxHh4eXLp06Z7tmZmZeHh4/OHfbTQadUdFRERERP4tFxcXa5cg8rusNrFJ0Vi4ixcvlth+4cKFEo//9meSk5PvaZG7cOHC7z5fRERERETkYWO1EBccHEyFChXuWRNu+/btVKpUifLly9/zM82aNSMtLY29e/datt28eZODBw/StGnTv71mERERERERa3N444033rDWL/f09GTBggXcunULg8HAsmXL2Lx5M5MnTyYsLIybN29y+vRpPDw8cHJyIjAwkISEBP7nf/4Hb29vrly5woQJEzCbzbz11ltq8hYRERERkYee1WanLLJ27Vo+/vhjrl69SlBQEAMHDqRr164AbNq0iZiYGFauXEmTJk0AuHPnDm+//TY7duygsLCQiIgIxo8fT+XKla25GyIiIiIiIveF1UOciIiIiIiI/HFWGxMnIiIiIiIif55CnIiIiIiIiA1RiBMREREREbEhCnEiImLzCgsLATCbzfesJSolpaenW7uEh0LRZ05ExBoU4kRExOZlZGQAkJ+fj8FgsGzXhXZJ+/btY8SIESQmJgKwY8cOsrKyrFyV7Th8+DCrVq0CwGg0kpmZaeWKRMRemaxdgDxcCgoKcHBwoLCwkLt37+Lq6oqDg4O1y7JJhYWFGI26z/KgMJvNJcKBPDhycnIYMWIELi4uFBQUEB0dTVBQEM2aNdPf0G+EhYVx48YNnnvuOSIjIzl58iTVqlWjQoUK1i7NJqSmprJq1Spyc3Px8vLi7NmzvPzyy7i7u1u7NPmb6LtYHlRWXexbHi5FAQ7gzTffZM6cOcTHx1OvXj28vLx0Efw7CgoKMBqN/PLLL/z0008cPHiQlJQUAgMDMZlMOmYPiOLvw8WLF0lNTSU/P18Xbg8Io9FI7dq1KV26NK6urqxatYr4+Hg+++wzQkJCKFOmDCaT7lkCODs707RpU/71r3+RkJBAy5YtefLJJ61dls1wcnLizJkzxMbGsmXLFho1akR0dLS1y5K/idlstgS4AwcOsHfvXnbv3o3ZbMbZ2VnfAWJVWidO/iuKX+T269ePK1eu0KRJEwIDA3nmmWfw9PQsEfLkf0Pv1atXGTBgABkZGaSmpuLj44OHhwdvvfUWERER1i5T+N/P96pVq1i2bBl37tzBy8uLcePG0bZtW2uXZ9cOHjxIeHg4np6elm3Xr18nISGBNWvWcOTIEQYMGECvXr3w8/OzYqUPjtu3b9O7d28MBgOnT5+mf//+jB49Wq0Nf1BycjKdO3fGaDRSr149xo4dS7Vq1QC12jxsis79H330EatXr6agoACDwUBeXh41atRgyJAhNG7c2Nplip1SS5z8VxQFuGXLlhEfH8/ixYt58skniYiI4NChQ8yZM4e4uDiysrIsX3b2zmg0YjabefbZZ6lYsSLjxo3j9ddfp0qVKqxZs4bc3FwiIyNxdna2dql2z2AwsHfvXsaPH8/TTz9NREQEd+/eZfHixZhMJho2bGjtEu3S559/ztChQ/H09CQoKMhyV9zNzY2wsDAee+wxHB0dWbhwITk5OURGRuLg4GCXLdyFhYWWfXZwcOCRRx6ha9eulCpViqVLl3L69GmaNm2Ki4uLlSt9sBUUFHDp0iWcnZ2pX78+hw4dYv/+/fj5+VGpUiUMBkOJYy22zWAw8N133/HGG28wfvx4hg4dyujRo0lJSWHbtm20atUKb29v3NzcrF2q2CGFOPmviouLw8nJiV69epGUlMTHH3/MpEmTyMjI4PLly5w5c4YmTZrg4+Nj7VIfCHv37uXrr7/m1VdfpUGDBjg5ObF69Wpu3LjB+PHjWbduHc7OzpQvX97apdql/Px8jEYjubm5rFu3jqpVqzJ69GgiIyOpXr06hYWFLF26lEuXLtG6dWtrl2t3qlSpwu3bt/n444/JysqicuXKeHl5WR53cXGhSZMmlClThnnz5nH37l2aNWtmdxfYxXtBHDt2jAMHDlC1alX8/f2pXr06FSpUYP369ezevZumTZvi5OREfHw8AQEB6oZKyQBsNBrx9vamRYsWREZG4uHhwYEDB/j2229xcnKiZs2aGAwG7ty5o0D8kNi0aRN5eXkMGTKEwMBA8vPz+ec//0n79u2pXbs269evp379+jg5OVm7VLEzOjvLf6To4iApKQkfHx8KCws5evQo8+fPZ8eOHZw5c4bBgwfTr18/zp8/zzPPPMOlS5cICQmxdukPhIKCAksXSgcHB1atWsWaNWuYN28e/v7+xMXF4eDgoG6VVmIymcjKyuL1118nNTWViIgIyxd1eHg4Q4YMwc/Pj8WLF3P58mUWLFiAh4eHlau2H46OjkyePJng4GDeeecdUlJSGD58ONWrVy8R1Lp160ZWVhZLly6lZs2aPPHEE3bTGmc2my0BbsKECezfv58rV67Qt29fhg0bhqenJ507d6Zs2bJMnz6dbt26Ubt2bX766SfWr1+vG0hg6R65fv164uLiSE5OZtiwYXTs2JEnn3ySMmXKsGDBApYsWUJSUhLBwcGcOHGC0aNHU6pUKStXL/+pa9eucePGDUqXLg1Az5498fPzIyYmhhMnTrBy5Uoee+wxmjRpYuVKxd6o47b8ZUUXB0XjA3bt2sUrr7xCUFAQCxcuxNvbm2nTpjFixAg8PT1xcXEhJCRE4wWK8fPzo7CwkMOHD3Py5EneeustYmJiaNmyJZ6enjg7O1uOl4avWkd+fj4HDhzg+++/54cffiA1NdXyWEBAAL179yYmJoYffviBF154gcLCQr1X91nfvn354IMPOHjwIK+//jr79+8nNzfX8riTkxNt2rShWrVq7N69G8AuAhz8735++OGH7Nmzh5EjR7Jt2zZGjRpFQUEB58+f59KlS0RHRzN79mxatWrFrVu3+Oc//6kAx/+edzds2MCMGTNwc3OjWbNmVKlSxfKcRx99lGnTphEWFsamTZuYNGkSfn5+CnA2rui9b9CgAbdu3SIhIYGRI0dy8+ZN3n77bVxdXcnLy6NMmTIlxuSK3Ddmkb+goKDAbDabzYWFhebNmzebR40aZb5w4YLl8Z9//rnE8xMTE80vvfSSuXPnzvezzAdKYWHhPdsKCgrM48aNM9erV89cvXp189tvv23ZfvToUXOzZs3MGzduvN+lyu8YNWqUOTw83Dxt2rR7Pt85OTnm2NhY8/nz561TnJ0pOv/81o8//mh+/PHHzc2bNzdv2bLFnJ6eXuLxH374wdygQQPzrl277keZD4zLly+bO3ToYF69erXZbDab8/LyzN988425TZs25gYNGpjbtGlj3rp1q+X5vz1u9i4pKcnctGlT85IlSyzb0tPTzbNmzTIPHz7cvH79erPZbDbfuHHDvG/fPvP3339vrVLlP/B739Fms9l87tw5c1RUlLlu3brmiIgI8/79+81ms9mclZVlnjdvnvmxxx4zp6Sk3M9SRcxms9msMXHylxTd3R0+fDibNm2ibNmydO/e3dJtx9vbm4KCAvr06cPHH3/Mtm3bSE1NZdmyZXh4eFim1rcX+fn5ODg4WMYGJiQkYDabcXFxoW7duvz444+kp6cTEBBApUqViI2NZf78+QQFBTFu3Dhrl2+X8vLyuHv3Lrdv38bDw4O2bduSm5vLsmXLSE1NJTAwkHLlygG/ThQRGhqKj4+P3XTTs5bi47tOnDhBfHw8V69eJTMzk1q1atGhQwfLzJSenp4EBwfj6uoKgL+/P5cvX8ZsNtOgQQNr7sZ9lZ+fz4YNG4iMjMRkMrF48WKmT59OaGgo3bp1IykpiUuXLtG+fXuMRqPG9vzGqVOnOHjwIFOmTKGwsJCdO3fyyiuvsHfvXtLT0/nyyy8JCQmhTp06VKhQgcDAQGuXLH9B0Xn79OnTHD16lO+//x5HR0eqVq1KdHQ0e/fu5c6dOwQGBpKUlMQXX3zBypUrmTx5MvXq1dOENnLfaUyc/GXp6emEhIRw/Phxzpw5Q1JSElWrVrVcxGZnZ9OuXTsSEhKoV68eLVq0oFy5cna31EBBQQEmk4nc3FyGDBlCUlISqampeHh4ULt2bYYMGcLMmTNZuHAh8fHxbN68mUqVKlGzZk1mzpxp7fLtStFnc+/evXzxxRfs378fZ2dnateuzZgxYxg1ahQhISHExMRw/fp1hgwZQtOmTUvckNCX+N/HXGx8V0xMDN9//73lbykvL4+ePXsyduxYFi1axBtvvMHMmTO5ePEi/fv3tyxmHRERwdq1a3n++edxdHR86N+vwsJCPD098fT0ZPr06ZjNZsusuCNHjsTFxYWcnBy+/fZbsrOzNabzd3h6enL27Fnmzp3L1atX2b17NzVq1GDGjBnUqFGDPn368M0339ChQwdrlyp/UdHSEF988QVz587l5s2buLq6cv36dTp06ECfPn1YsWIFkyZN4ssvvyQ5OZmmTZsyceJE2rVrV2I9OZH7RevEyZ9WvKXhxo0b7N69m5kzZ+Lj48P8+fMJDg7+P59vz60UL7zwAllZWfTr14+mTZuyY8cOxo8fzyOPPMKSJUtwcHAgMzOTxMREKleujJOTk+6I30dFX+InT57k2WefpUmTJlSsWJHMzEz27NmDi4sL7777LvXr1ychIYHhw4eTm5vL8uXLqVu3rrXLtytLlixhyZIlTJo0icjISDIzMxk3bhyHDh3iiy++oGrVqsCv48A+/PBDNm7cSM2aNQFIS0sjKyvL0or6sPntOmX5+fmYTCYyMzOZO3cuJpOJJk2a8OijjwK/3owrau3/8MMP7f5C9Pe+o/Lz85kzZw7r16/Hy8uL6Ohoxo8fj8FgICsriyFDhhAeHk5MTIyVqpb/htOnT9OzZ08GDBhAdHQ0VatWZcWKFcycOZNZs2bRrl07jEYjV69exWw2U7p0acsSQPZ8bSPWoxAnf1jRxUFeXh6ZmZmWbjcuLi7Ex8fz5ptvkpGRwXvvvUdkZCSgE1uRY8eOMXz4cKZMmWKZwnvNmjW89957LFq0iOPHj9OgQQPq1Klj7VLtWlFrqb+/P6NHj8bX15fc3FxOnz7NtGnTuHTpEmvXriUoKIiUlBTmzJnD9OnTrV223TCbzWRmZjJkyBBq1KhhaUk6d+4cTz75JBMmTKBUqVJcuHCBwYMHYzQaSUpKolKlStYu/b4oHuC++uorjh49yt27d6lYsSK9evWydCsFSEhI4Pr16xw5coQNGzawceNGKleubK3SHwjFv6+uXLnChQsXAChfvjzBwcH88ssvODk54e3tDUBGRgaxsbG8+eabLFq0SIs+27gFCxawZ88e5syZY7nJ06NHD3x8fBg2bBixsbEa3iAPFPu+5SZ/WNF6WUlJSYwePZouXbrQokULBg4cyKZNm2jevDlz5syhcuXKDBw4kA0bNgDqWlakoKCAX375hfLly+Pk5MSOHTuYMmUKY8eOpXr16qxatYrt27dbu0y7c/369RL/z87O5vLly4SFheHr6wtgWftpzJgx5Ofn8z//8z/k5ubi7+9vCXD5+fn3vXZ7ZDAYcHNz45dffgF+XQfuxo0bPP/883To0IEePXpw4sQJPv/8c27fvg1gCXD2cL+yKMC9/fbbTJs2jW+//ZazZ8+yYMECunbtyu7duyksLGTr1q307t2bSZMmcejQIebNm2f3AQ5KzkT54osvMmDAAAYMGECnTp2YOXMmfn5+eHt7s2rVKpo1a8YLL7zArFmzGDlypAKcDSosLCzx/1u3bpGSkkKZMmUAGDZsGHfu3GHKlCmYTCZWrlxJXFycNUoV+V0aEyf/T2azGZPJxN27d+nXrx8hISE89dRTODs7Exsby4wZM0hKSmLUqFG89dZbzJ8/n4kTJ5KXl8czzzxj7fKtomhsVWJiIo6OjpQrVw43Nze+/fZbHBwcGD16NC+++CJPP/00AF5eXhQUFFi5avsyZcoUDAYD48ePt3RbvXPnDpcuXbIscFz0njg4ONCoUSNCQkI4e/bsPd1ctSDy/VHU0lShQgUOHDhAbm4ugwYNIjQ0lAkTJmAymShVqhSOjo73/Ky93FDavXs369evZ+rUqbRo0QIPDw/Wrl3LG2+8gclk4ubNm3Ts2JGwsDDy8/MpXbr0Q9u19M8oGtN0+PBhpk2bRu/evWndujWZmZns3LmTZcuWceHCBT744AMCAgJo1aoVbm5uvPjii7Rp08ba5cufVHwM2+7du4mOjrZ0n7948SIbN25k3759LF++nICAAC5fvoyHh4elFVbkQaArD/l/Krr4mT17NqVLl+att96yfOm3bNmSpUuXsnHjRgICAnjmmWcYOnQoVapUsdtB3kWTL5w9e5a+ffsyceJEmjdvTpMmTVi2bBkfffQRzZs3Z/To0RQWFnLx4kXS0tLw8/Ozdul2xc/Pj0aNGuHk5ERycjJBQUEEBQURHR3NwoULqVWrFvXq1bM8Pzc313Invmickfy9fju+q+jfzzzzDK+99hqPPvoobm5ubNu2DVdXVzIyMvjxxx8pV64cXl5e1irbKopuHB07doywsDAiIyPx8PDg+vXrzJo1i759++Lm5kZMTAxvvPGGZdyg/MpgMJCfn8+aNWuIjo5m0KBBuLu7A1CtWjVCQkJ48803+eCDD3j55Zdp3bq1lSuWv6Koy2zRdc2MGTPYtGkTn3zyCV27dmXFihU8++yz3L59m7lz51K7dm0KCgpITEzEzc1NIU4eKOpOKX/YjRs3qFixIuXKlcNsNlNYWEiVKlUYMWIE/v7+bN68GYCgoCD69etnl61LBQUFGAwGS5eliIgIatWqhbu7O6+//jq+vr6WKYrPnj3Ljh07mDx5Mh4eHrzwwgvWLt+uDBkyhIYNG7J7924GDhzIxo0bAXj++efx8PBg6tSpbNu2Dfh1MowDBw6wb98+ateurQB3HxQPcFu3bmXWrFls3bqVS5cu0apVK8vskh4eHnz//fd89dVXvP322+zdu5cJEybg4OBwT3eph1FRV96iWTuzsrL45Zdf8PHxAX79PNeuXZsRI0ZgNBrZt28fJ0+etFq9DzKTyURKSgouLi64u7tbvr98fHzo0qULUVFR7N69m+zsbCtXKn+VwWCwvK/nz58nNTWVyZMnU7FiRct5v2hisbS0NHbs2MGyZct4++23ef755wkLC7OLrtliG3QlIn9IQUEBd+7c4datW8D/ts7l5+fj5+dHt27deO+997h58ya+vr6Wiy97W0qgaH8HDx7MpUuXaNiwIUFBQQAEBASwYcMGxo4dS1xcHMuXL7cMmF++fLkVK7dvHh4eZGVlsWTJEu7evctzzz3HuHHjmDt3LuPHj2f58uVkZWWRlpZG69at6dGjh7VLtgtF55AZM2awdu1aTCYT6enpNG7cmLFjxzJs2DACAgLYsmULw4YNw93dnfDwcObNm0eVKlXsYimTwsJCTCYTZrOZmTNnMnToUGrUqMGyZcvYvXs369atw2Aw8Oabb+Li4oLBYKBMmTIlJjiRkhwdHfnpp5/Izc3FycnJsqaph4cHDRs25NNPPyU3NxcXFxdrlyp/wvHjx1m1ahXvvPMODg4OJCUlMWTIEFJTU4mKirLMMtmwYUNGjx7NJ598wtSpUzGZTFSuXJl+/frRr18/K++FSEkKcfKHGAwGHnnkEWbPns2mTZto164dbm5umEwm8vPzLWvGGQwGu5uRMi4ujho1algWeM3JyaFKlSrEx8eTnZ1NQkICjRo1wmAw4OjoyOzZszlz5gyZmZmUKVMGHx8frc10H/22i17Dhg1ZtWoVMTExLFiwwLL+W2hoKLt27eK7776jWrVq1K1bl+7duwPYRUCwpqJzSNHMiW+++SYNGzbk9OnTvPzyy4wZM4Zx48bRvXt3OnXqRGpqKmaz2fK3VHw9uYdV8TE9Y8aM4fz583Tq1ImOHTvy9ddf89JLL2EymYiNjcXf398yk6KDgwM1atSwcvUPrq5duzJ16lTeeecdBg8ebJnkIjc3l9TUVAICAux+GQZbk5GRwaBBg2jQoIElnKelpVG7dm3+9a9/8a9//YvOnTvj4OCAo6Mj9evXp379+pw/fx6z2Yyvr69loqvffn+IWJPDG2+88Ya1i5AHR2FhoeUkdffuXe7cucONGzfw9vambt26/PDDD6xbt45SpUrh6+uLl5cXhw8fZsGCBURERNC+fXu7CnD5+fmsWrWKxYsXEx0dzfbt27l48SK9e/emYsWKbN26lcTERMLCwihbtqzl5F+6dGkCAgLw8vLSWnD3UfHwdeTIEeLj43FxcSEkJIROnTpx+vRptmzZQnJyMhEREbRo0YInnniCli1bWi58FeD+XkUBLjk5maSkJPLz8+nXrx/e3t4EBwfToUMHNm/eTGxsLB4eHpYu3l5eXpbFu+3hHFS0jzt27ODAgQP079+fRo0aYTQaCQwMJCcnh1OnTnH79m1OnDjB5s2b2bp1K++88w7h4eFWrt76ij5nKSkpnDlzhvPnz5Oens6jjz7KL7/8wueff05iYiK+vr5kZWWxfft2li5dyqBBg0qMlZUHX/fu3QkMDGTGjBmWG6blypWjWrVqGAwGNm/ezJEjR+jSpQsAeXl5GI1GS3grarnWgt7yoNE6cQLcu57bnTt3GDlyJGfOnOH69evUq1ePnj170qJFC2bPns2nn35K2bJlcXBwwGg0Ur58eT755BPA/u5Ubdu2jSVLlnD58mXu3LnDJ598QsOGDQHYt28fY8aMwdfXlzFjxvDII49Yum3I/VX8czl27FgSEhLIzMzkmWeeYdCgQZYv9/fff5/Vq1fTqFEjnn76aVq0aGHNsu1K0XkoO3t0YAAAIABJREFUKyuL4cOHs3//fmrWrMm6deswm83k5eXh5OREXl4e/fv35+jRo/Ts2ZPBgwdb7pTbk4MHDzJp0iQuX77MtGnTeOKJJyyP/fzzz3z//fesXbuW/Px8atSoQfv27WnZsqUVK34wFN2I+eabb5g+fTpXr14lKysLd3d32rRpw/Tp01m+fDmrVq3i8uXLAAQGBtK5c2deeeUVK1cvf8bEiRPZvXs3cXFxljBW/Lvg5s2bfP755yxYsICyZcuydOlS/P39dbNObIJCnAC/dscxGAy8++67AHTp0gVPT09at25NuXLlWLp0KdevX6d58+ZMnDiR+Ph4jh8/Tnp6Og0bNqRVq1a4uLjY7ax9q1ev5p///Ceurq68+OKLDB061PLY+fPnGT16NKmpqYwZM4ZWrVrZ3cx5D5J33nmHL7/8kilTpuDl5UVERARpaWkkJiYSGhqKh4cHmzZtYvr06QQFBbF48WJKly5t7bIfesUvmvbt20diYiI7d+4kISGB999/n8cffxz4tbty0Y2QESNGcPr0abZu3WpXN46KpKSksHz5cjZv3oyLiwtLliwhLCzsnudlZmZaZlq0d8Vbev/xj3/QqVMn2rZti4+PD5s2bWL9+vWEhoby8ccfc+PGDa5cuUJaWhqhoaGEhIRYu3z5Ez744APmz5+Po6Mjs2bNspxDfhvQilpa58+fT1paGrNnz6ZJkybWKlvkD1N3SuHWrVvEx8ezf/9+Dh06hL+/PydOnODVV1+lTZs2hIWF0bFjR1JSUoiNjcXR0ZF//OMfNGnShOjoaMLCwjCZTBQWFtrdnauiQe+HDx8mKCgIf39/tmzZwsWLF2nVqhUAvr6+tG3blp9++olFixYRFhZGtWrVrFy5fUpOTmbBggW89NJLtG/fHj8/P/bs2cPw4cNZuXIlW7dupXbt2rRq1YqqVatSu3Zt6tata+2yH3rFuyn17duXq1evMmjQIAIDA0lJSWHRokWUKlWKunXrYjKZyMnJwWQy0a5dO7p27YqzszOFhYUPfTfK3+6jh4cHjRo1olSpUhw8eJCvv/6asLAwy2RKubm5lnE+D/ux+aOKjsPixYvJzs5m8uTJhIWFUbp0aRo0aIC/vz9fffUVFy9epHPnzlSsWJEqVapYZvsU2/DZZ58xY8YMXnrpJVxcXFiwYAFOTk5ERERgNBrJz8/HaDRiNptxdHQkPDycSpUqcebMGRYuXEinTp3w8vLS34080BTiBFdXVxo3bozZbGbv3r3s3LmTc+fO8dxzz+Hj40NOTg6urq60aNGCkydPsnXrVsti38XZ48mu6MKzXr16REdHU7NmTbKzs9m+fTt79+6lWbNmuLm5WY6ft7c3Xbp0+d3FiOXvl5aWxooVK+jYsSPu7u7MmjXLMkaoT58+nDt3joSEBLp06UKVKlUsrRr2NlnP/VZ0bI8fP05sbCzPP/88oaGhBAUFER4eTnZ2NkuWLOHGjRtER0eXCHL2Mg6uqPWgoKCAEydOEB8fz6VLl/Dx8eGRRx4hODiY48eP89lnn+Hp6UnNmjUtyyzYYyvl7ykegr/88kvS0tJ47rnngF/HNzs7O1OjRg0uXrxIXFwcPXr00EyeNujYsWMMGjSI8ePHM3DgQMLDw8nLy2Pp0qVcvnyZ1q1bW4Kcg4OD5fweHBxMSEgIUVFRNG7c+KE/p4jts79+b3KPolndBg8ejJeXF5s3b+bixYscO3aMypUr4+zsbOnC1KNHD+Li4rh27ZpmVPz/Fb/ADwoKYtiwYfj6+rJhwwZefPFFxo4dy9WrV8nMzKRv3766oLqPfhu+AgMDqVu3LiNGjMBsNuPi4sKAAQN4+eWXMZlMXL58mdOnT9/Toqwv87/fpk2bWLp0Kfn5+SW6rVWvXp1hw4bh5+fHihUrSEpKYvHixZabSPbw3hTv5TBy5Eh+/PFHrl+/Tn5+PkFBQTz//PP06tWL0qVL8+GHHzJnzhzOnj3LhAkT7P58k5GRQXZ2NiaTCW9vb8s5wcvLi3PnznHx4kUqVqyIyWQiLy8PR0dHmjVrxtatW8nJybF2+fInmc1mDh06RP/+/XnqqacACAsLY+jQofj5+bFkyRKSk5NZvHgxrq6ulpsjRZ+LiIgIy2vpBog86NQSZ+eKX+QajUbq1q1LqVKluHLlCjt27MDb25vq1atbxrmdPn2a/fv306lTJ8vUy/au+EVkUTBo0KABHh4eHDlyhJUrVxIXF0evXr2oXLmyFSu1L0VdXeHXFrirV6/i4+NDaGgoAQEB1KhRg6FDh9K9e3eMRiMZGRmsWbMGd3d3OnToYOXq7U9GRgZfffUVycnJlC5dmrCwMMtaXF5eXlSvXp3SpUuzZs0aPD09qV+/vpUrvn+KzjH//Oc/OXDgAOPHj2fQoEF0796d/fv3s337dgwGAx06dCA8PJyUlBTi4uJo166dXY+/jYuL47333mPhwoWsX7+e0qVLW2bmzMvLY/v27aSlpRESEoKPjw8ODg7k5eXx7bffcuHCBbp27aqblTbGYDAQGhpKs2bNSvQW8vT0pFq1avj5+bFjxw7WrVtHs2bNKF26tGXx7t/eELKHG0Ri2zSxiZ0ym80UFBRgMpnIzs7m+vXrllkoARISEli5ciVnz56lRYsW9O/fn2PHjrFs2TIcHBxYtWqVlffgwfB7d+qKbztw4AApKSmUKlVKsxzeR8XfgzfffJPvvvsOBwcHxowZU+J9yM/PJzY2Fjc3N3bs2MGuXbvYvHkz5cuX111YK7h06RJjx47l5MmTvPrqq5ZxKUWysrI4ceKEZfbXh1l2dnaJBaWTk5MZPnw4zz77rOXGA/z6GS6ayXPlypXUrl3bMhmHPY+93bdvH0OHDqVz5874+PiQlJREXFwcS5YsITIyEoDZs2ezcOFCoqOj6dChAxEREXz77bfMnj2bXr16MXz4cCvvhfy35ebmsnPnTubPn09ycjIffPABzZs3t3ZZIn+JQpwdKuoyUqR///6cPHmSW7du0bp1a95++208PDw4c+YMK1euZMOGDbi6ulK5cmUiIiIYPXo0zs7OdjcFb9H+ZmZmcvv2bfz9/f/P/VcAeDBMmjSJPXv20KNHD1xcXOjevTs+Pj6kp6fj6enJ+vXrmTZtGiaTicqVKzNu3DgaNmxod5/t++23xzc7OxuDwWDpuj127Fh27drFgAEDePrpp/H397/nNR7mv7HDhw+zd+9e+vbta2kJKmoZmjp1qmUpgaLZgHNzc2nXrh2NGzdm+vTpdt+CcPnyZTp27EjPnj0ZPnw4Hh4eXLx4kcGDBzNo0CAee+wxy3H96quvmDFjBqmpqeTn5xMYGEhUVBRTp0618l7If1vxnkf79u1j6dKlfPvtt4wfP56+fftatziRv0Bj4uzM+++/T1ZWFkOGDMHHx4eXX36Za9eu8cILL5CWlsbGjRvp2bMnc+bMoWrVqowYMYKyZcvy2WefERgYyMiRI3F2dra7pQSKLjpTU1MZNWoU58+fx8nJiYEDB9KuXbt71qh6WC8ubcm3337L7t27mT59OlFRUQAcPXqU119/nfPnz9OjRw969+5N48aNSU9Pp1y5cpQtW9YuZ1m9n4of3/fff5/jx49z5swZatasSVRUFL179+aDDz5g5syZLFy4kOvXr9OnT597ps5/WP/GUlJS6NWrF/369cPNzc2yvaCgADc3N86fP28JsEXnYLPZjLe3N87OznYf4MxmM3379qV58+a8/PLLlolJKlasSOnSpVm4cCHTp0/H1dWV1q1b88orr1C/fn0uXbrE9evXLbMUysPHYDBYglxkZCReXl64ublpCRmxWRoTZyfMZjP5+fl89NFH7N27l+vXr2M0Gjl48CAjRoyga9euNG7cmPDwcEu3nLCwMGrUqEH9+vUpKCjgySefpHz58pjNZru6yC0+/fk//vEPnJyceOyxx3BxcWH58uUUFBRQqVIlSpUqZeVKpbgjR47www8/MGbMGG7evMmqVauIiYkhJycHPz8/1q9fT4UKFYiMjMTPz8+yjpa9XwT/3YqOb0xMDNu3b6dOnTo0btyY1NRUVq1axa1bt2jRogVNmzalXLlyzJs3j8TERNq2bWsXs7p27dqVJk2aEBMTU2JMj4+PD+fOnWPDhg2EhYWVGF9748YNvv76a8LDw3nkkUesUfYD48UXXyQxMZG5c+daxm0bDAbi4+NZtGgRERERtGrVCpPJxObNm7l16xZPPPEE5cuXJzQ0FG9vbyvvgfyn/t1swsWDXNmyZWnWrBm1a9e+zxWK/HcoxNkJg8GAg4MDTz75JDdu3GDr1q1cvXqVn3/+mT59+uDl5YXRaKRixYrUrFnTMgOch4cHERERRERE4OfnZ3dTrRff36NHj3L8+HHeeecd2rZtS/v27XFwcGDhwoVcu3aNypUra7KXB8jNmzf5+uuvOX78OMuXLycuLo6nnnqKt99+m8cff5w9e/YQFBRkF+OrHgTF/5b27t3L4sWLmTFjBk899RRRUVFkZ2cTHx9P3759SU1NpXTp0tSrV486depQs2ZNuxjf1bt3b+DXBel9fX0tU+LfuXMHFxcX6tevz/nz5/noo4+4e/cuBoOB06dPs2LFCk6ePMl7771XYhydvUlLS+PIkSP8+OOPuLi4EB4ejpubG5cvX6Z379706tWL1157jWbNmtG+fXvS09NZu3Ytbdq0UWuMDSuaxCozMxOj0VhitsnfUzzI/XapJBFbohBnR/Ly8nBwcKBZs2YAbN26lRs3bhASEmK5E2UwGPD396dWrVrk5OSwcOFCqlevbrnra08BDv53fzds2MDq1au5desWzz77LI6OjpjNZho1akTlypVZtmwZp06donz58pQrV86uWiofBL+30LOPjw83b97k1KlTlClThv79+zNkyBDc3NzIzc0lNjaW0NBQhbi/2fnz58nMzCwxQcnJkyf55ptv6NOnD2XLliUxMZGhQ4fSr18/GjVqxFtvvYWfnx9VqlQhODiYKlWqWHEP7o9JkyZx5MgR3n//fUJCQixdJq9evcqUKVOoWrUqFSpUoFatWjg5ObF8+XJiY2PZtWsXDg4OzJw507LIt71ydnamUaNGuLu789FHH5GSkkL16tXp27cvDRs2ZOLEibi7u1su4HNycvj666/p3Lnz7467lAdfUU+ZnJwcXn31Va5fv05oaOj/M5zZ27WMPJzsZ1CTnSssLMTR0ZFbt27h4+PDgAEDCA0NZerUqSxfvhyj0cg//vEPyxiLypUrM3jwYB555BEee+wxK1dvXWlpaezdu5eff/6ZgoICCgsLgf+dWKFDhw6UL1+ekSNHMnz4cDZv3kyFChWsXLX9KD5JxtmzZ0lKSsLV1ZUaNWoQExNjmeXPbDaTmppKVlYWy5YtIyUlhR49eli5+ofbrVu36Nmz5z0B4+7du1y/ft3Sda1fv360bNmSkSNHkpGRwfnz57lw4YK1yr7vtmzZwqeffsonn3xCjRo1yM3NxcnJicLCQnr16kW1atUIDAwEIDg4mDFjxtCrVy9++ukny1hOPz8/K+/Fg6FUqVL07dsXf39/pk2bxpdffklYWBivv/66ZYxh0QV8bm4ufn5+uulmo4qf+8+dO8c333zDzz//THZ2Nt26dbP8TdhbDyKxHwpxdqD4LG6vvfYaBQUFzJ07l+joaPz8/HjttddYtGgRN27c4Nlnn7VcWJUvX57y5csD984mZw+KTvylSpXi1VdfZd26dSxfvpzevXvz7rvvWloHCgsLqVevHitXriQ+Pl4B7j4qPj5z2rRp7Nq1ixs3bpCTk0N4eDjNmzdnyJAhwK8tHdu3b6dUqVLk5+fz4Ycf4uvra5ef7fslOTmZjIyMeyYhiYqKomLFiowfP57U1FTKlSvH5MmTKSgo4Nq1a5QtW/aeyYIeVjk5OZw5cwZHR0fmz5/P/PnzLV0iX3jhBXx9fZk6daplgg749ZxT/PwsJS/UnZ2d6dChA+XKlWPmzJlcvXqVM2fOULp0actn8cKFC8yZM4emTZtSq1Yta5Yuf0Hxc/+4ceP45Zdf8PDwIC8vj9mzZ3P9+nWeeeYZqlSpUqL7pMjDRN0pH3LFA9yOHTs4ePAg3333HceOHaNBgwaEhoby+OOPc/jwYXbu3Mnt27cJCgrCx8enxOs8rDPB/Z6i/vVFJ/yMjAx8fX1p1KgRPj4+JCQksG3bNoKDgwkODsZgMJCfn4+Pjw916tSxcvX2peg9WrlyJStWrODVV19l6NChvPTSS6xdu5Zjx47x6KOP4ufnh6OjIwEBAURGRtK/f3/q1KmjmSj/ZhkZGXz66adERUURHh5uuZByd3cnLS2NL7/8krS0NF555RXq1q3L5cuXWbVqFWfPnuW1116zi/EqJpOJ8PBw/P39iY2NZcOGDTz55JPMnj2b/fv3M3fuXCpWrFjiZ3QxWlJ+fr5loe4rV65gNBpxdXWlQoUK1KtXj2PHjrFixQrLIvJpaWmMHTsWgI8++sjK1ctfUfQ3sHbtWtasWUNMTIxlCQmDwcD69eu5dOkSAQEB+Pv7K8jJQ0kh7iFXdMKaPHkyn332GQBlypTh1KlT/Otf/6JKlSpUrVqVzp07k5yczIYNG0hJSaF58+Y4OTlZs3SrKGqVKSwsZM6cOcyfP58lS5Zw5swZPD09adeuHYGBgRw7dozPP/8cT09PatasaVch90GTmZnJ3LlziYqK4vnnn8ff35/s7Gw+/PBDXnnlFfLz8zlx4gRt2rShUaNG1K5dm7JlywK6GP675eXlsW7dOurUqUO9evVKXEg1btwYk8nE2bNn2bdvH59//jlffvklJ0+etASXohsqDztXV1eqVatGhQoVOHjwILNmzeLw4cOsWbOmRPiVexUUFFiGAbz66qssWLCAxYsXAxAUFETFihWJiori2rVrLFq0iIKCArZu3cqxY8f47LPP7GLG04fZxo0bcXJyYtCgQXh7e1vOLQEBASxYsIDk5GS8vLyoVKmS/obkoaMQ95Ap+rIvPtHDjh07WLhwIW+88QbDhg3j2WefJSgoiEuXLrFu3Tp8fX2pVq0a0dHR5Obm0qRJE2rWrGnlPbn/ii8l0L9/f44fP07ZsmWpVasWX3zxBfHx8ZhMJrp27UrlypW5cuUKK1aswGw207hxYytXb1+Kf74LCwtZtGgRYWFhREdHk52dTYcOHYiMjGT8+PF8+umnrFu3js6dO9v1zH3WYDQa+fzzzy1LORRNrpSbm4uDgwMNGjSgatWqVKxYkZycHDp27Ei/fv2oVauWXbWSms1mTCYTYWFhhISEkJmZSVJSElWqVKFOnTqW1n57CLR/RvFz9uDBgzlx4gStW7cmKCiIlStXkpOTQ1BQEBUqVKBp06Y4OjqyYMECzpw5w/r16ylXrpyV90D+U1988QWJiYmWxbrz8vIwGAyEh4eTlpbGtm3bSEpKwt/fn0qVKv3uJFgitkpj4h4yt27dwtfXt8SXfVJSEr6+vlSvXt0yrqJDhw6ULl2aqVOnMmHCBK5du8bAgQN55ZVXrFW61RQF3+Jd85KSkpg9eza1atXCaDQyePBgRo0axezZs/H39+fxxx/H1dUVDw8PmjdvbuU9sC9FraXZ2dkkJiZSvXp1qlSpwvHjx7lx4wYvvfQS5cuXZ/LkycCvEx24urraRdc8azt+/Di3b98mKioKo9GIo6Mj3t7eZGRkWCZXAkq08kdGRhIZGXnPa9lTYCneQhkVFYWnpycuLi5MmzaNpKQkJk6ciMlkIj8/39LqJP/bkr58+XJOnTrFvHnzqFWrFj/88ANHjx7lk08+4cKFCwwfPpzatWszdOhQypUrR0hICKGhoVauXv6s32uR7tChA9988w2rV6+mR48eODk5YTabgV9buJs2bcrly5d57733aNKkiW7kyUNFLXEPkTt37tC3b1/Kly9PpUqVLNu///57EhIS6N+/P05OTpa74BUqVMDJyYmdO3dy+vRpSzdKe1J8zGDRF8SXX36Js7Mzffv2xWg0kpubS6lSpejWrRvbt2/nwIEDdO/eHX9/f5o3b26ZNU7+fsXvvPfo0YPk5GTatGnD7du3Wbt2LXFxcWRkZLBo0SLKlSvH3bt32bRpE0ajkc6dO9tNy4415OfnM2XKFObOnUtgYCCVK1fGycmJ7du3s23bNvbs2cP27dvZuXMne/fu5ejRoxw8eJAbN26wd+9eqlevbtcBpXiQK1euHNWrV8fR0ZHVq1dz7NgxOnTooM8vJVvhzWazpctu3bp16datG3l5eWzevBmj0cgLL7zAxo0bOXXqFC4uLnh7e9O4cWNNCGODinetTk9Pt6x96+3tTUJCArt378bNzY0KFSrg6urKzZs32bp1K9WqVWPcuHF8+OGHVKtWzS6WKxH7YT+3Ou3AzZs3admyJS1atKCgoID09HQAHn/8cbKzs3nzzTeBX++C5+fnA7/O4lWjRg1atWrF7t27+eGHH6xW//127Ngx2rZtS1paWok7fKmpqVy7ds2ylEBR8AV46qmnOHfuHFevXgVQ6859VvQeHT9+HBcXF5566ikAnnvuOV566SWSk5MJCgri7Nmz7N+/n5kzZ7Jjxw7Gjh1rmbJd/h4mk4mJEyfSrVs3Jk6cyPz588nKyqJs2bIYjUbCwsK4du0aJ06cYNu2bXzyySfMmzePUaNGsWHDBv0tUXKMZlBQEAMGDGDMmDEcPnyYZs2akZmZaWllsEcXLlygT58+XLlyBfj1eDk5OZGamsqJEycAOHXqFB9//DHR0dF0796ddu3acejQIWJiYoiPj7dm+fIXFZ+J8t1336V///60bt2aJUuWWNYEDAwM5IMPPmDQoEGMHz+el19+mV27dtG2bVvLpFaenp5W3hOR/y77ve35EAoJCbF0h4yJiSElJYVJkyYRGhrKoEGD+Oijj8jKymL69OmWO1WHDh2iYsWKvPLKK7Rp04aEhATq1atn5T25P1JSUujSpQulSpUiJyfHchEZFhZGXFwc3333HS1atKCwsNDS/ctgMBAYGGhXXb0eNMuXL2fZsmW4uLgQEBBg2T58+HAqVqzIu+++y+uvv87t27epWbMmM2fOpHr16iVaXeXvUaFCBcaNG4e/vz8LFizgl19+wdHRkfr16zNlyhSMRiMmk4krV65gMBi4du0aOTk5llld7eE9+r/28fe6ivn6+vL000/j5uZGamoq7u7u96vMB9KFCxc4deoUffr04b333qNu3boUFBTQokULbt26BcDUqVNp2bIlXbp0AX69CffEE0/QokULOnXqZM3y5S8q+rt49913+fTTT2nRogVlypRh5syZnDp1igkTJrB69WpWrlzJ0aNH+fHHH6lbty5jx44lPDyc5cuXk52drTGQ8tBRiHvIFHXJ8fT05MiRI0ydOpVRo0bRv39/cnNzWbduHdHR0dSqVYubN2/y888/s3jxYsqWLUt4eLhljTh70KZNG9q0aUNeXh5dunShW7duDBw4kGHDhvH9998zduxYpk2bRlRUFO7u7ly6dIm4uDgCAgLUhdKKgoKC8PT05Ny5c+zdu5dOnTpZxnp26dKFqKgorl+/TmFhIQEBAZblMh72cPCg8PLyYuDAgQQEBPDuu++SlpZG5cqVyfv/2LvPsKiu9fH735mhjnSR3qs0QQQBjYJA7AUNscRurFGjxpYYe4slGltiVwQLWFCMsWLDFhsW1AiIDRVBsQHSZpjnRf4zB2NynnPOLzqG2Z9XXsJw3XvPzNrrXuVelZWqJMTa2hqRSPRGEq4J5/Up97QVFRVx8OBBFAoFgYGBODo6IpFI3krkFArFGzPOmi4kJIQVK1awYMECevXqxdy5c2nZsiW9e/emuLiYoqIinj17pipy8ezZM3Jzc3FxcRESuH8o5aBHYWEhv/32GzNnziQyMhJtbW22b9/O9OnTefz4Md9++y29evUCfv+eFRUVcfXqVdatW8fSpUsZP348Li4uar4ageDvJeyJq0GqF+gIDw9HIpFw5swZ0tLSsLGxoWvXrjRu3BgtLS2eP39OvXr1+Prrr6lfvz7JycmkpKTQt29frKys1H0p71z1ztKDBw+4evUqSUlJFBYWEhERQcOGDbl58yYrV67k0qVL7Nq1i71793Lnzh3Wrl2LVCpV8xVoDuUemMLCQl6+fKkqV5+VlcWuXbtwcHDAxcVF1QmuVasW5ubm1KlT540DkgXvj5aWFh4eHvj6+nLnzh1u375N/fr1cXBweKOIUHU1PcmuviSsa9eu7NmzhwMHDnDy5EmkUikODg5vLSkVzrb6F5lMho6ODjY2NlhYWHD9+nU2b96Mvr4+gYGB6Orq8vz5czZs2IChoSEymYzdu3dz4sQJZs6ciZGRkbovQfAfUn7mlXugb9++zcqVK8nMzCQmJgYbGxsUCgU+Pj7Ur1+fbdu2cfToUezs7KhTpw7a2tocO3aM2bNn8/DhQ3r06EHPnj3VfVkCwd9OSOJqAGUnt7KykrKyMp49e4aBgQH16tXD3t6eixcvcujQIaqqqoiOjqZx48a0adOGevXqkZGRQXJyMmvWrGHw4MG0bt1a3Zfzzv2xU2RiYkJAQAA6Ojps3LiRjIwMunbtStu2bbG2tiY7OxuJREKDBg0YP348dnZ2aoxesyhHYcvKyhg4cCDZ2dk4Ojri6elJo0aNuHfvHmvXrkVXV5e6detq5NmGHyqJRIKtrS1+fn5kZmaSlJSEhYWFKuHWNMo254cffiA3N5cZM2YwcOBATp06RXJyMnp6ejg4OKhmK/9YNVfTKZP8BQsWkJycjEwmQy6Xc+TIER4+fEh4eDjGxsaUl5ezdu1ajh49Sm5uLnPnztXII3P+qZRtfZMmTZBKpSgUChISEkhNTeXJkyfExsZiYWGBQqFAoVBgb29PZGQkaWlprF27luDgYJycnKhTpw6tWrUiNjbfh/m9AAAgAElEQVSWRo0aqfuyBIJ3Qkji/uFkMhkSiYTbt28zf/58FixYwO7duzl27Bj29vaEhYXh5+fHjRs3OHbsGI8ePcLf3x89PT1ycnKYMmUKcrmcTp060a9fP3VfznsjEol49OgRV69eRSQSYW9vj5eXF+bm5qpDhyMjIwkNDaVly5bExsYSFhaGmZmZukPXGNWX12VkZKiqzD148ABra2s8PDxo2rQppaWlrFy5kqKiIry8vDR+39CHRCwWU6dOHUJCQnj48CFr1qyhoqKCoKAgjUnk/jholJOTg6OjIzExMZiZmfHpp5+SnZ1NQkICFRUVODg4YGpqKiRvf2Lv3r0sWLCA0aNHM2rUKDp06ICbmxubN2/m+PHjhIeHExERQZMmTWjRogXdu3fH399f3WEL/gvp6encunWLNm3aIJFIEIlE+Pr6oqWlRUZGBvv378fHxwc7OztEIhFyuRwzMzOaN29OWVkZXbp0QSKRqKqRCqsxBDWZkMT9gymX5xQXF/Ppp5+ir69PSEgIoaGhXL58mXXr1uHv709gYCChoaHk5uaSnJyMlpYWISEhWFhYEBMTQ0xMDEFBQeq+nPdCWab48OHDTJgwgfj4eK5fv05wcDBWVlZ4eHjg5OTEyZMniY+Px9fXF2dnZ+DPCw8I3p3qh/ju27ePOnXq4O3tzaFDh8jOzsbU1BQPDw8++ugjTE1N+fHHH7l37x4tWrSo8UvzPgR/9n34q++IkZERDRs25OXLl9y5c4eOHTu+rzDVTnlPTp8+TXJyMhs3bsTCwoLw8HDV77Ro0YKqqirWrVvH48ePsbS0FMrgV6OsyLlt2zZ0dHQYMWIEBgYGGBsb4+npiY+PD0lJSaSmphIQEICvry92dnaq/bCCfw4LCwuioqLQ1dVl3Lhx5OXlERwcTP369alduzZZWVns3r0bExMT6tati1gsRiaTIZVKadq0KRKJBJlMJjwDBBpBSOL+wZSdpYULF/Lq1Su+++47WrduTUBAAKdOnUKhUBATE8Pdu3dVnV0dHR169+6tKreup6enUSPiYrGY7OxsBgwYQMuWLRk2bBixsbE4ODhQUFBAaWkpgYGBeHt7k52dzfLly2nYsCG2trZCAqcGO3fuZPv27cybN4/+/fvTunVroqKi2Lt3L0eOHMHAwAAnJyeCgoJwc3MjJCQEBwcHdYdd41U/s+nu3buUlJQAqA7S/bNkTl9fn7CwMGJiYhCJRG+c91VTKducM2fO0L9/fx49esTTp0959OgRTk5O2NraqtrfkJAQnJycWLlyJd7e3qqKnQJUy0qPHz/OjRs3VIc6w+/Ldu3t7Xny5AlHjhwhKSmJjz76SCP2dtckly5dorS0FHNzc7S0tHjx4gU///wzGzdupLKykrCwMOrWrYuVlRW5ubns2rWLyspKgoOD30rYhAROoCmEJO4fTNlR2rp1K1paWnTv3h2RSMRPP/3Ezp07WbRoEbm5ucyePZtGjRphYWFBUFAQOjo6GlEJrjrl3qqqqipmz56NjY0Nc+bMURVamDBhAgsXLiQlJYXy8nLatWuHk5MTlpaWdOrUSd3ha6xTp05x7949hg4dip6eHhUVFVhaWtKmTRt2795NWloaIpEIFxcX/Pz8sLOzE2ZM37HqB65PmjSJZcuWkZiYyM2bN7GyslJVnvyzJE1LS0v1fzX9PVIudS8vL2fq1Kk0b96ccePGERMTQ1paGgcOHMDCwgI7OztVQuLh4UG7du1o2rSpmqNXr/Ly8jcOflcOGuTn57Nr1y7c3d3x9PRUfcZEIhG5ubnIZDJiY2OFSpT/MKWlpWzcuJFjx44RHBzMqVOnkEgktGjRAj09PVavXk12djaRkZG4urri4uLC8+fPSUlJ4fr167Ro0aLGtycCwZ8Rhiv+wZRLTEQiEUVFRYhEIvbt28eSJUuYNm0agYGByGQyHj169FbCpikJ3PDhw7lz546q0ykWixGJRJiZmVFcXExycjLdunXj5MmTNG/eHGdnZ7Zu3Up+fj6BgYEMHz5czVeg2cRiMXfv3uXVq1fAvw5eNzY2pk+fPjx58oSVK1eSlJREZWWlkMC9B9XPbDp06BCffPIJ0dHRZGRkMHv2bPbt2wegGjTRVFpaWlRVVXH48GHMzMxo2LAhrq6u+Pr6snHjRtzd3ZkyZQqJiYk8ffpU9TpHR0c1Rq1+c+fOZdasWezZs4fXr19TUVGhel517tyZ5s2bM2nSJHbs2EFpaSkARUVF3LhxA4lEQt++fdUZvuB/oK+vj7W1NcePH1cd83Pr1i1sbGzo2bMn3377LadOnaJHjx48evQIPz8/RowYQWRkJO7u7kKbL9BYwkzcP5iy4aqsrCQhIYHnz5+zaNEihg8fTo8ePQDIzMwkMzOTVq1aaVxRjhs3bnDkyBEiIiLe2Btx5swZdu/eTWpqKnv27MHMzIwVK1bQqVMnHBwcSE5OJjo6GgsLC+Hh8J4pkzDlCLu+vj6HDx+moKAAT09PjIyMVB267OxsdHV18ff3Z/369TRp0uSNc8cEfz/l+1JaWsqmTZv4/PPP6dmzJ82aNcPS0pILFy5w6tQp5HI5AQEBqsIDmrq86eDBg4waNYrs7Gx8fX0JCAgAfu+0tmnThsePH7N+/XpevHiBu7s7xsbGao5Yvc6cOcPUqVOpqqpiz549pKSkcPXqVaRSKWKxGGNjYwIDAyksLGT58uWcP3+e48ePs3fvXlJTU5k+fbqwnPofKiAgAG1tbXbv3o2FhQX169fHx8cHqVSKi4sL7u7uHDlyhK1btxIQEICnpyeNGzdWVZ7UhOXZAsEfCUncP4hys+6LFy/Iy8ujqKgIqVSKt7c3BQUFpKSkYGNjw7x58ygtLeXGjRvMnDmT4OBgYmNj1R3+e1enTh2ioqKwsrJi9erVpKWlERYWRkREBCUlJchkMj777DNGjx6NnZ0dMpmMEydOcP36dWJjYzXq4HN1U3b0lQ/h58+fo6+vj7m5OWVlZaxbt46SkhIMDQ2xtbXl3r17bN68GZlMxqxZs9i+fbvqwS94d0QiEZcuXWLOnDk8efKEyMhIVafZ3d0dDw8Prl27xsmTJ8nPz6d+/fpvnX2mSdzc3PD19eXMmTNkZWXh7OyMlZUVEokEsVhMs2bNkEgkrFu3jm7duml8IQ5jY2NOnz6Njo4Oa9asobCwkJs3b7Ju3TpOnjxJeno61tbWeHl50aFDB65cuUJlZSXW1tYMHz5cKCX/D1VZWYlEIuHYsWNUVVUhl8u5ePEi5eXleHp6YmBggIODA97e3mRmZrJkyRJ8fHxwd3dX/Q0hgRNoIpFCuSZP8EFT7mF79uwZw4cPJycnBwMDA5o0aaJa8rd69Wo2btyIjY0NWlpaKBQKLC0tWb9+PfCvfWGaoPqyuqKiIqZNm8aePXvo2LEj3333HQAVFRXo6OiQk5NDSUkJt27dYv78+XTv3p1hw4apM3yNUv1zuWTJEs6fP09WVhaenp60bduW2NhYtmzZwvLly1EoFKpiPK9evSIxMRErKyu6dOnCZ599RpcuXdR8NTXf8ePHmTBhAoWFhUyePJlPP/0UbW1t1c8fPHjADz/8wIEDBxg7diy9e/dWY7TqpWyHrl27xldffUVFRQXjxo2jadOmGBgYqH4vLy9P42eRle3AuXPn+Oqrrxg9ejQdO3bk6dOnPH36lO+++46zZ88ikUgwNTWldevW+Pj44OHhgZeXl7rDF/wP/qpPcv/+fSZOnMjdu3dp06YNvXv3VhWquXbtGomJiQwfPhxLS8v3HbJA8EERZuL+IZQNXY8ePRCJRHTp0gWZTMapU6fIyMigQYMGtGvXjsjISCoqKnB3d6d169aMGDFCVXJXU/bBwZtJnK6uLoGBgUilUjZv3szJkydp3bq16qy8Tz/9lM2bN5OZmUnLli0ZPXq0mqPXLMr3acaMGezZswd/f3/atGnD/fv3iY+P59GjR4waNYrQ0FAcHByQSqU0a9aMr776CmdnZ9auXcuJEyf48ssvNX4m41344z5DOzs7QkJCyMrK4sCBAzg6OuLk5IRYLEahUGBsbEzjxo2pVasWffr00ZgRcuVscnZ2NocOHSI+Pp709HQAgoKCaNGiBWfOnCExMRETExPs7e1VZ1gZGhqqM/QPgvJzUqtWLS5evMjZs2dp164dxsbGmJqasnjxYlq3bk1sbCwKhYK0tDRSUlIQiURERESoN3jBf616gaRDhw6xa9cuTpw4gbu7O7a2trRs2ZLMzEwOHjzI06dPMTExIScnh8rKSrp06YKJiYlGL9UWCECYifvgVe9A5eTkMHHiRKZPn65aRrBy5Uq2bdtG7dq1GTp06J9WNdOkGTj4fdmplpYWpaWlXL16lcLCQuzs7KhXrx6bNm1i+fLlquU6Li4u3L59m6ysLFxdXd9YniF4fy5dusSQIUOYOnUqERER6OnpceHCBXr06MHSpUsxMjLCz88PqVQKQEFBAXv27CE7O5vU1FQmTpxIhw4d1HwVNU/1tiM3NxcdHR1MTU3R0dEhNzeXWbNmcerUKb766iu6dOmCVCp9K+nThEq4ymu8c+cOPXv2xNDQkKqqKl6/fs2TJ0/o168f48aNQ6FQ8PXXX/PLL7/QsWNHRo0apXF7lf8T2dnZdO7cmZ49e/LVV1/RtWtXSktLWbZsGfb29pSWllJWVsbRo0eF6sH/UMp2Ii4ujjVr1qCvr4+Pjw/Dhg3Dzc1N9Xs//PADiYmJSKVS8vLymD17tvCeCwT/jzAT94FTdoZ++uknfvnlF4qKiujcubNqBDcoKAgjIyPOnTvHr7/+ilwuf+t8IU0ZCYffO53KDuOAAQNISUlh9+7d3Lhxgw4dOuDj44OdnR1XrlxhzZo11K1bl/r16+Pm5kbt2rXVHL3munXrFkeOHGHgwIHUqVOH3NxcevbsSWxsLFFRUYwbNw5LS0vc3d1RKBRcu3aNnTt3olAo+Pzzz2nfvr26L6HGqT5S/uOPP7Jo0SKWL1/OgwcPsLCwwMPDg6ZNm1JcXMzy5cspLS3F29tblWgr1eQBJGVHVCwWU1lZyZAhQ6hXrx5TpkxR7dEyNjZmzZo1PH78mKioKD7++GPKysq4fPmyRi81/StVVVWYmJjw/PlzTp8+zc8//8yDBw/4/vvvVZUItbW10dfXF5ZR/oOJRCIuXLjAhAkTGD16NF9//TXt2rXD0NCQY8eOceHCBVXxEgsLCywtLenYsSOffPKJukMXCD4YWv//vyJQN5lMRlVVFSdOnFCVXDcxMVF1jmJiYrC0tGTJkiUsXryYevXqaexBscp7MnbsWAoLC5kzZw66urqYmZmpzmKKiorC3t6eBQsWMHjwYL799lt69uypzrA1jnK2VHkelEgkorCwkJcvXwLQv39/goKCGD9+PLq6urx69UpVZVUkEhEaGoqfnx+1atVS85XUfOvWrWPVqlW0bduWyMhI1q5dS0FBAb1796ZJkyZ888032NnZMX/+fG7evMny5cs15n0RiUSqRC47O5vCwkKGDh2Kvb09AHXr1sXCwgL4/T62aNGCJk2aqPbHCd4mFosRi8V069aNX375hczMTBISEjT2mVZTKfs0TZo0oUuXLpSVlXHp0iWmTZvG3bt3kclk7Nq1i82bN9OhQ4c3ZvQ1bXWRQPBXhJm4fwCxWIyfnx/u7u6cPn2aixcv4ufnp+ocANjb2+Pr64u7uzuRkZFqjFb9cnNziY+Pp1evXkRFRWFpaanac3Lr1i3mzZtH27ZtCQoK4sWLF4SFhWn82Uzvk0KhQCKRUFlZyciRI3FwcMDR0ZGDBw+q3ruqqip++uknpFIpDx48YP/+/YSEhODv768qJa2tra1Rs8zvm0gk4urVq0ydOpUJEybwxRdfEBoayuXLl/n111+5cuUKxsbGODs7ExwcjI2NDbVr1yYsLEzdob9zCQkJ5OTk4OzsrCrqcufOHeLj4xkyZAjm5uaqinv6+vp4enoSHx+Pra0tDRs2BDTnrM7/RVVVFWZmZujp6XH16lX8/PyEWbcaRiQSce7cOQ4fPkxISAhxcXHMmzcPCwsLVVGbuLg4nJ2dcXd3fyNpE9p9geB3wlDGP4S+vj7R0dH88MMPlJWVMXToUI4fP45cLgd+7xjXrVuXTz/9FECjD9kFuHfvHnK5/I3GXrk87OjRoxw8eBA3NzemTZtGkyZN1BipZql+ls/333/P7du3kclkWFpaMnz4cFJTU7l8+TKff/45JiYmFBQUsHHjRoqLi2nVqhXwr9lW4UH+91K2JdXduHEDT09PIiMjEYvFHDlyBLFYzLx585BIJMyfP5+EhARu3rxJx44dGTRoEFBz2x+FQkFpaSkrVqxQHW3x/PlzAFxdXVVLJ8vLy9HW1kYmkwGgp6eHg4PDGxUpBX9N+R0PCwvD2NiYPXv2UFlZqeaoBH+3kJAQTE1N+eSTT9izZw+RkZEkJibSrl07HBwcsLS0fGt5tkAg+BdhJu4D88eiANWJxWJsbW0JCQkhIyODuLg4zM3NcXV1faPEN2h2B7esrIwjR45QXl5OcHCwav+gSCRCR0eHX375BTs7Oxo2bIiOjo5G36v3TXmvN23aRF5eHhEREarkzMvLC29vb9LS0sjIyGDr1q3s27ePy5cvs3DhQtzd3YVqZO9Qr169cHNze6Ns94kTJ/jll18YOHAgWlpaDB8+HBcXF4YMGYKDgwOJiYlkZGRQUFBAy5YtVa+rqd8p5Qxwnz59uHDhAps3b0YikWBlZYWNjQ35+fkcOXKEkpIS3NzcMDQ0pLKykuvXr7Njxw5atWolFE/6L5iammJtbc3KlSvR0dEhODhY3SEJ/gfV+zVVVVUUFBSgra2Nk5MTISEhBAYG0qdPH7p164ZEIuH169ccPHiQX3/9lb59+wpVhwWCvyAkcR+I7du3c/nyZdW6f+W+tz9jZmZGWFgYL1++ZNmyZWhpaREcHFxjO07/TvWZHSUDAwNKS0vZsGED2trauLm5qUbznj59SmpqKv7+/vj7+2vkPVO3p0+fMmHCBNWZT40bN1Yl2s7OznTo0AGFQkHt2rVp2rQpn3/+OfXr13+jaI3g7/Xw4UPVnkN9fX0qKiqQSCTo6upSVlZGs2bNWLp0KdeuXSMhIQGAy5cvc//+fQYPHsygQYPQ1dX9t4NQNUVZWRk6Ojq0b9+eu3fvsmvXLl6/fo2bmxsff/wxmZmZ7Nu3j9OnT1NSUsLPP/9MXFwcvr6+jBgxQt3h/+PUrl2bw4cPU1ZWRtu2bdUdjuB/JBKJ2LlzJ4sXL2b27NmkpKRw8eJFmjVrRoMGDahTpw6JiYls2bKFs2fPEhcXx9dff02jRo3+9DkvEAiEJO6DcfPmTSZPnoxMJuPAgQOcOHGCxo0bq4px/JGBgQENGjRAW1ub0NBQbG1t33PE6ld9o3NGRgaXLl0iKysLR0dHgoOD0dLSYvny5WRmZlJcXMzVq1dZt24djx8/Zs6cOUJCoCZSqZRGjRrx6tUrjhw5oiotra2tjUKhwMDAgODgYJo2bYqvr69q76fwEH93jIyMiIiIoFatWsyaNYsTJ07QoEED7O3t8fPzw8TEhDVr1uDn50ezZs2oqKjg6NGjPH/+nJEjR6pK6tf0WVK5XI6Ojg7Pnj1j/PjxlJSUcO/ePTIyMsjKysLb25sePXqgpaXFnTt3SElJQSaTERYWxpw5c9Qd/j+Snp4evr6+fPbZZ2+tOBF8+JQDO6mpqUyYMAEvLy86deqEjo4O6enpbNiwAX9/f6ytrVm9ejWXLl1CIpHQs2dP1fYQoe0XCP6ccE7cB+Lp06fs3buXuXPnIpfLWbRo0RvLk/6KsuOkCSPg1VXvMI4cOZLffvuNBw8eoK+vj4GBAQMGDCAyMpIrV64wb948Xr9+jZaWFr6+vkycOBE7Ozs1X4EgPz+fFStWkJiYSL9+/ejfv7+wbEbNKioqWLx4MevWraN169aqox0Ahg4dSl5eHgMGDODZs2csXLiQqVOn0q5dOzVH/f716NEDkUjEF198gZ2dHb/99hszZszA3NycESNGqA6fLiwsxMjISFWBVSDQRM+fP6dv376Eh4czaNAg1XmSaWlprFixgoKCAuLj41UHeEskElWFW00YHBII/lfCTNwHQiqVIpfL2b59O/r6+mRnZxMaGvr/26lVdgw0rYNQvTjGkSNHmDBhAkOGDKFXr17s3LmT69ev0759ewIDA+nWrRsRERH06tWLtm3bYm5uruboBfCv2WSpVMrq1avJzc2lXr16qkqigvej+gCQRCIhNDQUOzs7EhISuHjxIj4+PpibmyOVSklMTOTQoUNcvXqVnj17auTRHJmZmWzZsoVhw4YRHh6OsbExrq6uNG/enB07dnDgwAGkUilWVlaYm5sjkUg0rn0WCKorKSkhISGB9u3b4+3tjUwmQyKR4OTkhImJCcnJydjY2BAUFISurq5qBVL1syoFAsHbhG+HmlWfCDU1NWXZsmUsXLgQuVxOjx49OH36tBqj+7AVFBRw9uxZ+vTpQ3h4OG5ubty/f5+CggIGDBjA2bNn2bVrFzo6Onh6emJlZSVUh/vAGBoa0q9fP2bPns358+fp378/WVlZ6g5LY8hkMkQiEUVFRdy4cYPLly9TXl5Ox44dWbJkCQUFBQwbNozTp0/TrFkzdu7cyfLly1m5ciXDhg0Dam4lyr8il8vJy8tT7bOVy+XIZDLs7OxYu3YtFRUVrFq1ivj4eMrKytQcrUCgfsXFxeTl5fHixQsAtLS0VNVGo6OjcXFxIT09/a3XCYMfAsG/JyRxalS9BP7Tp0/R0tIiOjqaiIgIZs+ejaenJwMHDmTbtm3A7xvq79+/r86Q1a560iuVSnn8+DElJSXo6Ojw6NEjvvjiC7p27Ur79u3JyMhg1apVFBcXa1xHU93+m1XaOjo6tGnThu+///6/ep3g/0Yul6OlpUVxcTFffvkl/fv3Z/z48Vy9ehWA0NBQli1bhrm5OcOGDWPHjh04OzvTuHFjfH19Ac1c6uTg4ICnpycJCQnk5eUhkUjQ0tICfp/JNDY2xsTEBCsrK/T09NQcrUCgPsr23NramtDQULZs2cKFCxcA0NbWRi6X8/z5cwwMDDRyX79A8H8lLKdUk+qV9iZPnszSpUtZvXo1rq6uuLq6YmlpSb169SgqKmL58uU8fPiQY8eOkZKSQuvWrVWdBk2iXIKhVFRUxP79+9HW1iY6OprY2Fh8fHyYPn06urq6HD16lKKiIrp166ZxHc0PgUgk4vjx46SkpLBt2zZKS0txcnL608+uWCzG3t6emJgYrK2thWpk74HyO9G9e3eqqqoYMWIEUVFRhIWFUVpaypMnT3B2diY6Oprc3FxWrFiBgYEBAQEBqr9R09+j6kdaKP+to6NDYWEhP//8MzKZDCsrK8zMzADIy8vjzJkzzJ07l/DwcHWGLhColVwuVw1O6+jo4ODgQEJCAllZWejr6+Ph4UFRURHHjx9n27ZtDBgwAEdHR3WHLRD8o2heJvCBUHYMJk6cyKlTp+jatSv29vY0aNCAyspKFAoF9vb2TJ06FTs7O5KSkpDL5SxevFhjSnkr7d+/nzNnzvDo0SPc3Nzw8fHh448/pnbt2vTt25cJEyZw6tQpbG1tWbBgAfr6+jx79oyHDx9iZ2cnJHDvmXJ2Zv/+/UyaNAkPDw8KCwtxdnZ+a2bij59j5XJX4T17P5KSkigpKWHNmjXY2NgAsHHjRjZt2kRhYSHh4eHMnz9f1Q41atRIzRG/PzKZDC0tLV69esWGDRu4desW9evX57PPPmPYsGGUlJQQFxfH5cuXCQkJQaFQcOLECUQiEa6uruoOXyB475Rt//Hjx9m8eTMXLlxAV1eXgIAARowYwY4dO+jfvz+TJk1i/vz5aGtrU1RURP/+/WnSpIm6wxcI/nGE6pRqdPr0acaNG8eCBQsICQkB4Pz58yxevJji4mK8vb0ZM2YMJiYmvHjxgqKiIhwdHTVqCdM333xDeno6VVVVWFpacu3aNeRyOXXr1mXu3Lm4uLgwf/58kpOTcXR0ZOrUqdy6dYtTp06RmprKnj173ji8WPDuVP9c5uXl0a1bN7p160b37t3f2ItYVFREfn4+bm5u6gpV8P9s3bqVuLg4Nm7cyP3794mLi2P//v2Eh4dTu3ZtDh06xMSJE+nQoYOqalz1oz1qquqDC+3bt+fFixfo6uqSl5dHw4YNmTFjBra2tvzyyy8kJyeTlZWFmZkZ3t7eTJkyRVhGKdA4yvb/+vXr9OjRg4iICPz9/cnPz+f48eO8ePGCmTNnEhoaytatW7lz5w61a9dWHVtS/W8IBIL/jDATp0ZVVVWYmJhgZ2dHbm4u27dvZ/Xq1Xh6emJiYsKxY8do2LAhMTExmJmZqZbsaEojN3fuXI4dO8aUKVNo2rQpUqmUFy9esHLlSpKTk4mNjWX16tWMGjUKOzs74uLi6N69O9ra2nh6erJq1SohgXuPqn8uHzx4QFVVFYGBgaoErqqqCplMxokTJ4iPj2fx4sXC+6NmBgYGvHr1imHDhpGTk4NIJGLOnDnExMTw5MkT1YHVgCpx06QEbtOmTWhpaREfH49UKuXq1avMmDGD/v37M2vWLNq0aUNUVBRisRiZTIaurm6Nvz8CwZ8Ri8VUVFTw/fffEx0dzZQpU1Rtf7NmzVi5ciUzZsxgwYIF9OnT563XCwmcQPDfE5I4NZJIJNy6dYv58+fz22+/kZeXR+/evRkyZAhGRka0bt2amzdvqjtMtUhKSiI+Pp5FixYRGRmJRCKhoqICExMTxo8fj6enJ8uWLWPEiBH8+OOPdOvWjdjYWK5cuYKVlRVSqVSV9ArerTFjxhAREUHbtm1V/1dcXMyTJ0/Q19cHoLKyEm1tbdXeiMuXL5OZmSkkce9R9Zm00tJSDAwMaN26Nbm5uVy6dFgTnE4AACAASURBVIlOnToRHR1NgwYNAHj58iW6uroaN6ukTODOnz/P9evXCQwMxN7eHolEQrNmzbCwsGDq1KkMGTKEadOmERkZiY6OjqosukCgKZQDHsoETC6X8+LFCz766CMMDAxUbU7Dhg3R09NjwIAB7N69m8DAwLf+lpDACQT/PSGJe0/+bJQpLCyMSZMmERcXh7e3N4MGDaJTp04APHnyhFq1amFiYqKOcNXqypUrTJkyhVmzZhEdHa3qVOno6KjuY0xMDBKJhIkTJ5KUlIS/vz/a2toEBQUB/111RMH/RqFQUFhYyKtXr/D39wd+T96U579ZW1szf/58NmzYgLa2tmqPUZ06dXBxcdGYPZ0fAuW9f/nyJYsWLeL06dMEBQUxfvx4Bg0axOvXr5FKpRQWFnL58mVKSkpYsWIFJiYmqjZJk9y4cYOJEyfy8uVL2rZt+8YsZL169Vi0aBHz589nzJgxDBkyhEGDBmlksSmBZlMmcWKxmKqqKsrLy3n69Cm5ubkAqkEjsVhMvXr1aNSoEdeuXVNz1AJBzSE8dd6D6ntIrl69yr1796isrKRJkyZ0796d2NhYdHV1kclklJWVkZmZyc8//8yDBw/o3Lkz8HYBiJqqtLSUSZMmUatWLSwtLVXXrLyHyoeFWCymXbt2ZGdns2HDBvr16/fGHitNuFfqJhKJMDc3Z8WKFYjFYlWRhyFDhuDp6ckXX3zBnDlz6Nu3L3PnzsXCwoLXr19z9uxZCgsLNXKAQh2qqqpUCUavXr0Qi8UEBQXh6ur6Rml8gJEjR3Lp0iXMzMywt7dn3bp1ABqxD646b29vunXrxubNm9m2bRs+Pj507NhR9XMHBwemTp2qWo4qJHACTZKWlkZaWhr5+fnY29szbtw4xGIxUqmUhg0bcvjwYZo3b07Dhg1V342ysjJEIhF2dnaqxE54TgsE/zfCEQPvmEKhUM3ATZ48mXXr1rFr1y4uXrzI+vXrKSsrw8nJCSMjI6ZMmcKXX35JWloaDx8+ZMmSJTg7O79R5rqm09bWprKykgcPHnDmzBn09fWpW7euKnkTiUSIRCJkMhlisRhtbW127NhBVFQU9vb26g5fY+Tm5vL69WsMDQ1VD+JDhw6xZ88eHj9+jJ2dHeHh4ejo6HDkyBHWrl3LrVu32LZtG7t27aJnz55vLL8UvDvK9+e7777jwYMHrFq1ivbt21O/fn2OHj3K6tWrOX78OBYWFsTExODt7U1sbCw9evTAyMhINYtXk1VvY5UDZgEBAdja2pKTk8Pp06cpLy9XLQNTKBRIpVKaNGlCRESEGiMXCN6vw4cP8+2336JQKCgqKiItLY309HSaN2+Orq4uHh4eJCcnc+bMGUxNTbG1tUUkEnHixAnWr19P165d8fX1FRI4geBvULOfzB8AZUO1YcMG9u3bx7Rp0/Dy8sLU1JQvv/ySDRs20LZtWyoqKhgwYAC2trY4Ozvj4uKCu7v7G+fJaYpevXrh4ODAjz/+yE8//UR+fj4DBw58YxZO2al8/fo1tra2mJubqzlqzZKcnMz+/ftZt26d6qyfsWPHYmtry7Jly3jy5AlffPEFffv2pV69euzfv59ff/0VX19f2rRpw6effgpozgzzh+Dp06cEBQVhYWFBVlYWSUlJbNq0CRsbG8rKynj06BHLly+nXbt2qtcoFAqNSOCUbeyqVavIy8vD0NCQr776isjISMzMzFi6dCmbN28mPz+fESNGqAo26OrqqjN0geC9yszMZMSIEfTv35++fftiYGDAxo0b2bBhA69fv0ZbWxtnZ2e2bNnCmDFjGDt2LFZWVqqiJy1bthTafoHgbyQcMfAeFBcXM3LkSFxcXBg9ejS6urrcunWL2NhYhg8fjq2tLY8fP/7Tik2a7MaNGyxcuJDbt28TGRnJiBEjMDQ0VCVyFRUVLFu2jJMnT7Jq1SohkXuPbt68ycSJE7lz5w7l5eV8/vnnjBw5UnXA9/Tp09HR0aFfv3507NgRLS0tysrK3iiSIVQje7eqd5LKy8sZNmwYhYWFREVFceDAAe7cucOoUaPo2rUrqampfPPNN+zatQt3d3c1R/7+VP8MDh48mKtXr6Krq8uzZ8/w8vJS7Qt8+PAhCxcu5MqVK7i5ufHtt98KM/8CjVJSUkLz5s356KOPmDJlCvr6+ohEIu7evUvv3r35+OOPuXnzJn5+frRq1Yp69eqxadMmHj58iFgsJiAggOjoaEDzlmcLBO+K0IN6R6qqqlT/NjAwID8/X1XtraSkhL59+xIVFUXPnj25f/8+K1eu5NGjR2qM+MPj7e3N7NmzCQ4OJjU1lUmTJnH37l1Vp+vIkSOsW7eOfv36CQnce1a3bl1mzpxJaWkpcrkcbW1tVcIQHh7O8uXLqVWrFsuWLWPFihU8efLkrSqHQgL37igTuJs3b3Lv3j10dXUZO3Ysr169YvXq1VhYWLB48WL69euHVCrFxMQEJycn5HK5ukN/r5Sfwblz53L9+nUWLVrE9u3b+eSTT7h8+TLt27fn2rVr2NraMnPmTCIiIsjIyFAduyAQaIqePXvi6OjI5MmTkUqlquJhly9fJj8/n7NnzyKXy9myZQtjx47lypUrdO/enXHjxjFmzBhVAqeJq4sEgndF6EW9A8r9FSUlJZw8eRIAa2tr7t69i1wup1+/ftjY2DBp0iRVWWpDQ0Nq1aqlzrA/SBYWFsyYMYO2bdty9epVpk6dysWLF7lz5w6TJ0+md+/ewt6q90z58H716hWRkZHExMSwbNkypk6dSllZGQAeHh6sXbsWPz8/VqxYwerVq5HJZOoMW6OIRCIqKiqYNm0aLVq04MSJE3h4eJCamsqOHTtYs2YNkZGRyOVycnJy+Omnn7C2tqZu3brqDv29y8nJ4eTJk3zzzTc0bNhQ1Q5HRERgbGxM165dOXz4MPr6+kycOJH4+HiNvE8CzXXx4kXu3r1LUVERz58/B34fALl58yaTJk1iwIABrFmzho0bN7Jq1SqKi4tZvnw58HalaGHwTiD4+wiFTf5m1UeZunbtSnZ2NpGRkTg6OrJ27VoSEhKoqqoiLi4OU1NTiouL2bFjBxKJhE6dOgkjVH9CIpHQqFEjxGIxv/76KydPnmTNmjWEhIQwa9YsdYencZQzbpaWljRv3pzg4GDMzMyIj4/n0qVLNG7cGKlUip6eHq1bt6a4uJhmzZrh4OCg5sg1i0QiwdzcnGfPnvHTTz9haGhIQEAAZmZmPH36lJ49e7J161b27NlDVVUVGzZseKMkuKYoKipi06ZNREVF4eLiwtGjR9m0aRMjR44kIiKC5ORk9u7dS3p6Oo0bN8bOzk7dIQsE75WRkRFeXl6cOnWKzZs3ExoaipmZGTExMTRv3pyxY8diYmKCWCzG1taW7Oxs0tPT6dq1K9ra2uoOXyCosWr2jnU1UHZ+Dhw4gFQqZfz48ejr61OvXj1GjRpFYmIiMplMNfp75coVDh06RGJi4hvnoAne1rNnT6ytrZk1axbe3t78+OOP6g5Joyg/mxcuXODUqVM8fPiQoUOH4ujoSOfOnalTpw7z5s2jb9++zJw5k1evXnH9+nXGjRsHCBvZ37U/u78RERHUqVMHMzMzVXXKb7/9FiMjI8LCwrh//z4BAQG0bdsWHR2dGl2JsqKighs3blBcXIyZmRne3t7A72foVT+sePr06bRt25bGjRsD4OTkhKurK7Vr1xaWbQs0koGBAc2aNcPExITFixfTtWtXjIyM8PT0ZPz48W8tlTc2NsbOzo7Kykr09PSEdl8geEdq5tNazfbu3cvSpUupqKjA0NAQsViMjo4OsbGxWFhYsGnTJgYOHIiOjg5169ZlyZIleHl5CZt9/wPR0dHY2dnh5OSk7lA0ivKzefbsWYYPH46NjQ3a2trcvXsXR0dHDA0Nad++PZaWlsyePZvu3bsjEomIjY1V/Q3hQf5uKe9vZmYmjo6Oqo6Vj48PI0aMwMLCgvXr1/P48WOWLFnCmDFjqKioUC3prn6eXE00ZswYzp07x4sXLzA3N6dnz570798fV1dXlixZgqenJwcOHEChUDB06FAUCgWZmZlIpVJiY2Np1qyZui9BIFAbiURCUFAQkydPJj4+np07d+Li4vLWeZ9ZWVkcPHiQTz75BCMjIzVFKxBohpr7xFajkpISTExMuHTpEtu2bWPo0KHA7/ve2rRpQ5s2bbhx4wZGRkYYGBioGkEhgfvPCPtR3h9l8iaRSCgsLOTrr7+mS5cu9OrVizp16gCwZ88eXrx4gb29PeHh4SxevJi0tDTKysoYOHAgIMzCvS8HDx7kyy+/ZMKECbRv317Vttjb29OnTx9KSkpISkqiQ4cO/PTTT28sDazJKwCGDBnC7du36devH+7u7ixYsICkpCRat26Nvb09QUFBwO9LK8vLy5FIJLx69Yo9e/bw+vVr6tevr+YrEAjUS9mGu7u7M2DAAMzMzFi9ejWvXr1ixowZ6Onp8fjxY77++mvs7OwYNmzYG68TCAR/P2FP3N9AeQi1ko+PDw4ODhQWFrJt2zZKSkr46KOPgH8VPalTpw5GRkbo6uoKDZzgg1W9Y3/69GnS09P5+uuvsba25tq1a4wbN464uDhOnTrFlStXcHd3x9fXF39/fxo0aACgcXus1MnR0ZEzZ86we/duABwcHFSj4bVq1cLGxoa0tDRKS0sBCA0NVVus78ucOXM4evQoS5cu5eOPP8bV1RV7e3u2bNlCixYtsLGxUbXhWlpapKSksGnTJlJTUzl27Bg//PADrq6u6r4MgeC9kslkb7Tb1fspxsbGeHp6YmVlRWJiIidOnKB+/fosWrSIe/fusXPnTkQikdD2CwTvmDAT93+k3ENSWlpKVlYW+fn5NG/enIYNG2JmZoa5uTm7d+/m8ePHzJo1C6lU+saySSGBE3yIxo4dy7Nnz1i9erXqIaxQKMjPzycnJ4cdO3aQkpJCVVUV33//PVZWVnzxxRecOXPmrcRAmGF+d/64BFtLS4stW7YwceJEVq5cSUFBAb1791bNXj979gxXV1fGjRunEefBJSUlkZCQwOLFi6lXr56qUp6WlhaWlpaqQ7uVn3EHBwdmzZrFwYMHMTAwYPTo0TRs2FBt8QsE79uuXbsICwvD0tIS4I0l19WZmZnRqVMnzM3NWbRoEW3atEFXV5d9+/YhEolq9P5ageBDIXzD/g/kcrmqkRo2bBgZGRm8evWKunXr8sMPP+Dm5saIESMwMzNj9+7dDB48mClTpgijuoIPmkwmIzw8HDc3N8RiMQ8ePMDOzo7w8HDs7e0ZNGgQtWrVIiAggNmzZ6se9j4+PohEImH5zHuiTODkcjkpKSkUFRVhbm5OmzZtmDlzJs7OzsyfP5+8vDy6deuGtbU1mzZtorKyUpXA1eT36sqVK0yZMoVZs2YRFRX1RsK7Y8cOysrKeP78Odu3b0cqleLi4oKVlRXR0dE0bdr0TzuuAkFNVlFRwfHjx4mLi2PlypWkpKTg4uJCRETEnyZkUqmUqKgoDA0NSUhIYODAgdjY2LzRNxIIBO+OSPHHQzwE/7XBgwdz7949+vXrh0QiYebMmRgZGTF79mwaNWpEaWkpSUlJrFmzhjp16qgqUdbUzpOg5jh27BiDBw9m6tSpdO3aFYCff/4Zb29vrK2tkUqlvH79mgsXLjBy5EjmzJlD8+bN1Rx1zVe9iu3gwYO5dOkS5eXl6Ovr06ZNGyZOnAjAkSNHGD9+PGVlZejo6GBkZMSmTZtUSwhr6lKn0tJSunTpwsOHD1m8eLFqOTvA999/z5o1azAyMqKyslK1tBR+X3IqlUqZMGECrVq1UkfoAoFaJScns2HDBh4/fszLly/Zt28fzs7O/7a9UCgUvHz5EhMTkxrdrggEHxphT9z/0c6dOzl06BDz588nMjISkUhETk4OL1++JDExETMzM+rXr09AQACWlpZ06NABW1tbIYETfLCq72N49eoVT548IS4ujtevX9O4cWM8PT0xMzPj4MGDJCUlceDAAbZs2UKrVq3o16+fmqPXDMr2Y/LkyZw/f57vvvuOHj16cPPmTQ4fPsy1a9cICwvD29ubVq1a4ejoSMuWLRk0aBB2dnY1vhKutrY2lZWVPHjwgNOnT6Onp4eXlxc7d+5k4cKFTJw4kZEjRxIbG0vHjh1xd3enXr16yOVyHB0d+fLLL9V9CQKBWnh5efH8+XPS0tKQSqVYWlpSv379v9zjppzNV1bDFfo2AsH7IyRx/yNlw7V//36Ki4vp2bMnOjo6JCcnc+vWLUaPHs3Tp09JSEigpKQEAwMDmjZtqlp6JhB8iBQKBWKxmKdPn1JYWIiHhwf+/v4oFAo2bNjAzZs3ad26NQCHDh1i27ZtGBoaEhUVxZgxY4C3C/0I3o309HTWrVvH9OnTadq0KbVq1eLChQvo6elx69Yt0tLSqF+/Pk5OTvj5+eHh4YGxsbHqPa7p/P39sba2Jj09nTNnzpCVlcXSpUuZNm0anTt3pnbt2tSuXRtLS0v8/f0JCQmhdevWtGrVSiPuj0DwR5WVlUgkEi5cuIC3tzempqakpKTw+PFjwsPDVWcp/lXBE4FA8H4Ji5b/R8qGq7S0lOfPn1OrVi2ysrJYtmwZ06dPJzw8nIcPH3L69GnWr1+PVColICBAzVELBH+t+uzM+PHjefjwIXPmzCEgIIAvvvgCCwsLVqxYQceOHVm7di1Dhw6lV69e1KpVS/VQr+kzPOpWfQ9beXk5VVVVWFtbA78n1VlZWUyYMIFLly6xaNEievfuTdeuXRk8eLBqj4omdboiIiKwsLBg4cKF/Pzzz3h6ehIVFYVYLH5jsEG5l1PYByfQRMp2W1tbG/j9SA6AW7dusWnTJvbu3cujR4+YN28ehoaGVFRUAAjfF4FAzYSZuP8jU1NTSkpKaNq0Kd988w12dnaMHTsW+H2kXE9Pj/j4eOGgWMEHraqqSpV87d69m4KCAs6dO8fZs2exsrLC19dXtQ/u+PHjbNmyBT8/P5ydnd9ICoQZjHer+r2+ffs2W7ZsoU+fPkgkEgYOHEhMTAydOnXCw8ODLVu24OjoCEDLli3VFbLa1alTh7CwMJ48ecKtW7fIzMzEy8sLU1PTN5I4TUpuBQKl6gNv69evZ+/evZw7dw4/Pz+sra3x9vZGJBJx8uRJUlNTsbe35+jRo+Tk5Kh+JhAI1ENI4v5D1ZcQVFZWUl5ejkwmw9bWlkaNGiEWi9m8eTOBgYGEhoZSVFTEli1bqKqq4pNPPkEkEgnLzAQfLOXn8ttvv2Xbtm14eHhQv359ioqK2LhxI1KplMDAQLy8vHBxcSE9PZ21a9fSo0cP4azDdywtLY3NmzeTmJjIkydPsLa2xsDAACcnJywtLQkMDGT9+vXcv3+fpUuXIpfLuXTpElevXuWbb76hT58+gGYvc61VqxZNmzbl+fPnnDx5kvT0dOzt7bGxsdHYeyIQwL8G3saMGUNycjJ3797l0qVLbN++ndDQUJycnAgICEBPT4/09HTi4+M5ceIEQ4cOFbaHCARqJiyn/A9Un6VYtmwZly5dIi8vDwMDA/r06UN0dDRyuRw9PT12796NkZER6enpnD17VnXopVCxSfChu3DhAqmpqcyaNYvo6GgAiouLWb16NXPmzOHx48cMGTKE8PBwatWqBYCJiYk6Q67xvvnmG65du0ZJSQkODg4kJibi6emp6jx9+umnADx//hypVAr8vsT72LFj6Ojo4OXlBaAx++D+HR0dHcaMGYOlpSWbN29mzpw5fPbZZ3Tq1EndoQkEanXkyBHS09NZvHgxbm5uXLhwgVWrVtG5c2fmz59Py5Yt6d69Oy4uLrx48QJLS0t8fX2Ffo1AoGbCTNx/QDlSO23aNFJSUvD398fV1RUtLS2WLVuGWCwmODgYV1dXLl68yL59+9DX12fGjBl4eXkJ+4QE/wg3btwgNTWVgQMHYmpqCvze8Q0ODub169fExcXx4sUL/Pz8cHd3x8bGBqjZZ42p03fffcfRo0f56quvGDt2LF26dKFNmzaqM94SEhIwMzPD2NiYgoICtm/fzvnz5zl8+DAHDhxg4cKFODg4CB2tP/D398fKyoqDBw8il8uFIzEEGknZbufm5pKRkUFVVRXdu3fH2NgYNzc3XFxcKCgoYOXKlWhpaREUFIS9vT3u7u6qfbhCuy8QqJcwE/dvVO+cnjt3jkOHDjF9+nSaNGmCjo4O+/btY+/evbi7u3PlyhUCAwNJSkri/v37GBoaYmpq+sYsnkDwITMzM6OqqorffvtNtddNoVCgpaVFZGQkcXFxquU2CQkJb+wnEvy9jh07xi+//MLUqVNp1qwZEomEiooKVXI9c+ZMNm7cyLFjxxg9ejTt27fn/v37/Prrr+jp6bFkyRICAgKEBO4vREdHY2tri7Ozs7pDEQjU5tWrV7Rq1QqZTEbDhg3R19dX/Sw4OBgjIyMsLCxYvHgxGRkZ/Pjjj2qMViAQ/JEwE/cnKioqkEgkb+xju3btGsePH6d79+5YWVnx+PFj+vXrR69evWjatCkTJ05ELBbj4+ODsbGxqjEUOriCD9GfzZ7p6upy+PBhrly5QoMGDTAxMVElAIWFhTx48IBRo0aRnJxMcXExYWFhwizc30x5P7dt24a2tjbdu3dXLV1VFt+Ij49n1apVfPLJJ+Tk5JCamoqPjw+dOnWiU6dOREdHq2brhPfmr9WpU0dVsVMg0DQikQhdXV0aNWrE6dOnuXnzJqampri4uKiqTpqbm+Pl5UVVVRX5+fnExMSoOWqBQFCdMET7By9fvmT8+PGcP3/+jWImr1+/pri4GAsLCwC6d+9OWFgYo0aNQiqV8ujRIwoLC9UZukDwH5HL5arOfX5+PufOnSMzMxMzMzMWLVpEUVERI0aM4MCBA+Tn56uqID59+pSPPvoIT09Pnjx5AghJwt9NOft5/vx5jI2NqV27tupnYrGYkydPMnv2bGbOnMmMGTMYPnw4BQUFJCUl8fr1ayQSiWpvnEAgEFSnUCje+r+AgABVMasFCxawdevWN/oy1tbWDB8+nFWrVgG/Pz8EAsGHQRiG/IPc3Fz27dtHTk4Oo0aNolGjRujq6hIeHo6JiQmjR49GoVBgamrKlClT0NbWprKyEgsLCwwMDNQdvkDwbykUCtXy3rlz53L8+HHu3LmDRCIhKCiIzp07s3DhQhYvXsy4ceMwMDBAW1sbuVzOypUrMTIywtDQEAMDA2Gp3jtSXl6Orq4uMpkM+H1lgHJkvKKigjFjxtCqVSsAmjdvzpQpU6hdu7aQvAkEgr9Uvb3Ozc3l9u3bGBoaIpVKqVu3Lrt372bIkCHMmzePgoICPvvsMxwcHAAwNDT8/9q776gorvaB4186AqKCCKIoiiwoSBELaBARsJdorFEsMYnRaBCNsbwmMZpiiSVgiT32iCVBgwVFY+FNJHYUaywoYqUoiNT5/eHZeV2BqPmZ2J7POTknzN69c2d2d5xn7r3PVeuR6SFCvDgkiHuEu7s7W7duJSwsjLFjxzJixAiCgoKwsrJi0KBBfPfdd9y5c4d58+Zha2vL1atX+eGHH0hLS6NNmzbPu/lC/CVtz9m8efPYsGEDgwYNwtHRkYyMDKZNm8Z3333HwIEDWbhwIbGxsWRkZGBsbIynpyc1atTgxx9/ZN++fSxatEgCuH+Iqakp1atXJzY2lps3b2JjY6O+1rx5cwIDA9XPcffu3ZiZmVG/fn1AkswIIUqmvV7PmTOH6Ohorly5grm5OYaGhnTt2pXw8HDmzp3L5MmTWbJkCVevXuXdd9/Fw8PjObdcCFEamRNXgvLly9OuXTsSExNZtmwZpqamaDQaPDw8MDIy4urVq0RFRZGQkMCaNWs4ceIEM2fOxMnJSWcIphAvotTUVKZNm0bfvn3p3bs3NWvWxNXVldatWxMbG8vOnTsJCgrCx8cHNzc3KlSowLZt25gxYwaxsbF89NFHak+QeDa+//57ADXjp6mpKZs2beLWrVu4u7urvfwPX19yc3PZsGEDqampDBgwAAsLCwnghBCl2rZtG9988w3vvfcen3zyCZ06deLChQusX78ePz8/7O3teeONN6hUqRILFy7Ez89PnV8rhHjxvPZB3MNPrrVJTAoKCihTpgxt2rQhMzOTBQsWkJaWhoeHB/7+/nh5eWFnZ0dmZiYhISG88847eHt7SyZK8VJIT09n6dKltG3bFhcXFwAKCgooV64cISEhrFixgpycHJo2bQpAdnY2P/30Ew4ODnTp0kVdm0w8O3FxcSQnJ9O4cWMAHBwcuHDhAj///DP5+fnY2dlRsWJFNYBLS0sjNjaWyMhIxo0bh6en52u9mLcQonSKopCfn09ERAR16tThww8/xM7OjrJlyzJp0iRCQ0OxtbVly5Yt+Pj44O7uTocOHdQefiHEi+m1H06Znp5O2bJl1cW6ATVjmb6+PmPHjqVatWp88803XL9+nfDwcOrUqYObm1uxoUvSAydeBnl5eSiKQmpqKoC6jICiKJQtW5ZKlSqRm5urlre1tWXy5MkYGRk9rya/8urWrcu8efP48MMP1cy2kydPxtjYmJUrV3Ls2DFatmxJ/fr1uXr1Klu3buXUqVMMGTKEFi1ayGLeQghVXl4e+/btw9/fHyMjI/T09DA2Nub27duYmZmpGW+7detGrVq1eP/991m9ejXr1q2jTZs2ODg4qPPhZO6zEC+u1zqIu337NsHBwdja2gJQu3ZtDAwMqFGjBnZ2dtjY2ODk5ETLli2pXLkyw4cPZ8aMGfTp04cGDRpIIgHxUqpVqxZ169Zl8eLFNGrUCDc3N+DBfLmsrCwsLCzU9ci02cwkgPtntWnThh9++IHJkyczZswYTExMAJg4cSIeHh4sXbqU7777Tv08WrVqxdChQ2nfvv3zbLYQ4gUUHh5OTk4OAQEBOtuto0AjKwAAIABJREFUrKw4deoUACNGjODevXvMnj1bTVaVn59f7FovAZwQLy49paScs6+Jw4cP069fP3JzczE1NaVq1aoYGhpy5coV7t27R1FREQAWFhY4OjqSnJzMnTt3qFixIl9//bU63EyIF1FhYWGx4b3ap6rJyckMHz6cy5cvM3LkSOrXr09OTg4bN27kxx9/ZMOGDbIQ8r9E+5nMnz+fqKgoJk2aRP369SkoKFBHBdy6dYv79+9z8+ZNLC0tcXJyKvZ+IYQAyM/PJy0tDVtbW7Zu3Urt2rWpXr06e/fu5eOPP8bMzIyMjAyio6OpVq0aeXl5TJkyhSNHjrB48WIsLS2f9yEIIZ7Aax3EAZw7d47Ro0dz7949WrduTd++fcnPz+fu3bskJyeTlpZGamoqZ86cAeDMmTMYGhqyfPlyudCJF9Lx48dxc3NTF6sv6QZfURROnDjBggUL2LZtG+XKlaOoqIjy5cszatQogoODSwwCxT9HURS6d+/OnTt3WLp0Kba2tuTn52NoaChz3YQQj7V9+3ZMTEzUB8w7duxgyJAhdO7cmQEDBuDo6Mjs2bP58ccfsba25tNPP8XExIS9e/cyf/58IiMjCQgIkCy3QrwkXusgTnuDm5aWxsiRI0lKSqJt27b06dNHHQ/+qKysLOBB79zDT8qFeBGcOnWKd955h8DAQD7++GMqVKjwl/8gK4rCgQMHOHnyJBUrVsTZ2RlnZ2f5R/xfpg2Yk5KS+Pjjj8nOzmbRokXUqlVLLSM9bkKI0ly9epVevXphbm7O4MGD1SWP5s+fz/Tp02ncuDEjRozAzc2N1atXs3HjRo4dO4axsTGOjo707NmTbt26ybVfiJfIa52dUk9Pj8LCQszNzenYsSOXL18mJiaGlJQU7O3t1blyeXl5GBgYUFRUhImJibrwrtxQiReNvr4+Fy5c4ODBgyQmJuLi4oKVlRWgm4lVm6q+oKCASpUq4ePjg7OzM9bW1mpd8g/5P+fRTJLaa4m1tTUODg4cPXqUOXPmUL58eapWrUqZMmXU65W+vr7caAkhdGiTUp04cYK9e/eSlZVF/fr18fHxwdXVlSVLlpCQkICDgwPt2rUjODiYdu3a0a5dO3r16kWDBg3UuuTaIsTL4bUO4uDBzZP2xki7iG50dDRnz56lQoUKODo6qgGcBG3iRVemTBlCQkLIyMggPj6egwcPUrlyZRwcHNThlfC/oCE+Pp558+ZRvnx5qlSpotYj/4g/e/v37yc6Opo6depgYmJSYiCmp6eHo6MjdevWpaioiO+//57Dhw9z48YN7OzssLS0RE9PTz4fIYRKURQURUGj0eDs7MzJkyeJj4/n2rVreHl54erqSkhICOvWrVOHz9eqVYsqVapgZ2enJmmTh0NCvFxe+yAO0EnPXa9ePapXr84vv/zCsWPHMDIywsXFRQI48dLQ09PD19cXExMTfv/9d3777TfKlCmDq6urTgBw7NgxvvrqKy5cuMDHH38s89/+QdevX6dHjx4cOXKE7OxsHB0dKVeuHKB746T9/0qVKtGkSRNatGjBiRMnOHPmDBEREZw6dYr169fj7++vLkUghHi9adeoLSoqIiYmhhMnTnD+/HkOHz5MSkoKGo0GJycnunfvzq+//qquP6nRaHSybEsAJ8TL5bUP4h5eoPujjz7CzMyMoKAgvL29+fXXX9m4cSMODg5oNJrn3FIhno67uzv29vYcOnSI+Ph4cnJy8PHxAR4k9ImMjOTEiROsWrWKihUrqj3S4tm7du0amzdvJjMzk+zsbM6dO4e9vT2VKlVSe0gf7WHT19fHxsaGVq1a0bFjRxo2bIinpyfly5fXGfokhHi9aa/b7777LomJiXTq1Ilhw4apCawOHDhA5cqVcXR0pGvXriQmJrJlyxbeeecddX1cIcTL57UJ4h6+Qc3KysLY2FinB27EiBHs37+fgQMHUrZsWezs7GjWrBk3btygf//+sk6WeCnVqFEDd3d3kpKS2Lt3L6mpqVSuXJn169fz008/MXv2bLy8vCQT5T/MysqKmjVrkpCQgKWlJefOnePMmTNYW1tTrVo1nUBO6+Hsovr6+tjb22Nvb4+7u/tzPBIhxIvo0KFDrFixgnHjxtG+fXtsbGwIDAzE1NSU3377jfj4eKytrXFycqJt27ZqmUevO0KIl8drEcQ93Ns2cuRIADQajXrhmjx5Mlu3bmXBggU4Ozujp6dHQUEBlpaWhISEYGRkJL0U4qVlY2ODr68vly5dYu/evezcuZO4uDhGjx7Nm2++qfMwQzx72gTAVlZWXL9+HVdXV4KCgti8eTPHjh3DxMREHepaUiAnhBCPc/PmTVauXEmvXr2wtbUlNzcXQ0NDateuTZkyZfjxxx85ceIEGRkZNGrUCDMzM/T19eUaI8RL7JW+c7tw4YJ6g1pYWMj169fZs2ePOnygqKiIzMxM4uPjGTt2LB4eHup7H106QHopxMusUqVKTJw4kXbt2nH27Fn69etHv379nnezXlkPr9yiHSZZrlw5mjVrxpIlS6hatSo//PAD+fn5zJ07l0WLFpGTk6NmnhRCiKdhZWWFnp4eO3bsAMDExITc3FwAgoKCqFSpEhYWFmoGYrmnEeLl98r2xB05coSBAweSl5eHq6srpqamKIrC6tWrcXJywtvbG0VRKFOmDO3bt6du3bryREq80gwMDPDz88PDw4NevXoBsvbYP0X7FFyroKAAfX19HB0d0dPTY+nSpXTr1o1WrVrx22+/sXv3bu7cuYOTkxMWFhbPseVCiBddSSODLC0tycjIYP369RQUFNCgQQP1GnTx4kWOHj3Khx9+SLt27QDJRCnEq+CVDeL09fXZs2cP+/fvJz09HUdHR2xsbIiLi8PMzAx/f3/gwVNyY2PjYtnhhHhVVa9eHUDmwf1DTp48SWhoqDo00traGiMjI/XaYmFhwe+//052djaBgYG0atWKM2fOEBsby5kzZ9ShTkII8SjtdbugoICoqCi2b99Ofn4+FStWxM3NjbNnz7Jjxw4SEhKws7Njz549rFixghs3bjB27FhA7nOEeFW8skGcubk5nTt35syZM2zZsoVr167h5OREamoqW7ZsITk5mf3795ORkUFycjLm5uZcv35dXRhZiFed9MA9e/fv3+eDDz7gwoUL7Nu3j3PnzhEdHY2VlRVlypTBwsKCihUroqenx5QpU6hTpw7Ozs60aNGClJQUjIyMaN++/fM+DCHEC+jh+f3vvfceP//8M3/88QebNm0iNzeXhg0bEhwcTF5eHocOHWLRokUcP36c8uXLM3v2bMzNzWV+vxCvED3lNZiAMWvWLJYvX46Xlxe3b9/mwoUL2NracufOHW7duoWhoSEFBQW4ubmxdu1aucAJIf62GTNmsHv3bu7evYunpycWFhasXbsWDw8PXFxcGDhwIDY2NsyaNYvbt2/z0UcfYWtrq1OHDHMVQpTms88+Iz4+ns8++4wmTZowZMgQ9u3bR8uWLRkyZAg1atTg/v376lImxsbGWFhYyOgLIV4xho8v8vIbMmQIlStXZsaMGWRkZODt7c3y5csBOHPmDHfu3OHUqVMEBgaqiQVkqIEQ4mmMGzcOHx8fwsLCsLa2ZvPmzSQmJjJq1CjeffddfvjhB/773/+yadMmWrVqxZ07d7hy5QonT54sFsRJACeEKEliYiIJCQmMGTOGgIAA0tLSsLe3p169esTExJCamsrgwYPx9fVVlyPRPquXAE6IV8srO5zyUXXq1MHT05PDhw/z559/YmZmRo0aNdS1lzw8PLC0tJShBkKIp6IoCjdv3mTy5MlkZGTQsWNHPD09MTMz48KFC6xfv55atWoxcOBAunfvjrm5OWfOnCElJYUzZ85w6NAhunbtiqGhoTw8EkIU8/CD5YsXLxIdHc2AAQOwsrJi27ZtrF+/nsjISPz9/VmwYAEXLlzgt99+w9/fX2fOvxDi1fLaBHEA9vb2NGvWjKNHj7Jnz54Ss8FJACeEeBp6enqYm5tTtWpV5s2bR/ny5fHw8MDZ2Zlq1apx/fp11qxZw+3btwkICMDLy4uGDRvSvHlz9PX16dKlC3Xq1JEbLSFEMQUFBRgYGHDr1i1u3rwJwPbt2/H19cXKyor+/fvTrVs3QkJC0NfXZ9WqVZibm1O3bl0CAwOfc+uFEP+k1yqIgwdpeNu2bUtSUhJr1qzB09MTZ2fn590sIcRLztramtOnT3P8+HGaNGlC2bJlsbe3p3bt2mRlZbFlyxYOHTrEG2+8gbW1NVZWVjRt2lS9/sgwbiHEwxRFwcDAgOzsbLp160aFChXw9fXFxsaGgIAAli9fTnJyMpGRkQCcP3+e5ORkPv/8czp16gSgZskVQrx6XstuJxMTE6ZNm8bkyZNp3br1826OEOIVUL58efr27UtiYiJbt25Vt9esWZOwsDC6du1KUlISgwYN4tSpU8XeLzdaQgith4OvX375BWtra+rXr4+FhQVvvvkmxsbGFBQUoKenR2pqKvfu3WPTpk3cvn2bmjVrAg+CQBldJMSr67XriXuYq6srUPLCmUII8Ve0PWcP96BVq1aNrKws1q1bh4+PD7a2thQWFmJubo6Pjw+mpqbExsZy/fp1WrVq9ZyPQAjxotFeT7TXlAkTJrBr1y4qVapEnz59gAdDLPX19Tl79iybNm0iKSmJ6Oho4uPjmTVrFlWrVpUMt0K8Bl7rIE5LLnRCiKeRk5ODkZERQLFAzszMjE2bNmFpaUnDhg3R19ensLAQIyMjPDw8qFmzJqGhoRgbG8sQSiGEjkevB0ePHmXz5s1cu3YNFxcXHBwcMDR8kFi8bt266OnpcenSJezs7AgPD6dBgwYSwAnxmngt1okTQohnZcaMGSxatIiePXvi4+NTYo/aggULmD17Nt9//z2+vr5A8bXf5EZLCKGVlpbGpUuXOHLkCJaWljRu3JjKlSsDsGvXLj755BPs7OwYN24c9erVUx8iAdy/fx9TU9Pn1XQhxHMiPXFCCPEEFEWhoKCAs2fPUrZsWbZv386WLVvYtm0bd+7cQV9fn8qVK6MoCjVq1CAhIYG0tDQCAgKA4j3+0gMnhABITU3lP//5D8uXL+ePP/5g27ZtKIqCt7c3xsbG1KhRA19fXzZt2sTOnTuxs7OjSpUqaiBnYGAg1xMhXkPSEyeEEH/DmTNnOH78OFu3buXEiRPk5OTg5+dHYGAgLVq04Pfffyc8PJylS5dSv379591cIcQLqKCggLZt21KjRg06depEw4YNycnJITMzk9q1a3Pjxg3S09NxcXHh2rVrhIeHc+bMGYYPH07r1q2xsrJ63ocghHhOJIgTQojHyMnJ4fr16yQlJWFsbIyvr6/O+pKHDx/mzz//ZNWqVaSnp5OdnU3//v3ZsGEDlSpVIjIyUm62hBDFDBs2jGvXrjF9+nTs7e11Xrt//z5dunQhPz+fsWPHEhAQQH5+Pp9++ik///wzw4cP5/33339OLRdCPG+Gz7sBQgjxIissLGTUqFEkJSVx5coVLC0t+fjjj+nUqRN6enoYGhri7e2Nt7c3Xbp04dChQxw7doz169eTlpZGlSpVJIATQhSTmJjIxYsXGTRoEHZ2dsCDnjlt4pK+ffuSkZGBjY0NU6ZMISUlhbfffptJkybh6OhIUFDQ82y+EOI5k544IYT4C6GhoeTl5dGtWzeaNGnCoUOHcHNzo3r16jrl8vPzdZINZGdnc+7cOVxdXTExMaGwsBADA4N/u/lCiBeUNmHJpk2b1CBOa/DgwZw5c4aoqChOnz7NtGnTuHv3LjNnzqR27dpqObmuCPH6kp44IYQoxYwZM7h27RqRkZHUqlULQ0ND2rRpQ1FREenp6Rw5cgR9fX38/f0xMjJSM04WFRVhbm6Op6cn8CATpdxoCSEelpaWpl4btOvVahfvTk9PZ+LEiVhZWeHn50fr1q2ZOXMmlpaWOnXIdUWI15cEcUIIUYI7d+5w9OhR3nrrLWrUqKEOcQKYPn06W7du5cqVKwDUq1ePiIgIKlasCBTPRClLCQghHmVra0t2djbHjh3TGRpZuXJlFixYoM67zcnJ4ciRIzRu3Bh7e3tZX1IIAYDcWQghRAkKCgo4deoUVlZWmJiYkJeXR3JyMu+99x4LFy6kTJkyfPrpp4wfP57r168zYcIECgsLn3ezhRAvCW9vb9zc3Jg9ezanTp1StyuKopM46fLly6SkpKiLe0sAJ4QA6YkTQogSlStXjtq1a7NhwwacnZ05ePAg0dHRXLp0iQ4dOjBy5EhsbGwAuHv3Lhs3biQnJ0fn5ksIIUpjbm5Ox44d+frrr5k7dy4DBgzAw8NDJ0i7ceMG8+fPJzs7W81EKT1xQgiQIE4IIUpkYGBAcHAw8+fPp2fPngDY2dkxadIkfH19sbKyUpMKFBQUUK5cOZ0hl0II8Th9+vQhKyuLiIgIUlJS6NChA926daOgoIA9e/awc+dO9u/fz7JlyzA2NpZEJkIIlWSnFEKIv7B3716uXr2KkZERTZo0wdbWFvjf0/D09HQ+/fRTTE1NmTp1qjwhF0I8ta1bt/LVV1+Rnp5OYWEhlpaWFBYW0qxZM0JDQ/H09JQATgihQ4I4IYR4Crm5uZiYmAAPhjqtWLGCFStW8NNPP1G9enUZ6iSE+Ftu3brF6dOnuXDhAhYWFri6uuLo6IipqalcV4QQxUgQJ4QQTygzM5OpU6fi6OiIiYkJ8fHxHD16lEmTJhEQECBPyoUQz5QEb0KI0sgEDiGEeEKXL18mLi6OnJwccnNzCQkJYeLEiQQEBKAoigRwQohnSgI4IURppCdOCCGeQl5eHhcvXsTY2JiqVauqyUzkibkQQggh/i0SxAkhhBBCCCHES0QW+xZCCCGEEEKIl4gEcUIIIYQQQgjxEpEgTgghhBBCCCFeIhLECSGEEEIIIcRLRII4IYQQQgghhHiJSBAnhBBCCCGEEC8RCeKE+JeMHj0aFxcXnf9q165NvXr16Nq1Kz/99NMz3Z+iKEydOpVGjRrh5eXFypUrn2n9r7rQ0FBcXFxo2LAhBQUFJZYpLCykUaNGuLi4EBkZ+bf2c/ny5SduT/Pmzf/WPsSTiYyMxMXFhStXrvy/yjwru3fv5oMPPqBp06a4u7vTvHlzJkyYwM2bN59bm0py5cqVYr+BzMxMBg0ahJeXFw0aNCApKQkXFxdGjx79XNoohBCvGsPn3QAhXjdjxoyhQoUKwINAKysri40bNzJ69GjS09N55513nsl+fv31VxYuXEizZs0IDg7Gx8fnmdT7usnMzOTAgQP4+voWe+2PP/4gIyPjb9e9fv16vvjiC44dO/bYsh988AE5OTl/e1/i5ZGfn8/48eNZt24dXl5e9OrVi3LlypGUlERUVBTbt29n9erVVK1a9Xk3FQArKyumTJmCi4uLuu37779n586d9OvXj5o1a+Lo6MiUKVOoVq3ac2ypEEK8OiSIE+JfFhwcXOzmq0uXLrRp04bZs2fTu3dvjI2N/9/7OX36NADDhw/XubkST65ixYpkZ2cTFxdXYhC3Y8cOrKysSEtL+1v1//HHH+Tm5j5R2SZNmvytfYiXz9y5c1m3bh3Dhg1j0KBBOq916NCBfv36MXTo0Gfee/93mZmZ0bFjR51tp0+fpnz58owZM0bd9mgZIYQQf58MpxTiBWBqakrz5s3Jysri7Nmzz6TO/Px8AMzNzZ9Jfa8jExMTmjRpQlxcXImv79ixQ4Y4imfq1q1bzJ8/n0aNGhUL4ADq169P586dOXnyJEeOHHkOLXwy+fn5cu0RQoh/kARxQrwg9PT0gAfzrLQOHz5M//798fb2xtvbm3feeafY0LvmzZszbtw4xo4dS926dWnatCleXl7MmjULgKCgIJ1A48CBA/Tr10+ts0+fPvzxxx+PrTMtLU2dk7N27VpatmyJh4cHb731FseOHePmzZuEhYXh7e2Nv78/M2bMoKioSK0zPz+fefPm0aFDB7y8vPDw8KBDhw6sW7dOZ98uLi7Mnz+fJUuWEBwcjLu7O+3bt2fLli3Fztnu3bvp3bs33t7eNGnShPDw8GLzgnbt2kWPHj3w9PSkQYMGDB06lAsXLjzx5xIcHExKSoras6mVmJhIamoqLVq0KPF9j9tvaGio2pPy8Fyh0s59SXPi/vzzT8LCwmjUqBE+Pj6EhoZy4MAB9fW8vDy++uorgoKCcHd3JyAggC+++ILMzMzHHvf58+cZNGgQ9evXp1GjRnz55ZdERUXpzL2KjIykbt26bN++nSZNmuDt7c3atWsBSE9PZ/z48fj7++Pu7k7Lli2ZP3++zve7tLlcj26PjIykTp06nD9/ntDQUDw9PWnevDlz5szRqQ8eDH+dOHGiut/WrVuzdOlSFEXRKZecnMzQoUNp0KABjRo1YsaMGcXKPO789OnTBw8PD5o1a8Z3332nPjg5f/48Li4uTJkypdj7pk6diru7e6mfwbZt28jPz6d79+6l7nvYsGHEx8fj5eVVapkTJ04wdOhQGjdujJubG35+fowYMYJr167plFu9ejXt27fH09OTRo0a8eGHHxZ7kLRt2zbeeustvL298fHxoX///hw8eFB9/eE5cdr/T0hIICUlRee7XdKcuCf5fbq4uDBz5kw++OAD3N3dadOmTanzVIUQ4nUhwymFeAEUFRWRkJCAsbExTk5OAMTHxzNw4EBcXV0JCwsjLy+PDRs20KtXL5YsWUL9+vXV98fExFCjRg3+85//cOvWLVxdXfn555/Zvn07Y8aMUYdvxsXFMWTIEKpVq6Y+5V+7di39+vUjIiKCoKCgUuu0srJS64iNjaVv374oisLcuXMZOnQoZcuWxdnZmdGjRxMbG8v333+Po6MjnTp1Ah7MBdyyZQs9e/YkNDSU9PR0oqKi+M9//kO1atVo2LChuu/Vq1dTVFREr169MDU1ZenSpYSHh+Pk5IRGo1HbN2LECJydnRk6dCj5+fksXryYxMRENmzYgKWlJRs2bGDs2LH4+fkxcuRIMjMzWb16Nd26dSMqKooaNWo89rMJDAzEwMCAuLg4nWGpO3bswNHRUf28HvYk+/3ggw8oKiriwIEDxeYKlXbuH3bx4kW6deuGoaEhvXv3xsrKih9//JH+/fuzcuVKPDw8mDBhAr/88gt9+vTBwcGBs2fPsnLlSi5dusTixYtLPearV6/y9ttvA/DOO+9gaGjIypUr2bRpU7GyBQUFjBs3jgEDBpCXl4ePjw+ZmZn06NGDlJQUevToQY0aNYiPj2fatGkkJSUxc+bMx573RymKQv/+/XF2dmbkyJHs37+f7777jmvXrjFhwgQA7t27R+/evUlNTeXtt9/Gzs6O33//na+//pqLFy/y+eefAw96u3r06EF+fj59+/bF1NSUVatWcefOnSdujzZ4HjVqFAkJCcyZM4fU1FQmTZpEzZo1cXNzY+vWrXzyySc679uyZQv+/v6UK1euxHpPnDgBgKenZ6n7Lun78LDTp0/z9ttvU716dd5//33KlCnDoUOHiI6O5saNGyxfvhyAjRs3Mn78eN58801CQ0NJS0tj6dKlhIaGsn37dsqWLUtCQgLh4eE0bdqUrl27kpOTw4oVK+jfvz8xMTE4ODgUa9uUKVP4/vvvSU9PZ8yYMaXOg3ua3+fSpUvx9PRk3Lhx3L9/H0NDuX0RQrzmFCHEv2LUqFGKRqNRTpw4ody+fVu5ffu2cuPGDeXw4cNKWFiYotFolK+//lpRFEUpLCxUgoKClB49eigFBQVqHdnZ2UpISIjSsWNHdVtgYKDi6uqqXLp0SWd/ERERikajUS5fvqwoiqLk5+crTZs2VQICApS7d++q5TIzMxV/f3/F399fycvL+8s6AwMDFRcXF+XUqVPqtsmTJysajUYZNmyYTjvd3NyU4cOHK4qiKDdu3FBcXFyUb7/9Vqe+P//8U9FoNMrEiRPVbRqNRvHy8lJu3Lihbjty5Iii0WiU6dOnq+enSZMmSvv27ZWcnBy1XHx8vKLRaJQVK1Yod+/eVerVq6eEh4fr7PPGjRtKgwYNlMGDByt/pXfv3kpgYKD6/507d9Z5vXXr1sq3336rXL58WdFoNEpERISiKMpT7Vf7nXhYaef+4fYoiqKEhYUpHh4eysWLF9VtaWlpio+Pj/LRRx8piqIoHh4eyhdffKFTz4wZM5TOnTsrWVlZpR77mDFjlDp16ijnzp1Tt127dk3x8vLS+U5pv2PaY9eaOnWqotFolO3bt+tsHz9+vKLRaJRff/1V5/3a+rQe3a79e/DgwUpRUZFabsSIEYqLi4vazoiICMXNzU3n+6koijJt2jRFo9EoJ0+eVBRFUSZNmqS4uLgox48fV8vcunVL8fX1LbE9JbUtLCxMZ/vo0aMVjUaj7nvx4sWKRqNRjh49qpY5dOiQotFolJiYmFLrf++99xSNRqPk5uaWWqa0Nmnb/dlnnymenp5Kenq6Trnw8HBFo9Go2999912lbdu2OmV+/fVXpU2bNsqBAwcURVGUzz//XPH29tY576dOnVJatGihbNmyRVEUpdhvQFGKf18V5cFve9SoUYqiPN3vRKPRKD4+PkpmZuYTnxMhhHjVyXBKIf5lnTp1ws/PDz8/P9544w26d+9OXFwcoaGhjBgxAoCkpCQuX75McHAwmZmZpKWlkZaWxv379wkMDOTkyZM6w6KqVav22KxvSUlJXLt2jV69emFhYaFut7S0pHfv3ly/fp3jx48/ts5q1arp9Ehpn5aHhISo28zMzLC2tlZTodvY2HDw4EEGDx6sllEURR0SlZ2drbMPHx8fbGxs1L9r164NoNZ3/Phxbt68Sbdu3TA1NVXLNW7cmLVr19KxY0fi4+PJysoiODhYPX9paWkYGBjg6+vLvn37nnhIVnBwMCdOnOD69evAg+Fyf/75p84xaz2L/T7u8ywqKmJQYpfjAAANcElEQVT37t0EBARQvXp1dXuFChVYtWoV48aNA8DOzo7NmzezYcMGtZdp2LBhrF+/vtT5SoqiEBcXh7+/v04vo62tLR06dCjxPW+88YbO3zt37sTJyYng4GCd7drPv7Q5ho/z/vvvq8OOAfr374+iKOzatQuA2NhYNBoNNjY2Oude2w5tuT179lC3bl3c3NzUuqytrWnbtu0Tt2XAgAE6f4eGhgIPhvgCtGnTBn19fZ1hwDExMZiZmREYGFhqvfr6D/5ZfnSY6NMYP348O3fupHz58uq2rKwsTExMgAc9lvDg+3H+/HlmzZqlDl0NCAggJiZGzWZrZ2dHdnY2X375JX/++SfwYHjjtm3baNWq1d9u49P+Tjw9PbG0tPzb+xNCiFeNjEcQ4l82depUKlasCDy4YbO0tMTJyUm9wYIH83UApkyZUuK8GoDU1FTs7OyABzegj6O9SStpCGHNmjWBB8PovL29/7LOR7cbGBgAxYd4GRgY6MwxMjY2ZuPGjezbt4+LFy9y6dIlNXhTHpmL9Ghd2myd2jl2KSkpADoBjJaHhwfwv3MYHh5e4nEApKWlUalSpVJf1woKCuLrr79m586d9OzZkx07dmBnZ0fdunXVtmg9i/0+7vPMyMjg3r17JR6/drgpPLiZHzZsGGPGjOHTTz/Fy8uLkJAQ3nrrLcqWLVtq3RkZGTg6OhZ7Tfs9eVx7r1y5gr+/f7FyNjY2WFpaFjtnT+rRoava49fWl5yczP379/Hz8yvx/ampqWr5h4cOa5V2fCV5tKw26Nb+zmxtbWnQoAHbtm1j1KhRFBUVsXXrVoKCgihTpkyp9WofXty+fftvLyGgp6dHeno68+bN4/Tp0yQnJ3P16lX1d6b9HX344YccOXKEyMhIIiMjqVWrFs2bN6dr167q8fTu3Zt9+/axYsUKVqxYQdWqVQkMDKRLly64urr+rfbB0/9OHjeEVAghXjcSxAnxL6tXr95jb860N1lhYWGlJi94+CZSG0j9lUcDpZJeMzIyemydpc1FebiH5FF5eXkMGDCAgwcP0qhRI/z8/OjXrx8NGzakWbNmxcpreyNKoz0/f1VOW2bixImlnu/S5iU9qmrVqri6uhIXF0fPnj3Zvn07ISEhJR7zs9jv4z5PbS/N486Tn58fu3btUv+Lj4/nm2++4YcffmDDhg0l3hhrez9KWubi4QcND3u0HX/1XSsqKtL5npWktF6oR9+nPdfa81VYWIiPjw9Dhgwp8f3agEBPT6/EpR3+qt2PevSz17734c+uffv2jBs3jqNHj3L//n1u3rxJu3bt/rJeb29voqKiOHr0aKnfn8TERKZMmULfvn2L9XbCgzUiBw8eTKVKlfD19VUXC9+3bx/z5s1Ty9nZ2REdHc3+/fuJi4tj7969alKhxYsX07BhQywsLFixYgVHjhxhx44d7Nmzh+XLl7Ny5UqmTJlC+/btn/icPexpfydPco0TQojXiQRxQryAqlSpAjwYlti4cWOd144dO0ZmZqbOMMKnqfP8+fPFXtNmg9P27D1rmzdvJiEhga+++oouXbqo27XDE59W5cqVAbh06VKx9dPGjBlDvXr11OO1srIqdg73799PUVHRU63HFxwczPz58zl//jyJiYnq0NdHPev9lqRChQqYmppy6dKlYq8tWrSIW7duER4ezsmTJ7Gzs6Nt27a0bduWoqIilixZwpQpU4iJiVGHAD7M2toaMzMzLl68WOy1kvZXkipVqpT4Pbt58yZZWVnq56cN/vLy8nTK3bp1q8R6L1++TK1atdS/tW3U9shVqVKF7OzsYuc9MzOT3377TS1XtWrVEo/v8uXLT3B0D6SkpODs7Kz+rf0NPTwMtmXLlkyYMIGdO3eSk5ND+fLlH7veX0BAAMbGxqxbt67U4Z3R0dEkJCTQp0+fEl+fOHEi1atXZ/369ZiZmanbH01Mo824qh3eDXDw4EH69u3L8uXLadiwIRcuXODu3bt4eXnh5eXFxx9/zLlz59QES383iPs3fidCCPEqkzlxQryA3N3dsbGxYfny5TrzxbKystThcU/7ZNrNzQ0bGxtWr15NVlaWTp2rVq3CxsYGd3f3Z3YMD8vIyADQuQEHWLZsGcBTpwt3d3fHysqKDRs26AQAhw4dYsOGDdy7d4/GjRtjYmLCwoUL1dTv8CBwHDx4MN9+++1f9h4+Kjg4WE3ZX65cORo0aFBiuafZrzaIeXgphidhaGhIkyZN2L17tzpEEB4EK4sWLSI5OZmMjAy6d++u0/Oir69P3bp1dfb9KH19fZo3b86ePXt0gprMzEx++eWXJ2pfYGAg58+fZ8eOHTrb58+fD6D2vmqHDp46dUotk5WVpc4re5Q2q6LWkiVLMDQ0VJdeaN68OadOneLXX3/VKTd37lzCwsLU1PktWrTg7Nmz7NmzRy1z9+5doqOjn+j4AKKiooq1RU9PT2cZCEtLSwICAti9eze7d++mZcuWj+2FtLa2pm/fvvz3v/9l0aJFxV7ft28fq1atwsXFpcQhofDg92Zvb68TwKWmphIbGwv8r6czLCyMTz75RKfns06dOhgZGanfjy+//JLBgwfrXIdq1qyJpaXlY3uC/8qz/n0KIcTrRnrihHgBGRkZ8emnnzJs2DA6d+5Mly5dMDExYe3atVy9epVvv/32qVNsP1znW2+9pfaIrVu3jhs3bhAREfH/uin7K40bN8bQ0JBPPvmEXr16YWhoyK5du9i3bx9GRkbFEps8jrGxMaNHj2bUqFH07NmTDh06kJ2dzbJly3BycqJr166YmZkxfPhwvvnmG7p3706HDh0oKChg1apV5ObmMmrUqKfaZ+3atalSpQr79u2jc+fOpQbRVlZWT7xf7XDGiIgIdZjpkxoxYgRdu3ala9euarKaqKgo7t27x7Bhw6hUqRLt27dn1apV5OTk4O3tTUZGBitWrKBixYq0bt261LrDwsLYvXs33bt3JzQ0FGNjY3788Uc1Ocrjbq4HDhxIbGwsw4YNo2fPnjg6OvL7778TGxtLixYtCAgIAB4Exl9++SUTJkwgJSUFY2NjoqKidIKPh/30009kZWVRr1499u7dy65du/jwww/VXh3tfocMGUKPHj1wdnbm4MGDREdH07RpU5o2bQo8SIiyceNGhg4dSt++fbGysmLNmjVPNZxy06ZNZGVl4eHhwe7du9m1axfvvvtusXmK7dq1IywsDHgQED2JoUOHcvbsWaZMmcKOHTsICQnBxMSEw4cPExMTg7W1NTNnziz199q0aVM2b97MZ599Rt26dbly5QpRUVHk5OQA/0skNGDAAMaNG0e/fv1o1aoViqIQHR1Nbm6uusRE//79ee+99+jVqxdvvvkmJiYm7Nixg+TkZCZPnvzE5+tRT/M7EUIIUZwEcUK8oFq2bMnixYuZO3cuc+bMQV9fH2dnZ+bOnfuX2e2epM45c+Ywe/ZsDA0N8fT05KuvvtJZd+5Z02g0REREMGvWLKZPn465uTnOzs4sWbKEVatWkZCQQH5+/mN7KR7WsWNHLCwsmDdvHtOmTcPS0pLAwEBGjBihBgH9+vXD1taWJUuWMGPGDExNTXFzc2Pq1Klq9r2nERQUxLJly0pd4FvrSffbs2dPfv/9dxYuXEhiYuJTBXFOTk6sWbOG6dOns3DhQvT19fHw8GDy5MnqML+JEyfi4OBATEwMMTExlClTBj8/P8LDw/8yUUS1atVYsWIFkydPZt68eZiYmPDmm29iYGDAokWLHjvMrXz58qxZs4aZM2eyefNm7ty5g4ODA5988gn9+vVTy1lZWbFgwQKmTZtGREQEFSpUoFu3btSsWbPEhBezZs1i9uzZxMbG4uDgwMSJE+nWrVux/UZERLB161bWrFmDvb09gwcP5v3331eDHgsLC1atWsXUqVNZs2YNhYWFtGnTBmdn5ycOtBYsWMCXX37JL7/8gq2tLWPGjNE5Nq3AwEAsLCywsLB44t+YiYkJs2fPZuPGjaxfv55FixaRmZmJra0toaGhfPDBB3/5+Y0fPx4zMzN27txJdHQ0dnZ2vPnmm4SEhKjfuTp16tC1a1eMjIxYtmwZ06dPp6ioCHd3dxYsWECjRo2AB5lH586dy7x585gzZw65ubk4Ozszffr0p8rmWZJn/fsUQojXiZ7yNI8ehRBCvPJu376NlZVVsR63iRMnsnr1ao4ePfpUAff/V2RkJLNmzSIuLu5vZ2x8XvLy8mjcuDHdu3dn5MiRz7s5QgghXhEyJ04IIYSOsLAwNRGKVk5ODrt27cLV1fVfDeBedjExMdy9e5fOnTs/76YIIYR4hchwSiGEEDo6duzIuHHjeP/99wkKCiI3N5eNGzdy7do1vvjii+fdvJfC4sWLOXToEHv27CEwMLDYGndCCCHE/4cEcUIIIXR07doVExMTli1bxtSpU9HX18fd3Z0ffviBhg0bPu/mvRQKCwvZt28fnp6eTzzPTgghhHhSMidOCCGEEEIIIV4iMidOCCGEEEIIIV4iEsQJIYQQQgghxEtEgjghhBBCCCGEeIlIECeEEEIIIYQQLxEJ4oQQQgghhBDiJSJBnBBCCCGEEEK8RP4P8L1QO7rYQVUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')\n",
    "### Balkendiagramm\n",
    "df = df_visual\n",
    "\n",
    "import scipy as sp\n",
    "from matplotlib import pyplot\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set(style=\"whitegrid\",font_scale=1.5)\n",
    "sns.set_palette(sns.color_palette(\"Set2\"))\n",
    "\n",
    "a4_dims = (11.7, 8.27)\n",
    "fig, ax = pyplot.subplots(figsize=a4_dims)\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"value\", hue=\"variable\", data=df, alpha=1)\n",
    "\n",
    "ax.set_xticklabels(ax.get_xticklabels(), rotation= 50)\n",
    "\n",
    "\n",
    "sns.set_palette(sns.color_palette(\"Set3\"))\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"paper\", hue=\"variable\", data=df, alpha=0.7)\n",
    "handles, _ = ax.get_legend_handles_labels()\n",
    "ax.legend(handles, [\"Precision\", \"Recall\", \"F1\"], bbox_to_anchor=(1.05,1),loc=2, borderaxespad=0.)\n",
    "ax.set(xlabel='Performance Metrics grouped by Classifier')\n",
    "ax.set(ylabel='value')\n",
    "ax.set_title('Classifier comparison for visual data')\n",
    "\n",
    "fig.savefig(\"feature_clf_visual.pdf\",bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')\n",
    "### Balkendiagramm\n",
    "df = df_meta\n",
    "\n",
    "import scipy as sp\n",
    "from matplotlib import pyplot\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set(style=\"whitegrid\",font_scale=1.5)\n",
    "sns.set_palette(sns.color_palette(\"Set2\"))\n",
    "\n",
    "a4_dims = (11.7, 8.27)\n",
    "fig, ax = pyplot.subplots(figsize=a4_dims)\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"value\", hue=\"variable\", data=df, alpha=1)\n",
    "\n",
    "ax.set_xticklabels(ax.get_xticklabels(), rotation= 50)\n",
    "\n",
    "\n",
    "sns.set_palette(sns.color_palette(\"Set3\"))\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"paper\", hue=\"variable\", data=df, alpha=0.7)\n",
    "handles, _ = ax.get_legend_handles_labels()\n",
    "ax.legend(handles, [\"Precision\", \"Recall\", \"F1\"], bbox_to_anchor=(1.05,1),loc=2, borderaxespad=0.)\n",
    "ax.set(xlabel='Performance Metrics grouped by Classifier')\n",
    "ax.set(ylabel='value')\n",
    "ax.set_title('Classifier comparison for meta data')\n",
    "\n",
    "fig.savefig(\"feature_clf_meta.pdf\",bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3EAAAKSCAYAAAB801LUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvIxREBQAAIABJREFUeJzs3Xt8z/X///H77Igtcg4LYXPYwYgl5wlDk5aETOQTRSmlmMohKtRozLE+TjmmnD5YOfORT4YOPvqIytjIoZo5DDu9X78//Pb+erepTdtee9nterm4XHo/X8/X6/V4v997ZXfP1/P5cjIMwxAAAAAAwBJKmF0AAAAAACD3CHEAAAAAYCGEOAAAAACwEEIcAAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDiqkrV65o/vz5Cg8PV5MmTdSoUSP16NFDK1eulM1mc+gbEhKiiIgIkyqVZsyYIV9fX506dcretnfvXnXu3Fl+fn7q06ePVq9eLV9fX+3bt8+0Ou80RfkzPX36tCIiIhQQEKDg4GAlJSUV2LkSExML7Nh/9/i+vr4aNWpUoZ8XAGAuF7MLAFD4jh8/rueee06nT59WWFiYHnvsMaWmpmrbtm0aM2aM9u/fr/fee09OTk5mlypJ6tChg+69916VK1dOkmSz2fTKK6/I2dlZkZGRqlKlinx8fDRlyhTVrl3b5GrvHE2bNi2yn+nkyZN14MABPf/886pYsaL9ZyO/jRkzRvHx8fr4448L5PizZs3SmjVrtGXLlgI5/q189tlnGj9+vA4dOlSo5wUA5A9CHFDMpKamasiQIUpOTtann36qevXq2bc9/fTTGj9+vJYtW6aAgAD169fPxEr/T7169Rzq/PXXX5WUlKQBAwboySeftLd7e3ubUd4dy9vbu8h+pkePHlX9+vU1dOjQAj3Pnj17VK1atQI7/n/+8x9lZmYW2PFvZf/+/UpNTS308wIA8ge3UwLFzLJlyxQfH6/IyEiHYJRl5MiRKlOmjFasWGFCdbmTnp4uSSpdurTJlcAs6enpfP8AgGKLEAcUMxs3blSpUqXUtWvXHLd7eHjok08+0dq1a295DMMwtHz5cvXo0UNBQUHy9/dXaGio5s2bJ8Mw7P0uXryoUaNGqW3btvLz89NDDz2kqKgohxGAtLQ0vf3222rfvr38/PzUpk0bjR8/XhcvXrT3uXlO3IwZM9S+fXtJUkxMjH3OVk7zt1JTUzVt2jSFhITIz89P7du3V3R0tNLS0ux9svb74osvFBISosDAQM2YMeOW7z0tLU0zZsxQx44dFRAQoE6dOmnevHkOoynXrl1TVFSU/bwhISF6//33de3atWzn/eGHHzRs2DAFBQXpgQce0OTJk5WZmak1a9aoU6dOatSokXr16qUffvjB4fNo0KCBjh8/roiICAUGBiokJESzZs3KNqrz/fff64UXXtCDDz6ohg0bqnnz5nrllVd09uxZh+P5+/try5YtatGihYKCgrRq1aocP9MvvvhCjz32mIKCgtSkSRMNGDBABw8edDhnZmamPvroI3Xq1El+fn5q2bKlxo4d6zBvbd++ffL19dWXX36p8ePHq3nz5goMDNRTTz3l8F7/KGu/06dPKy4uTr6+vvbvKy/nXbNmjcLCwuTv76/IyMgcz/XH86xevdrh++vevbv8/f31wAMPaNSoUTp//rx9+/vvvy9fX18tXbrU3paWlqawsDAFBwfr/PnzCgkJUVxcnE6fPu3wPm5l6dKl6tSpkwICAtSjRw8dPXo0W5/09HTNnTtX3bp1U6NGjRQQEKBu3brp008/tfeJiIjQmjVr7O8xa05dbq9rAID5uJ0SKEYMw9CRI0fUuHFjubq63rJfzZo1//Q4H3zwgebMmaNHH31UPXv2VEpKitauXauoqChVrFhRjz76qCTppZde0v/+9z/169dPlSpV0jfffKN58+YpOTlZEyZMkCS99dZb2rBhg/r16ydvb2/9+OOPWrp0qU6ePKn58+dnO3eHDh3k5eWld999Vx06dFCHDh1Uu3ZtnT592qFfZmamBg8erK+//lo9e/ZU7dq1dfjwYc2ZM0dHjhzR7NmzHeb8RUZGKiIiQl5eXmrUqNEt3/vQoUO1e/duhYWFacCAATp06JCioqL0+++/KzIyUmlpaRowYIC+/fZbhYeHy8/PT4cOHdKHH36ogwcPavHixQ6f/aBBg9SkSRONGjVKmzdv1vz583Xs2DEdPXpUTz31lAzD0OzZszVs2DBt2rRJLi4u9u9ywIABqlu3rl599VXt27dP0dHROnv2rN566y1JN2457NOnj2rUqKFBgwapZMmS+vrrr7Vu3TqdP3/eYZ5XRkaG3njjDQ0cOFBpaWlq0qSJvv32W4f3HhcXp+HDh6t169Z6/PHHde3aNS1ZskQDBgzQxo0b7bdeDh8+XF988YU6duyofv36KT4+XsuXL9dXX32lVatW6a677rIf84033lClSpU0ZMgQXbx4UR999JGeeeYZ7dixw/5eb1a7dm1NmTJF7777ru6++249++yz8vX1zfN533rrLT3yyCN6/PHHVbVq1Ry/6z+ep3HjxpJu/OPBjBkz1KlTJ/Xs2VPnzp3TkiVLFBcXp08//VTlypXTCy+8oG3btumDDz5Qp06dVKFCBc2cOVPHjh3TtGnTVKlSJY0ePVpRUVG6cOGCIiMj7e8jJzNmzFBMTIxatWqlfv366dChQw63EmeJjIxUbGysevfurYiICF24cEGffPKJXn/9dd17771q1qyZnn32WdlsNh04cEBTpkzRvffeKyn31zUAoAgwABQbv//+u+Hj42MMHz48T/u1a9fO6Nu3r2EYhpGWlmY0btw42zEuX75s+Pn5GYMHDzYMwzB+++03w8fHx/joo48c+o0aNcp46qmn7K8DAgKM8ePHO/SZNm2aER4ebly5csUwDMOYPn264ePjYyQmJhqGYRiJiYmGj4+PMX36dPs+n332meHj42N89dVXDq93797tcOwVK1YYPj4+xpYtWxz6jRw58i8/h507dxo+Pj7G7NmzHdpfeeUVo2HDhkZycrKxbNkyw8fHx1iwYIFDnw8//NDw8fExli5d6nDeF154wd7n0qVLRsOGDY169eoZx44ds7dPnTrV8PHxMeLj4x0+jyFDhhg2m82hDl9fX+Onn34yDMMwxowZYwQGBhoXLlxwqGX48OGGj4+PvT3reDd/njl9pmPHjjWCgoIczvnDDz8YHTt2NGJjYw3DMIxdu3YZPj4+xsSJEx2OtWnTJsPHx8eYMmWKYRiG8dVXXxk+Pj7GY489ZmRkZNj7zZ071/Dx8TH27NmT7fO/2c0/k7dz3pv3zct5EhISjHr16hnvv/++Q7+jR48aDRs2NN5++21729dff23Uq1fPePXVV43Dhw8bDRo0yHbd9O3b12jXrt2f1vD7778bfn5+2b7vrO8t62f3/Pnzhq+vb7bafv75Z8PHx8eYMGGCvW3kyJGGj4+P/XVur2sAQNHA7ZRAMVKixI1L/u8spODq6qq9e/faR3uyXLhwQZ6enrp69aokycvLS6VKldKyZcv0xRdf2NvfffddLVy40L5flSpVtGnTJq1evVqXLl2SdGME77PPPvtbc542b96scuXKqWHDhkpKSrL/adOmjZydnbVz506H/i1btvzLY+7cuVMlSpRQ3759HdpHjhypdevWydPTU9u3b5enp2e2UZJ+/frJ09NT27Ztc2h/6KGH7P/t5eWlcuXKqWbNmqpbt669vXr16pJuLOhys0GDBjmMJg4YMECGYWjHjh2SpHHjxmn79u0qW7asvc+VK1fk7u4uSfbvJLefQZUqVZSSkqKJEyfq559/liT7raihoaGSpO3bt0uSBg8e7LBv586dVatWLW3dutWhvWPHjnJ2dra/rl+/fo7v9a/k9by5+b5zsmXLFtlsNoWEhDj8XFWoUEH169d3+LkKCgpSRESE1q9fr2HDhqlcuXIaM2ZMns+5b98+paWlqWfPng7f9x8f+1GxYkUdPHhQQ4YMsbcZhqGMjAxJUkpKyi3PkdvrGgBQNHA7JVCMlClTRq6urn/7mVqurq7auXOntm3bpvj4eJ08edI+h834/3Nn3Nzc9NZbb+nNN9/UsGHD5ObmpmbNmqljx47q3r27PUiMGzdOL730kiIjI/Xmm2+qUaNG6tChgx577DF5eXnddo0JCQlKSkpS8+bNc9x+5swZh9fly5f/y2OePn1a5cuXl6enp0N7xYoVVbFiRUnSqVOn5O3tne12VTc3N3l7e2e77bNChQoOr11cXLLVkhVy/vj8vj8u/V+jRg17nZLk5OSkCxcuaO7cuTp69KgSEhL0yy+/2L+jPx7vrz6Dvn37as+ePVqyZImWLFmi6tWrq127durRo4d9kZxTp07prrvuyva+surdvXu3Q9sfHw3g5uaWY21/5e+eN7cSEhIkSb169cpx+x+/96xbPE+dOqVp06Y5BOrcyvo+s257zFK2bNls35mbm5vWr1+vPXv26MSJEzp58qQ9vBl/Ma8tN9c1AKBoIMQBxYiTk5OCgoJ0+PBhZWRk5DjnSJKmTZumxMRERUZG2sNJFsMw9Oqrr2rDhg1q0qSJgoKC9MQTT6hp06Z66qmnHPqGhYWpVatW2rp1q3bt2qW9e/dqz549WrZsmVatWiU3Nzc1b95cO3bssP/58ssv7aN1q1evvu1ftjMzM1WzZk2NHTs2x+03z4+S/m+U8q+O+VfPzvuzX3ZtNlu2X/JvHoXKktvn8/3xWFnBJ+uYO3fu1JAhQ1SpUiU98MADat26tfz8/LRnzx7NnTs32/H+6jPw9PTUkiVL9O2332rr1q3avXu3Pv74Yy1dulRTpkxRWFhYnt9/bj733MiPzz03sj7j2bNny8PD4y/7nzx5Ur///rukG6PDXbp0yfM5s34ecnokwM1hNy0tTQMHDtTBgwcVHBys5s2bq3///mrWrJnatm37p+fIy3UNADAfIQ4oZjp06KC4uDht2rRJ3bp1y7b9+vXr+vTTT5WZmZnjqMGBAwe0YcMGDRkyRC+++KK9PSMjQ8nJyfbFLVJSUnTkyBHVrVtXPXr0UI8ePZSWlqb33ntPixcv1p49e9SyZUsdOXJEVapUUdeuXdW1a1fZbDYtWLBAU6ZM0caNG7PdMpZb1atX1+HDh/XAAw84BIX09HRt2bJFVapUyfMxq1atqr179yolJcXhVs/vv/9e8+fP13PPPadq1arp22+/VXp6ukNwSEtL06lTp3T//fff1vvJSWJiourUqWN/feLECUn/NyI3YcIE1ahRQ5999plKlSpl7/evf/3rts4XHx+vy5cvq1GjRmrUqJFGjBihn376SU8++aQWLFigsLAwVatWTXv27NFvv/2WbVQsPj5e99xzz22d+68U1nmznhl3zz332G/9zLJr1y6HUdqMjAyNHj1aZcuWVffu3fXhhx+qS5cu6tixY57OmXVNnThxwuGxIFeuXNGFCxfsrzdt2qS4uDi9/fbb6tGjh7393Llzf3mO3F7XAICigTlxQDHzxBNPqFq1apo8ebKOHTvmsC0zM1Pjxo3Tb7/9pmeeeSbHFSyTk5MlySE8SNInn3yia9eu2eff/Pjjj3ryyScdljZ3c3NTgwYNJN0YCUlOTtYTTzzhMCpUokQJ+fv72//7doWEhCg5OVnLly93aF+xYoWGDx+u//znP3k+Zps2bWSz2bRq1SqH9uXLlys2NlYVKlRQSEiIrly54rC0vHTj+XwpKSl/OSKSFzevLilJCxYskIuLi0JCQiTd+K6qVq3qEODOnDmjzZs3S8r73MiJEydqyJAhDnOr7rvvPt1111327yrr3H8c6du6davi4+Pz9f3frKDOW6JECYfRrnbt2tnPc/Po35EjR/Tcc89p0aJF9rZ//vOf+v777xUZGakXX3xRtWvX1vjx4+3XUE7Hz8mDDz6oUqVKadGiRfbrS1K2n7FbXZuLFy+WJId9s76vrHPn9roGABQNjMQBxYy7u7tiYmL09NNPq0ePHvZnZSUnJ+vzzz/XkSNHFBoaqgEDBuS4f1BQkDw9PfXuu+/ql19+0V133aV9+/Zp06ZNcnd3t/+CHxgYqPvvv1/Tpk3TmTNn5OvrqzNnzmjJkiW677771Lx5c7m5uSksLEzLli3TtWvXFBQUpOTkZC1ZskQVKlRQ586db/t9Pv7441qzZo0mTJig77//XgEBATp27JhWrlyphg0bKjw8PM/HDAkJUYsWLTRp0iT9+OOP8vf31zfffKO1a9dq6NChKlu2rP28kyZN0rFjx+Tn56fDhw9r9erVCgwM1OOPP37b7+mP1qxZoytXrqhx48b697//rR07dmjo0KH20aLWrVtr06ZNGjNmjPz9/XXq1Cn7L+XSny90kZMBAwbomWee0ZNPPmmf17h161YlJCRo8uTJkm4E3fbt22vx4sU6d+6cgoODdeLECS1fvlze3t7ZFh7JLwV13nLlyumHH37QsmXL1KxZM/n4+CgiIkIff/yxkpOT9dBDD9l/ZkuXLm0fxfr5558VExOjli1b2p/JOHbsWPXr109vv/223nvvPfvx9+/frwULFqhx48YKDAzMVoOnp6deffVVjR8/Xk899ZQ6d+6sH3/8UevXr1fJkiXt/R588EG5uLjotdde05NPPikXFxft2LFDe/bskaurq8P3nXWb8vTp0xUcHJzr6xoAUDQQ4oBiqEGDBlq3bp0WLlyo3bt3a9OmTTIMQ76+vnrnnXcUHh5+y3lZFSpU0Lx58/T+++9r1qxZcnNzU61atTR16lQdOnRIixcvtt/SNnPmTMXExGjHjh1auXKlypQpo44dO+rFF1+0L2AxYcIEeXt7a+PGjdq4caNKliyp5s2ba/jw4bc9H066Meq3cOFCzZw5U1988YXWr1+vSpUqqXfv3ho6dKjDL7+5VaJECc2aNUuzZs3Sv/71L61fv1733nuvxowZo969e2c7b2xsrNavX68qVapo8ODBeu655/70+Xx5FRMTo5kzZ2rz5s3y9vbWhAkT1LNnT/v2cePGqVSpUtq+fbvWrVunKlWqqHv37urQoYN69+6tr776yj4ymhstW7bU7NmzNXfuXM2aNUupqamqW7eupk6dag8qTk5Oio6O1ocffqi1a9dq+/btKl++vJ544gm98MIL2eYi5peCOu8LL7ygsWPH6p133tHQoUNVp04dvf7667rvvvu0YsUKTZ48WV5eXrr//vvto202m02vv/66nJycHOZkBgcH65FHHtG6devUtWtXtW3bVv/4xz909OhRRUVFKTw8PMcQJ0l9+vSRl5eX5s2bp8mTJ6tmzZqaNWuWRo4cae/j4+Oj6dOnKyYmRlOnTlXp0qVVt25dLViwQMuWLVNcXJz9Nt+s7/+jjz7Sf//7X/3zn//M9XUNADCfk8GSUwBgKVkPft62bZv98QMAAKD4YE4cAAAAAFgIIQ4AAAAALIQQBwAAAAAWwpw4AAAAALCQYrk6pc1mU0pKilxdXW+5Ah8AAACKL8MwlJ6ertKlS/+t55YCBaFYhriUlJRsDzkGAAAA/sjHx0deXl6m1nDp0iWdP39e6enpptaBwuHi4iIPDw9VrFhRHh4eOfcp5JqKhKznNPn4+NifVQUAAABkSUtL07Fjx/L1+Z6349KlSzp37pyqVaumkiVLchfZHc4wDGVkZOjKlStKSEhQ5cqVVaZMmWz9imWIy/rhd3Nzk7u7u8nVAAAAoKgyOzSdP39e1apVU6lSpUytA4XDyclJrq6uuvvuu+Xu7q6zZ8/mGOK4wRcAAAAootLT01WyZEmzy4AJSpYsqdTU1By3EeIAAACAIszs0UCY48++d0IcAAAAAFgIIQ4AAAAALKRYLmwCAAAAWF16ZqZcnZ0tWUNERITi4uIc2lxdXVW5cmV16NBBL7300i2X188vERERcnZ21sKFCwukf0EixAEAAAAW5OrsrFdiF5taQ1Tnfre9r7+/v9544w3769TUVO3fv18zZ87UuXPnNG3atPwo8ZbGjh2bp/mGee1fkAhxAAAAAAqdp6enGjVq5NAWHByss2fP6tNPP1VkZKQqVapUYOevU6dOgfYvSMyJAwAAAFBkNGjQQIZh6MyZMwoJCdGkSZMUERGhxo0b691335UkXbhwQW+88YaaN2+ugIAA9e7dWwcPHnQ4Tlpamj744AOFhIQoMDBQYWFh2rRpk317RESE+vfvb3/95ZdfqmfPngoKClLTpk01ZMgQ/fzzz7fsf/36dUVHR6tTp07y9/dXly5dtHLlSocaQkJCFBMTo0mTJunBBx9UYGCgBg4cqJMnT/6tz4gQBwAAAKDIOHHihCTJ29tbkvTxxx/Lz89P0dHR6tq1q1JTU9W/f3/t3LlTL7/8sqZPn64yZcqof//+OnTokP04I0aM0MKFC9WrVy/NmTNHTZs21csvv6wdO3ZkO2diYqKGDBkiPz8/zZ49WxMnTtTx48c1ePBgGYaRrb9hGHrmmWe0aNEi9e7dW7Nnz9aDDz6osWPHaubMmQ59Fy5cqPj4eL377ruaMGGCDh8+rMjIyL/1GXE7JQAAAIBCZxiGMjIy7K8vXLig3bt3a8WKFQoNDVW5cuUkSVWqVNFrr71mn4/2ySef6OjRo1q1apX8/f0lSa1bt1aPHj00bdo0LViwQMeOHdMXX3yhMWPG6Mknn5QkNW/eXAkJCdq3b5/atWvnUMuhQ4d0/fp1DR48WJUrV5Yk3XPPPdq2bZtSUlLk6enp0H/Xrl2Ki4tTdHS0QkNDJUktW7ZURkaG5syZoz59+ujuu++WJJUtW1azZs2S8/9fACYhIUEzZszQ5cuX5eXldVufHSEOAAAAQKH76quv1LBhQ4c2Z2dnPfTQQxo3bpy9rW7dug4LivznP/9R5cqVVb9+fYcQ2K5dO82dO1dpaWn2Wys7dOjgcPyPPvoox1oCAwPl7u6uHj16KDQ0VK1bt1ZwcLACAgJy7L9//365urqqY8eODu1hYWFavny5vvvuO7Vt29Z+bOebVvCsUqWKJOnq1auEOAAAAADWERAQoDFjxkiSnJyc5OHhoWrVqqlkyZIO/cqXL+/wOjk5WWfPns0WALNcuHBBycnJOe57K9WrV9eSJUs0b948ffrpp1q8eLHuuusu9enTRy+99FK2VSkvXryo8uXLq0QJx9lpFSpUkCRdvnzZ3vbHRyVk7ZPTbZq5RYgDAAAAUOhKly5tvx0yL7y8vFS7dm1Nnjw5x+133323fYQrKSlJFStWtG87duyYrl27psDAwGz7BQQEKCYmxj6St3LlSs2ZM0cNGjRQp06dHPredddd+v3332Wz2RyC3K+//mqvoSAVmYVNjhw5ooYNG+rs2bN/2i8lJUXjx49XixYtFBQUpGeeecY++REAAADAna1p06b65ZdfVKlSJfn7+9v/bNu2TR9//LFcXV3VpEkTScq2iMnbb7+tqVOnZjvmxx9/rJCQEKWlpcnNzU3NmzfXhAkTJElnzpzJ1r9Zs2ZKT0/X5s2bHdo3bNggV1fXW96GmV+KxEhc1sovN9/TeivDhw/Xf//7X7322msqXbq0YmJi1K9fP23cuPG27ykFAAAAYA3h4eFasmSJBgwYYF+IZOfOnVqwYIGef/55OTk5qX79+urYsaPeffddXb16Vb6+vtq6davi4uL0z3/+M9sxH3jgAU2ZMkVDhw5V37595ezsrBUrVsjd3T3bIijSjYVUmjZtqtdff11nz55V3bp1tWvXLq1YsULPPfec7rrrrgL9DEwNcRkZGVq5cqWioqLk6ur6l/0PHDigXbt26cMPP1Tr1q0lSffff7/at2+v5cuXa9CgQQVdMgAAAFAkpGdmKqpzP9NrcL1p0Y7CULp0aS1dulRRUVGaNGmSUlJS5O3trTfffFN9+/a194uKilJ0dLTmz5+vixcvqnbt2vZHAfxR3bp1NXfuXM2YMUMvv/yyMjMz5efnp/nz56tGjRrZ+pcoUUJz587VBx98oI8++kgXL15UzZo1NW7cOPXq1atA378kORl/Z0bd37Rv3z49++yzGjhwoCpXrqw33nhDu3btsq/Y8kfR0dFatGiR9u/f77DCy6BBg3T16lUtWbIkV+dNTU3V4cOH5efnJ3d393x5LwAAALhzFJXfF48cOaL69eubdn6Y61bfv6lz4mrXrq2tW7fq+eefdwhlt3L8+HHVqFEjW997771X8fHxBVUmAAAAABQZpt5OmbUEZ25duXIl24P2pBtDqleuXMnz+Q8fPpznfQAAwJ2hUWADOZVw+uuOFmfYDH373f/MLgNAPioSC5vk1p/d+fnHZzTkhtnD4wAAwDw223Wl/LTG7DIKXOk6j9pX6kPuZd1OCRRFReYRA7nh6emplJSUbO0pKSk5jtABAAAAwJ3GUiGuVq1aSkxMzDYid/LkSdWqVcukqgAAAACg8FgqxLVs2VKXLl3S3r177W1JSUk6cOBAjkuFAgAAAMCdpkiHuKSkJH377bf2RUuaNm2qZs2a6eWXX9aqVau0ZcsW9e/fX15eXurdu7fJ1QIAAABAwSvSC5vs3LlTkZGRWrx4sYKDgyVJMTExmjRpkqZMmSKbzaYmTZrogw8+UJkyZUyutvgwjHQZRqbZZRQ4JydnOTn99UPoAQAAgMJk6sO+zVJUHt5oVTbbdaWnHzW7jALn6uqrEiU8zC4DAFBAitPqlPx9lndF5fdFHvZdvBXJh30DAAAAAPKGEAcAAABYkJGRbnYJt11DRESEfH19Hf7Uq1dPjRs3Vnh4uNatW5fPlf611atXy9fXV2fPnrXX2L9//0KvIzeK9Jw4AAAAADlzcnHV+dmvmVpDpeem3Pa+/v7+euONN+yvbTabzp49q0WLFum1115T2bJl1aZNm/wo845DiAMAAABQ6Dw9PdWoUaNs7a1bt1bz5s21evVqQtwtcDslAAAAgCLDzc1Nrq6ucnJyknRjhG7OnDl66KGH5Ofnp9DQUK1atSrbfmvXrlX37t0VGBiokJAQTZ8+XZmZ/7ei+hdffKHevXsrKChIfn5+6ty5s5YtW1Zo7ys/MRIHAAAAoNAZhqGMjAy081DMAAAgAElEQVT768zMTJ06dUqzZs1SSkqKHnnkEUnSuHHjtHr1aj333HMKDAzUl19+qTfffFPXr19XRESEJGnp0qV666239MQTT2jEiBE6fvy43nvvPV2/fl2vvfaatm3bpmHDhql///4aNmyYrl+/rmXLlmn8+PHy8/NTQECAKZ/B7SLEAQAAACh0X331lRo2bOjQ5uTkJF9fX0VHR6tdu3aKj4/XJ598otdee01PP/20JKlly5bKzMxUdHS0evToIXd3d82cOVOhoaF666237H0uXbqkL7/8UoZh6Oeff1Z4eLgiIyPt5woKClJwcLDi4uIIcQAAAADwVwICAjRmzBhJ0rlz5xQdHa2MjAxNmzZN9913n6QbQc8wDLVr185h1C4kJESLFi3SoUOHVKFCBf3+++/q0KGDw/Gff/55Pf/885KkQYMGSZJSUlIUHx+vhIQE/fe//5Ukpaebv8pnXhHiAAAAABS60qVLy9/fX9KNlSobNWqkbt26aeDAgfrss89Urlw5JScnS5JCQ0NzPMb58+fl4nIj0pQvX/6W50pKStLYsWO1detWOTk5qUaNGmrSpImkG7d1Wg0hDgAAAIDpKlSooDFjxujFF1/U22+/raioKHl5eUmSlixZIg8Pj2z7VK9eXb/++qukG0HtZr/99pt+/PFHNW7cWCNGjFB8fLwWLlyooKAgubm56dq1azkukGIFrE4JAAAAoEgIDQ1Vq1attGHDBsXFxen++++XJF28eFH+/v72P2fOnNH06dN17do13XfffSpbtqy2b9/ucKyVK1dqyJAhkqSDBw8qNDRUwcHBcnNzkyTt3r1b0o3VL62GkTgAAAAARcbo0aPVrVs3TZw4UWvWrNHDDz+s0aNHKzExUfXr19dPP/2kqVOnqmHDhqpataqkG/Pf3n77bd19990KCQnRsWPHNG/ePA0cOFDu7u4KCAjQ+vXrVb9+fVWuXFlff/215s2bJycnJ127ds3kd5x3hDgAAAAARcZ9992niIgIzZ8/X8uXL9ekSZM0Z84cLVmyROfOnVOFChXUo0cPDRs2zL5PRESESpYsqfnz52vFihWqWrWqhg0bpgEDBkiSJk2apAkTJthXr6xZs6bGjx+v9evX6+DBg6a8z7/DybDiTL6/KTU1VYcPH5afn5/c3d3NLsdybLbrSk8/anYZBc7V1VclSmS/9xoAcGew2a4r5ac1ZpdR4ErXeZS/z25DUfl98ciRI6pfv36O24yMdDm5uBZyRUWvhjvZrb5/5sQBAAAAFlQUwlNRqKE4IsQBAAAAgIUQ4gAAAADAQghxAAAAAGAhhDgAKIKMjHSzSygUxeV9AgCQn3jEAAAUQU4urjo/+zWzyyhwlZ6bYnYJAABYDiNxAAAAAGAhhDgAAAAAsBBCHAAAAABYCCEOAAAAACyEEAcAAABYUEamzewSbruGUaNGydfX95Z/9u7d69D/yJEjatiwoc6ePZsfZVseq1MCAAAAFuTiXEKzluwxtYYhfVve9r5VqlRRdHR0jtvq1Klj/+/jx49r8ODBysjIuO1z3WkIcQAAAAAKnZubmxo1anTL7RkZGVq5cqWioqLk6upaiJUVfdxOCQAAAKDIOXjwoN5//309/fTTGjFihNnlFCmMxAEAAAAwRU63SDo7O8vJyUm1a9fW1q1bVb58ea1evdqE6oouQhwAAACAQpeQkKCGDRtmax83bpx69+6tChUqmFCVNRDiAAAAABS6KlWqKCYmJlt7tWrVTKjGWghxAAAAAAqdm5ub/P39zS7DkljYBAAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFsLAJAAAAYEEZmTYN6dvS9BpcnPM+LjRp0qQ89Q8PD1d4eHiez3OnYiQOAAAAsKDbCU93Yg3FEZ86AAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDAAAAAAshxAEAAABFmGEYZpcAE/zZ906IAwAAAIooFxcXZWRkmF0GTJCeni5nZ+cctxHiAAAAgCLKw8NDV65cMbsMmODSpUvy8vLKcRshDgAAACiiKlasqF9//VVXr17ltspiwDAMpaWl6bffftOFCxdUrly5HPu5FHJdAAAAAHLJw8NDlStX1tmzZ5Wammp2OSgEzs7O8vLy0r333it3d/cc+xDiAAAAgCKsTJkyKlOmjNlloAjhdkoAAAAAsBBCHAAAAABYCCEOAAAAACyEEAcAAAAAFkKIAwAAAAALIcQBAAAAgIUQ4gAAAADAQghxAAAAAGAhhDgAAAAAsBBCHAAAAABYCCEOAAAAACzExewCAADZGUa6yj72rNllFDjDSJeTk6vZZQAAYCmEOAAoggwjU6nJ+8wuo8C5VKhKiAMAII+4nRIAAAAALIQQBwAAAAAWQogDAAAAAAshxAEAAACAhRDiAAAAAMBCCHEAAAAAYCGEOAAAAACwEEIcAAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDAAAAAAsxPcRt2LBBXbt2VUBAgDp37qy1a9f+af+kpCRFRkaqZcuWatasmQYPHqwTJ04UTrEAAAAAYDJTQ1xsbKxGjBihFi1aaObMmWrWrJlGjhypzz//PMf+hmFo6NCh2r17t0aMGKEpU6bo119/Vb9+/XTx4sVCrh4AAAAACp+LmSefOnWqOnfurNGjR0uSWrVqpYsXLyo6OlqhoaHZ+p84cUJff/21Jk+erO7du0uSateurYceekjbt2/Xo48+Wqj1AwAAAEBhM20kLjExUQkJCerYsaNDe6dOnXT8+HElJiZm2yc1NVWSVLp0aXtbmTJlJEnJyckFWC0AAAAAFA2mhbjjx49LkmrVquXQXqNGDUlSfHx8tn3q1aun4OBgzZw5Uz///LOSkpI0ceJElSpVSg899FDBFw0AAAAAJjPtdsrLly9Lkjw9PR3as0bZrly5kuN+48aN0z/+8Q916dJFkuTm5qaZM2fK29u7AKsFAAAAgKLBtBBnGIYkycnJKcf2EiWyDxL+/PPP6tWrl+69916NHj1aHh4e+uSTTzRs2DB99NFHuv/++/NUw+HDh2+z+uLN17eWrl69829fLVXqqo4e/d7sMlBMBQU1NLuEQnPw4EGzS0AxxXUGwKpMC3FeXl6Sso+4paSkOGy/2cKFCyVJ8+fPt8+Fa9Gihfr06aN33nlHq1evzlMNfn5+cnd3z2vpxZ7Ndl2urmXNLqPAubqWUpMmTcwuA8WUzXbd7BIKDdcZzMJ1hj+TmprKP/ijyDJtTlzWXLiEhASH9pMnTzpsv9kvv/yi2rVr2wOcdGMkr0mTJvrpp58KsFoAAAAAKBpMC3E1atRQ9erVsz0TbvPmzapZs6aqVq2abZ9atWrpxx9/zPZMuO+++07VqlUr0HoBAAAAoCgw9TlxQ4cOVWRkpMqUKaO2bdtq+/btio2N1bRp0yRJSUlJSkhIUJ06deTp6an+/ftr/fr1GjhwoAYNGiQPDw+tW7dOcXFx9n0AAAAA4E5maogLDw9XWlqa5s+fr1WrVsnb21uTJ0+2rzy5c+dORUZGavHixQoODlb16tW1fPlyvffeexo1apRKlCghHx8fLViwQA8++KCZbwUAAAAACoWTkbUcZDGSNVGVhU1uj812XenpR80uo8C5uvqqRAkPs8tAMWWzXVfKT2vMLqPAla7zKNcZTMN1hj/D74soykybEwcAAAAAyDtCHAAAAABYCCEOAAAAACyEEAcAAAAAFkKIAwAAAAALIcQBAAAAgIUQ4gAAAADAQghxAAAAAGAhhDgAAAAAsBBCHAAAAABYCCEOAAAAACyEEAcAAAAAFkKIAwAAAAALIcQBAAAAgIUQ4gAAKGAZmTazSwAA3EFczC4AAFC8zVqyx+wSCtyQvi3NLgEAcAdhJA4AAAAALIQQl4/SMzPNLqFQGGYXAAAAABRj3E6Zj1ydnfVK7GKzyyhw73XqaXYJAAAAQLHFSBwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHFHMZmTazSwAAAEAeuJhdAKzIMLuAQjNryR6zSyhwQ/q2NLsEAAAA5AEhDrfBSWkn/2d2EQXOtY6v2SUAAAAA2XA7JQAAACyP6QEoThiJAwAAuINlZJTQ+d8vmF1GgatcwUtyZnwCxQMhDgAA4A5ms0nffpNgdhkFrkP7BmaXABQaQhxwCxkZJdStfUOzyyhw6ekZcnXlfwUAAABWwW9uwC3wL5cAAAAoirhxGAAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEJczC4AAFB8ZWSUULf2Dc0uo8Clp2fI1ZW/cgEA+YO/UQAAprHZpG+/STC7jALXoX0Ds0sAANxBuJ0SAAAAACyEEAcAAAAAFkKIAwAAAAALIcQBAAAAgIUQ4gAAAADAQghxAAAAAGAhhDgAAAAAsBBCHAAAAABYCCEOAAAAACyEEAcAAAAAFkKIAwAAAAALIcQBsJT0zEyzSwAAADCVi9kFAEBeuDo765XYxWaXUeDe69TT7BIAAEARxUgcAAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDAAAAAAshxAEAAACAhRDiAAAAAMBCCHEAAAAAYCGEOAAAAACwENND3IYNG9S1a1cFBASoc+fOWrt27Z/2t9lsmj17ttq3b6+AgACFhYVp48aNhVQtAAAAAJjLxcyTx8bGasSIEerXr59atWqlrVu3auTIkfLw8FBoaGiO+7zzzjtauXKlXn75ZdWrV08bN27UK6+8Ik9PT7Vp06aQ3wEAAAAAFC5TQ9zUqVPVuXNnjR49WpLUqlUrXbx4UdHR0TmGuISEBC1dulRvvfWWHn/8cUlS8+bNdeLECf373/8mxAEAAAC445kW4hITE5WQkKCXX37Zob1Tp06KjY1VYmKivL29HbZt3bpVHh4e6t69u0P7kiVLCrxeAAAAACgKTJsTd/z4cUlSrVq1HNpr1KghSYqPj8+2z9GjR1WrVi3t3btX3bp1U4MGDdSxY0dt2rSp4AsGAAAAgCLAtJG4y5cvS5I8PT0d2kuXLi1JunLlSrZ9kpKSdObMGY0ePVovvviiqlevrlWrVmn48OEqV66cHnjggTzVcPjw4dusPmdNmjTJ1+MBheXgwYNml5BrXGewKitdZ8VFUFBDs0tAPuM6Q3FhWogzDEOS5OTklGN7iRLZBwnT09OVlJSkOXPmqF27dpJuzIk7fvy4YmJi8hzi/Pz85O7ufjvlA3cUghFQ8LjOih6b7brZJSCf5ed1lpqamu//4A/kF9Nup/Ty8pKUfcQtJSXFYfvNSpcuLWdnZ7Vo0cLe5uTkpAcffFBHjx4twGoBAAAAoGgwLcRlzYVLSEhwaD958qTD9pvVqFFDNptNGRkZDu3p6enZRvQAAAAA4E5kWoirUaOGqlevrs8//9yhffPmzapZs6aqVq2abZ9WrVrJMAzFxsba2zIyMvTvf/+b21QAAAAAFAumPidu6NChioyMVJkyZdS2bVtt375dsbGxmjZtmqQbC5kkJCSoTp068vT0VPPmzdWmTRtNnDhRV69eVc2aNbVs2TKdPn1aUVFRZr4VAAAAACgUpoa48PBwpaWlaf78+Vq1apW8vb01efJkdenSRZK0c+dORUZGavHixQoODpYkTZ8+XdHR0Zo3b54uXryoBg0aaP78+fLz8zPzrQAAAABAoTA1xElSr1691KtXrxy3hYeHKzw83KHNw8NDI0eO1MiRIwujPAAAAAAoUkybEwcAAAAAyDtCHAAAAABYCCEOAAAAACyEEAcAAAAAFkKIAwAAAAALua0Qd+7cOX333Xe6fPmy0tLSZLPZ8rsuAAAAAEAO8hTiDh48qPDwcLVt21a9evXS4cOHFRcXp7Zt22rTpk0FVSMAAAAA4P/LdYg7dOiQBgwYoJSUFD311FP29jJlysjFxUUjRozQrl27CqRIAAAAAMANuQ5x0dHRql69utatW6dBgwbJMAxJkr+/v9avX6/atWtr7ty5BVYoAAAAACAPIe6bb75ReHi4PDw85OTk5LDN09NTPXv21I8//pjvBQIAAAAA/k+e5sS5ubndcltqaioLnAAAAABAAct1iAsMDNSGDRty3Hb16lWtWrVK/v7++VYYAAAAACC7XIe4YcOG6X//+5/69u2rtWvXysnJSYcOHdLixYv1yCOP6NSpU3r22WcLslYAAAAAKPZcctsxKChIc+fO1dixYzV58mRJ0rRp0yRJFStW1LRp0/TAAw8UTJUAAAAAAEl5CHGS1KJFC23ZskXff/+9EhMTZbPZVK1aNfn5+cnFJU+HAoDbcj0jXYOadjC7jAJnyDC7BAAAUETlOXk5OTnJz89Pfn5+BVEPAPypDJtN/zr6ndllFLiXHmhjdgkoxq6lpynTuPP/IaGky53/HgHcmXId4vr165erfosXL77tYqyOEQIAwJ0g0zA0Z/8Os8socPxjCQCrynWIO3XqVLY2m82mCxcuKDU1VdWqVVPdunXztTirYYQAAAAAQEHLdYjbvn17ju2ZmZnatm2b3njjDQ0cODDfCgMAAAAAZJenh33nxNnZWR07dtTjjz+u999/Pz9qAgAAAADcwt8OcVlq1qypH374Ib8OBwAAAADIQb6EuLS0NK1fv17ly5fPj8MBAAAAAG7hb69OmZaWpvj4eF26dEkvvPBCvhUGAAAAAMjub61OKd2YE3fffffp4YcfVp8+ffKtMAAAAABAdn97dUoAAAAAQOHJt4VNAAAAAAAF75YjcbeaA/dnnJyctGjRor9VEAAAAADg1m4Z4m41Bw4AAAAAYJ5bhjjmwAEAAABA0ZOvc+KSkpLy83AAAAAAgD/I9eqUkrR27Vpt3rxZV69elc1ms7dnZmYqJSVFP/30kw4fPpzvRQIAAAAAbsh1iPvwww81depUubq6ytPTUxcuXFCVKlWUnJysa9euycPDQxEREQVZKwAAAAAUe7m+nXL16tWqV6+e9u7dq5UrV8owDC1evFgHDhzQmDFjlJqaqsDAwIKsFQAAAACKvVyHuNOnT+uRRx6Rp6envL29VaZMGR04cEDOzs7q06ePunTpwuMFAAAAAKCA5TrEubi4qHTp0vbXNWrU0NGjR+2vg4ODdeLEiXwtDgAAAADgKNchrnbt2vrmm2/sr2vVquWwiMmlS5eUlpaWv9UBAAAAABzkOsSFh4dr9erVGjFihK5evaqQkBAdOHBAMTEx2rRpkxYuXKh69eoVZK0AAAAAUOzlenXK3r176+zZs1q6dKlcXFzUsWNHde3aVTExMZIkT09PjRgxosAKBQAAAADkIcS9/vrrCgsL0/PPPy9XV1dJUlRUlHr37q3k5GQFBQWpfPnyBVYoAAAAACAPIW7jxo1avXq1KlasqIcfflhhYWGqX7++7r///oKsDwAAAABwk1zPifvqq68UFRWlgIAALV26VOHh4erSpYvmzJmjxMTEgqwRAAAAAPD/5XokzsPDQ126dFGXLl105coVbd26VbGxsZo5c6aio6MVGBiosLAwPfnkkwVZLwAAAAAUa7keibuZp6enunfvrrlz52rTpk1q06aNvv32W02cODG/6wMAAAAA3CTXI3E3S0pK0pYtWxQbG6sDBw4oMzNTTZs2Vbdu3fK7PgAAAADATXId4i5cuKDNmzfr888/1/79+5WRkSFfX1+99NJLevjhh1WlSpWCrBMAAAAAoDyEuJYtW8pms+mee+7RwIEDFRYWpjp16hRkbQAAAACAP8h1iHv88ccVFhamJk2aFGQ9AAAAAIA/kesQN27cuAIsAwAAAACQG7e1OiUAAAAAwByEOAAAAACwEEIcAAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDAAAAAAshxAEAAACAhRDiAAAAAMBCCHEAAAAAYCGEOAAAAACwEEIcAAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDAAAAAAshxAEAAACAhRDiAAAAAMBCCHEAAAAAYCGEOAAAAACwEEIcAAAAAFgIIQ4AAAAALIQQBwAAAAAWQogDAAAAAAshxAEAAACAhRDiAAAAAMBCTA9xGzZsUNeuXRUQEKDOnTtr7dq1ud73zJkzatKkiWbNmlWAFQIAAABA0WFqiIuNjdWIESPUokULzZw5U82aNdPIkSP1+eef/+W+hmFo9OjRunLlSiFUCgAAAABFg4uZJ586dao6d+6s0aNHS5JatWqlixcvKjo6WqGhoX+677Jly3T8+PHCKBMAAAAAigzTRuISExOVkJCgjh07OrR36tRJx48fV2Ji4p/u+/7772vChAkFXSYAAAAAFCmmhbisUbRatWo5tNeoUUOSFB8fn+N+NptNo0aNUufOndW6deuCLRIAAAAAihjTbqe8fPmyJMnT09OhvXTp0pJ0y7luixYtUmJioubMmVOwBQIAAABAEWRaiDMMQ5Lk5OSUY3uJEtkHCY8fP64PPvhA06dPl5eX19+u4fDhw3/7GDfz9ffL1+MBheXgwYNml5BrXGewKq4zoOBZ6ToD/g7TQlxWCPvjiFtKSorD9iyZmZkaNWqUQkND1aJFC2VkZNi32Ww2ZWRkyMUlb2/Hz89P7u7ut1N+jq6kpebbsYDC1KRJE7NLyDWuM1gV1xlQ8PLzOktNTc33f/AH8otpc+Ky5sIlJCQ4tJ88edJhe5YzZ87ou+++09q1a9WwYUP7H0maMWOG/b8BAAAA4E5m2khcjRo1VL16dX3++efq0KGDvX3z5s2qWbOmqlat6tC/UqVK+vTTT7Mdp0ePHurdu7cee+yxAq8ZAAAAAMxm6nPihg4dqsjISJUpU0Zt27bV9u3bFRsbq2nTpkmSkpKSlJCQoDp16sjT01P+/v45HqdSpUq33AYAAAAAdxLTbqeUpPDwcI0fP1579uzR0KFDFRcXp8mTJ6tLly6SpJ07d+qJJ57Q999/b2aZAAAAAFBkmDoSJ0m9evVSr169ctwWHh6u8PDwP93/6NGjBVEWAAAAABRJpo7EAQAAAADyhhAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAAAACyHEAQAAAICFEOIAAAAAwEIIcQAAAABgIYQ4AAAAALAQQhwAAAAAWAghDgAAAAAshBAHAAAAABZCiAMAAACA/8fencfXdO3/H3+d5GQeJCEJiQgiYgxKxVCztqaWq1VuW62hptZQpaZrKp1UTVVDxdSqokpppSlFcQlFjUUpSQgxRkQSmc7J+f3RR86X0vtrixwn3s9/1M7ZO+s8Hl32fu+11mfZEYU4ERERERERO6IQJyIiIiIiYkcU4kREREREROyIQpyIiIiIiIgdUYgTERERERGxIwpxIiIiIiIidkQhTkRERERExI4oxImIiIiIiNgRhTgRERERERE7ohAnIiIiIiJiRxTiRERERERE7IhCnIiIiIiIiB1RiBMREREREbEjCnEiIiIiIiJ2RCFORERERETEjijEiYiIiIiI2BGFOBERERERETuiECciIiIiImJHFOJERERERETsiEKciIiIiIiIHVGIExERERERsSMKcSIiIiIiInbE5iFu3bp1tG3blsjISFq3bs2aNWv+5+cvX77M6NGjadasGbVq1aJjx47ExsYWUmtFRERERERsy2jLXx4bG8vQoUN56aWXaNSoERs3bmT48OG4urrSqlWr2z6fm5vLK6+8Qnp6OgMHDiQgIID169fz+uuvYzabadeunQ2+hYiIiIiISOGxaYibOnUqrVu3ZtSoUQA0atSItLQ0ZsyYcccQt23bNn799VdWrlxJZGQkAA0bNiQ5OZno6GiFOBERERERKfJsNp0yKSmJM2fO8MQTT9xy/MknnyQ+Pp6kpKTbzvHw8KBz585Ur179luPly5fnzJkz97W9IiIiIiIiDwKbjcTFx8cDUK5cuVuOh4aGApCQkEBISMgtP6tfvz7169e/5VheXh5bt24lPDz8PrZWRERERETkwWCzkbj09HQAPD09bznu4eEBQEZGxl+6zocffkhiYiK9e/e+tw0UERERERF5ANlsJM5isQBgMBjueNzB4X/nS4vFwuTJk1m8eDE9e/akZcuWf7sNv/zyy98+53+JqF7tnl5PpLD8/PPPtm7CX6Z+JvZK/Uzk/rOnfiZyN2wW4ry8vIDbR9wyMzNv+fmd5ObmMmLECGJiYujZsyfDhg37R22oVq0aLi4u/+jcO8nIzbln1xIpTLVr17Z1E/4y9TOxV+pnIvffvexnOTk59/yFv8i9YrMQV7AW7syZM0RERFiPnz59+paf/1FGRgZ9+vRh3759jBo1ipdffvn+N1ZEREREROQBYbM1caGhoZQuXZrvv//+luMbNmygbNmyBAUF3XaO2WymX79+HDx4kKlTpyrAiYiIiIjIQ8em+8S99tprjBw5kvfaaaoAACAASURBVGLFitG0aVM2b95MbGws06ZNA+Dq1aucOXOGChUq4OnpyfLly9m9ezedO3emVKlSHDhwwHotg8FAjRo1bPVVRERERERECoVNQ1zHjh3Jzc1l4cKFrFy5kpCQECZNmkSbNm0A2LJlCyNHjuSzzz4jKiqK9evXA7BixQpWrFhxy7UcHR05evRooX8HERERERGRwmTTEAfQpUsXunTpcsefdezYkY4dO1r//tlnnxVWs0RERERERB5INlsTJyIiIiIiIn+fQpyIiIiIiIgdUYgTERERERGxIwpxIiIiIiIidkQhTkRERERExI4oxImIiIiIiNgRhTgRERERERE7ohAnIiIiIiJiRxTiRERERERE7IhCnIiIiIiIiB1RiBMREREREbEjCnEiIiIiIiJ2RCFORERERETEjijEiYiIiIiI2BGFOBERERERETuiECciIiIiImJHFOJERERERETsiEKciIiIiIiIHVGIExERERERsSMKcSIiIiIiInZEIU5ERERERMSOKMSJiIiIiIjYEYU4ERERERERO6IQJyIiIiIiYkcU4kREREREROyIQpyIiIiIiIgdUYgTERERERGxIwpxIiIiIiIidkQhTkRERERExI4oxImIiIiIiNgRhTgRERERERE7ohAnIiIiIiJiRxTiRERERERE7IhCnIiIiIiIiB1RiBMREREREbEjCnEiIiIiIiJ2RCFORERERETEjijEiYiIiIiI2BGFOBERERERETuiECciIiIiImJHFOJERERERETsiEKciIiIiIiIHVGIExERERERsSMKcSIiIiIiInZEIU5ERERERMSOKMSJiIiIiIjYEYU4ERERERERO6IQJyIiIiIiYkcU4kREREREROyIQpyIiIiIiIgdUYgTERERERGxIwpxIiIiIiIidkQhTkRERERExI4oxImIiIiIiNgRhTgRERERERE7ohAnIiIiIiJiRxTiRERERERE7IhCnIiIiIiIiB1RiBMREREREbEjCnEiIiIiIiJ2RCFORERERETEjijEiYiIiIiI2BGFOBERERERETuiECciIiIiImJHFOJERERERETsiEKciIiIiIiIHVGIExERERERsSMKcSIiIiIiInZEIU5ERERERMSOKMSJiIiIiIjYEYU4ERERERERO6IQJyIiIiIiYkcU4kREREREROyIzUPcunXraNu2LZGRkbRu3Zo1a9b8z89nZmby1ltv0bBhQ2rVqkWvXr1ITEwsnMaKiIiIiIjYmE1DXGxsLEOHDqVhw4bMmjWLunXrMnz4cL7//vs/PWfw4MF8//33DB06lEmTJnHx4kVeeukl0tPTC7HlIiIiIiIitmG05S+fOnUqrVu3ZtSoUQA0atSItLQ0ZsyYQatWrW77/N69e9m6dSvR0dE0btwYgDp16tCiRQuWLVtG7969C7X9IiIiIiIihc1mI3FJSUmcOXOGJ5544pbjTz75JPHx8SQlJd12zo4dO/Dw8KBhw4bWY35+fjz66KNs27btvrdZRERERETE1mw2EhcfHw9AuXLlbjkeGhoKQEJCAiEhIbedExoaiqOj4y3Hy5QpQ2xs7F/+3RaLBYDc3Ny/3e7/JS8vF1eDzZcZ3ne5Obnk5dt0ELdQWHJzcXSw2LoZ993v/SDf1s34y9TPihb1sweT+lnRon52N9f7v+dGkQeJzf7lKljD5unpectxDw8PADIyMm47JyMj47bPF5xzp8//mby8PABOnDjxl8/5q+q7FL/n13zQHDt2Eihv62bcf78dJ8Df1o24/06c+NXWTfjb1M+KEPWzB5b6WRGifnZX8vLycHV1vS/XFvmnbBbiCt5qGAyGOx53cLj9DeD/ehNyp8//GQ8PDypWrIiTk9Ntv19ERERExGKxkJeXZx1gEHmQ2CzEeXl5AbePuGVmZt7y85t5enpy9uzZ245nZmbecYTuzzg4ONzx+iIiIiIiBTQCJw8qm014L1gLd+bMmVuOnz59+paf//GcpKSk20bkTp8+fcfPi4iIiIiIFDU2C3GhoaGULl36tj3hNmzYQNmyZQkKCrrtnMcee4zr168TFxdnPXb16lX27t1LgwYN7nubRUREREREbM1x/Pjx4231y728vJgzZw6pqakYDAYWLVrE119/zbhx4wgPD+fq1ascP34cT09PnJ2dCQ4OZvfu3XzxxRf4+PiQnJzMqFGjsFgsvPvuuxryFhERERGRIs9gsXHd1OXLl7Nw4ULOnz9PSEgIvXv3pkOHDgCsXr2akSNH8tlnnxEVFQVAWloa77//Phs3biQ/P5/atWszYsQIypd/CKpLiYiIiIjIQ8/mIU5ERERERET+uqK/k6eIiIiIiEgRohAnIiIiIiJiRxTiRERERERE7IhCnIiIiIj8T/n5+QBYLJbb9usVkcKnECfyB7o5iYiI3CojIwMAk8mEwWCwHi8IdyJSuFSdUh5qZrMZR0dHWzdDRP6Cgv6an5/PjRs3cHNzU/8VKQQ5OTn069cPV1dXzGYzTZs2JSQkhMcee8zWTRN5aCnEyUPLZDJhNBrJyspi/vz5HDlyhIiICAYNGoSDgwapRR4kN79wefvtt/npp58ICgpi9OjRhISEYLFYbhkdEJF7Jz8/n4SEBI4fP87p06dZuXIljo6OeHl5MWLECGrUqIGLi4utmynyUDHaugEitpCfn4/R+Pv//t27dyczMxNPT088PDzIzs7G3d3dxi0UkQIWi8Ua4Hr06EFycjJRUVEEBwfj4+MD/N6nNSoncu/t3buXiIgIwsLCCAsLA6BTp07s3r2bZcuW0aNHD1555RWef/55AgICbNxakYeHQpw8lApG2iZOnEhqairz5s3D398fd3d3Dhw4wPbt2wkICKBRo0aUKlXKxq0VebgVjLAtWrSIxMREFi5cSNmyZQGIi4tj9erVWCwWGjVqRIcOHWzYUpGiZe3atQwfPpw33niDDh06WENaiRIlaNOmDQ0bNuSzzz5j9uzZXL9+nREjRuDs7KyRcZFCoBAnD62MjAzOnj3L888/T2hoKGfPnmXBggXMnj0bNzc3bty4QadOnZg4caKtmyoiwOXLl6lSpQply5YlMTGR1atXM2/ePIKCgjCbzcTHx1OjRg3KlStn66aKFAlt2rThwIEDzJgxg/Pnz9O9e3fKlClj/XmxYsUYMGAA/v7+TJgwAWdnZ0aMGKEAJ1IIFOLkoeXi4kJaWhqbNm0iKCiIBQsWcODAAfr06cNTTz3FTz/9xKRJk+jXrx9BQUG2bq7IQ6dgHVxiYiK+vr7k5+dz8OBBZs+ezcaNGzlx4gR9+/alR48enDp1in//+9+cPXtWIU7kHnFycmLcuHGEhobywQcfcOHCBQYMGEDlypVvCWodO3YkKyuLBQsWULVqVZ566imNxoncZ6reIA8tJycnXnrpJc6cOcOAAQPIz8/nvffeY/DgwVSoUAE3NzfKlCmjdTYiNlCwDi4pKYmnn36azZs3M2jQIEJCQpg7dy4+Pj68/fbbvP7663h5eeHq6kq5cuVUlEjkPujWrRsfffQRe/fuZcyYMezatYvc3Fzrz52dnXniiSeoVKkSW7ZsAVCAE7nPVJ1SHgo3lyY3mUxkZGTg4eGBi4sLly5dIiEhgVq1auHs7AxAUlISY8eOxcnJiXnz5tm49SIPl/z8fBwcHLBYLKxdu5b//ve/DBo0yDqNKzEx0bomDiAhIYEpU6aQlJTE2rVrbdRqkaKjoA/+0ZEjRxg8eDA5OTkMGzaMJk2a4Onpaf35wYMH6dGjBx9++CHNmjUrzCaLPHQU4qTIu7k0+ejRozl69Ci//fYblStX5rHHHmPgwIEApKenM2zYMPLz80lLS+Pq1at8/fXXeHh4aD85ERvo378/Bw4coGbNmkydOtX6kgV+79fdunUjNTUVi8VCXl4eS5YsITAwUP1V5C7c3H+OHDnCsWPH8PX1JSAggOrVq5OSksKQIUM4ePAgAwcOpH379vj5+VnPHzt2LKGhofTs2dNWX0HkoeA4fvz48bZuhMj9VPA2sV+/fhw5coSWLVvSoUMHsrKyWLJkCcnJyTRr1ozExEQOHz7M5cuXqVWrFqNHj6ZEiRLW/eREpPCkp6cTHx/Pb7/9RlpaGg0aNKB48eLWdTZZWVmYTCaysrJo3rw5/fv3JyQkRAFO5C5YLBbrPXPkyJFER0fz3XffsX37dlavXk1aWhotW7akTZs2nDt3jsWLF5ORkUGFChXw9vYG4Pr163zzzTc8/fTTODg4aFqlyH2ikTh5KGzZsoXRo0czffp0IiMjcXZ25vPPP2fq1KlMmzaNvLw8HnvsMVxdXcnLy8PBwcE6/VJrbEQKz83FEFJSUtiyZQtTpkzB19eX2bNnExoa+qefVyEFkXtj/vz5zJ8/n7Fjx1K/fn0yMzMZPnw4P//8M9988w0VK1YE4OOPP+bjjz9m1apVVK1aFfg9xGVlZREYGGjLryBS5OnpVB4K2dnZmM1mAgMDcXZ2Zt++fbz77rv069ePkiVLMmPGDDZv3gz8XvCk4E2+ApxI4cjPzwfAZDJx7do1rl+/joeHB8888wzvv/8+JpOJ559/np07d1rP+WNoU4ATuTsWi4WMjAy2bt1K+/btad68Ob6+vmRnZ3Po0CHGjRvHb7/9xuzZs8nPz6d///58//331gAH4O3trQAnUgj0hCpFkslkuuXPnJwcUlNTcXBwID8/n379+vHss8/Sq1cvQkJCuHjxIhcuXLBlk0UeWiaTCQcHBxITExkyZAjt27enSZMm9O7dm9WrV9O4cWNmzJhB+fLl6d27N1999RWg0CZyrxkMBtzd3bl06RIArq6upKSk0LVrV9q0aUOnTp04cuQIa9eu5dq1awDWIkOa2CVSuBTipEgyGo2kpKQwaNAgzGYzLVu2JCwsjGHDhvHEE09QrVo1Ro0ahclk4uzZsxQrVuyWhdkiUjgsFgtGo5EbN27Qo0cPMjMzee6553jttdfIzMxk0qRJTJ06lUqVKvHuu+/Srl07Ro8ezbJly2zddJEip2AJQenSpdmzZw+5ubn06dOHChUqMGrUKIxGI97e3jg5Od12rl6qiBQuFTaRIuvUqVNMmzYNNzc3oqKiyMvLIzY2luzsbHr27En16tU5ceIE8+bN48qVK4wbN04FEUQKWcGD3+TJk0lLS2PKlCk0a9aMRx55hDp16nDt2jViY2Nxd3enYcOGREREEBgYSIcOHXB1dbVx60XsW35+/h2nJHt4eLBmzRoWLVqEyWRi+fLleHp6kpGRwRdffIGLiwudOnXSkgMRG1JhEymyMjMzGTt2LPHx8axcuRKj0cjKlStZuHAh6enpODk54eTkhMFgYN68eYSGhqqynYiNDBkyBIApU6ZgsVisVfIuXbpEv379cHR05MsvvwT+b7RA/VXkn7u5cFdMTAzHjx8nIiKCGjVqEBQUxOzZs1m+fDm+vr4MGTKE7Oxstm/fznfffcfKlSsJCwtT8S8RG1KIkyLhz6rSJScn8/TTT/P0008zduxYAI4dO0ZCQgJHjx7lkUceITw8XKXJRWzIbDbTp08fUlNTWbVqFfB7nzabzRiNRpYuXcqHH37Ipk2bNO1Z5B6bNGkSy5cvx2g0kp6eTt26dXnzzTepXr06q1atYt26dezZswcPDw8iIiLo168f9evX1z1TxMa0+ZUUKWfOnMHT09P6oBcUFESvXr1Yt24dhw4dIjIyksqVK1O5cmXatGlzy7m6GYnYhsFgoF69ekyfPp3Vq1fTqlUr3N3dMRqNmEwm0tPTKVeuHAaDQdsIiNwDBf1o//79fPXVV7zzzjvUqVOH48ePM3DgQIYOHcrw4cN55plnaNeuHZcvX8ZiseDr64unpycWi0X3TBEb05o4KRIMBgMnT56kdevW/PTTTxw7doyyZcvi4uJCpUqVWL58OZcvX+bxxx8HtJ+UiC3k5+dbp1/duHGDtLQ0UlJS8PHxoUaNGhw4cIAVK1bg7e2Nn58fxYoVY9++fcyZM4fatWvTunVr9VuRu1Rw/0tKSiIxMRGTyUSPHj3w8fEhNDSUNm3a8PXXXxMbG4unpydlypQhMDCQYsWKWZcgqB+K2J6mU0qRcfnyZXbs2MHmzZs5cuQIFy9eJCoqilatWuHl5cW4ceN46623aNWqla2bKvJQ+eNLk7S0NAYPHsyJEye4cuUKNWvWpHPnzjRp0oTp06fz5Zdf4u/vj6OjIw4ODgQFBfH5558DaA2OyF0o6ItZWVkMGDCAXbt2UbVqVVasWIHFYiEvLw9nZ2fy8vLo2bMnBw8epHPnzvTt21dTmUUeMApxYrduno+fk5OD2WzG3d2d7OxscnJyWL9+Pdu3b+enn34iLy8Pi8VC8+bNmTJlCqDROJHCMnToUAwGA5MnTwagffv2eHl50bJlSwIDA1mwYAFXrlyhcePGjB49mm3btnH48GHS09OpU6cOzZs3x9XVFZPJhNGoVQAi/8TN98ydO3cSHx/Ppk2b2L17N9OmTbPOVMnJycHFxQWA119/nePHjxMTE6OXJyIPGE2nFLuUn59vvRlNmjSJuXPnsnjxYooXL07lypVxdXWlatWqNG7cmJdeeglPT0/c3d3ZsGEDPj4+VK9eXQFOpBCkpqaybds2du3axc8//0zJkiU5cuSIdc/G8PBw2rZty4ULF4iNjcXJyYlnn32WqKgomjZtSnh4OEaj8ZY+LyJ/T0G1V4Bu3bpx/vx5+vTpQ3BwMBcuXGDevHl4e3tTo0YNjEYjOTk5GI1GWrVqRYcOHXBxcbltOwIRsS2FOLFLBTeS//znP3z//feEhobi5ubGggULcHFxoXbt2sDvm367uLhQq1Yt6tWrR1ZWFocPH6ZFixZ33KxURO4tNzc36tati8ViIS4ujk2bNnHy5ElefPFFfH19ycnJwc3NjSZNmnD06FFiYmJ47rnnrCMBBfTwKPLPFfSfw4cPExsbS9euXalQoQIhISFERESQnZ3N/PnzSUlJoWnTprcEOa2DE3kwaWxc7NaqVavYuHGjdSSue/fuBAUFMWXKFIYPH05WVhYODg4UzBguVqwYlSpVYvv27Zw+fdrGrRd5OBRUtOvbt691VDw7O5tDhw4B4OLiQk5ODgCdOnXi2rVrXLx40ZZNFimSVq9ezYgRI8jKyqJcuXLW45UrV6Z///706tWLVatW0b17d0wmk/VFisKbyINJIU7sjsViIScnh23bttG+fXsee+wx8vPziY+Pp0KFCvTp04d169YxaNAgEhISyMjIAH6f1pWYmAhAVlaWDb+ByMPh5nWnRqORrl270qtXL6pXr87UqVP5+uuvAawPizdu3MDHx4e8vDybtVmkqAoJCSE7O5vTp08TExNDWlraLT/r0aMHQ4YMYefOnSxZssSGLRWRv0LTKcXuGAwGjEYjX3zxBQaDgVatWnHq1CnGjh3Lv//9b1555RUOHz7M1q1bWbp0KZUqVaJixYpYLBZ27NhBrVq1ePbZZ239NUSKrIKNuh0dHcnOzubChQvEx8dTsmRJwsPDKVu2LBcuXGDDhg0kJSVRsWJFdu7cyaJFi/Dz86N37962/goiRU5wcDAtWrSwTqn09/cnNDQUV1dX4PepzxERETRq1Ii2bdvauLUi8v+j6pTywDt27Bjx8fGEhoYSFBRkLXM8ceJE8vLymDBhAl27dsXV1ZXo6GgAxo4dS2ZmJo888ggvvPCC9VqqSClyf+Xl5d2y3rRnz54cPXqU1NRUWrZsyfvvv4+npycnTpzgs88+46uvvsLNzY3y5ctTu3ZthgwZgouLyy2V9ETk7/lj/8nOzsZgMFinL7/55pts3ryZV155hS5dulCyZMnbrqHtPEQebBqJkwfayJEjWbRoEV9++SXffvstV69epVatWtZCCFWrVsVkMjFv3jxeffVVypcvT2pqKjExMQQEBPDaa68BYDKZAHRDErmPpk2bxtatW6lWrRpubm4MHDiQs2fP0qVLF6pWrcqmTZuIiYkhKiqK8PBwIiMjcXFxITExkQoVKjBmzBhtJSByl26u5Dpt2jSio6OZNm0aP//8M6mpqdSuXZvWrVuTlZXFvHnzyMjIICQkhOLFi99yHb3wFHmwKcTJA2vChAls2bKFgQMHMmbMGJycnPjyyy9p3rw5gYGBAHh4eHD+/HkWLVpEWFgYxYsX55tvvmHdunWMGjUKf39/4PfwphuSyP1hsVgwmUx88sknxMXFceXKFRwcHNi7dy+vv/46HTp0oG7dukRERLBr1y4+++wzwsPDqVKlCrVq1cJsNvOvf/2LoKAgLBaLRuBE7kLBvW7kyJFs2LCByMhI6taty+XLl1myZAmpqak0adKEBg0aEBgYyKxZs4iPj+fJJ59U1WYRO6IQJw+kJUuWMG/ePD788EMef/xxvLy8iIqKIiYmBl9fX9LS0vjtt98ICwvD19eXy5cvEx0dzdq1a9m/fz+jR4+2FjxReBO5vwwGA46OjvzrX/8iJSWFmJgYzp8/T0JCAi+//DLFihXDwcGBMmXKULVqVRITE4mOjsbT05PatWtTu3ZtAgICNN1Z5C7c3H/i4uKIjo5m0qRJPPfcczRs2JDs7Gy2bdtGt27duHz5MsWLF6dmzZpERkZStWpVKlWqZONvICJ/h9bEyQNnz549dO3alQ8++IB27doBv4+kmUwmnnjiCfLz87lw4QIAYWFhzJ07l9KlS7NlyxYuXbpEhQoVrPvEiUjhuHkt3Pz58/nkk09IT09n7NixPP/887d8Nj4+nsWLF/Pll18ya9YsWrRoYYsmixQJp06dwtnZmZCQEOux9evXM3HiRJYsWUK5cuWIj4/nX//6F926daN169aMGTOG3r178/jjj9uw5SJyN7RASB4o6enpjBw5khIlShAeHo6Dg4N1Hdt7773HjRs36NevHytXruTNN9/kypUrDB48mLS0NJo1a0bnzp2tAU7vJ0QKR35+Pk5OTqSmpgLwyiuvMHnyZIKCgli8eDHLly+3rksFKF++PH379mXq1KkKcCJ3ITU1lc6dOxMfH3/L8Rs3bnDlyhV8fHwA6NGjB82aNWPw4MGULl2aU6dOab9UETunlePyQPHy8qJVq1b88MMPvP322/Tp04fGjRvzxRdfsGLFCmbNmkXDhg0xGo1Uq1aN69evs2jRIi5cuGC9WRXQtCyR++/mCnb/+c9/MJvNzJw5k6ZNmxIQEMB//vMf5s2bR0pKCi+88IK1nwYFBREUFATcXklPRP6apKQkMjIybiva1bBhQ8qUKcOIESO4fPkygYGBjBs3DrPZzMWLF/H397dWehYR+6SROHlgFIycDR06lFdeeYW0tDSmT5/O+++/z8SJE5kwYQKNGjXCaDRa17pVrlwZb29vcnNzbdx6kYfPzQFu48aNZGZmsnXrVvr160dSUhJVqlRh/vz5hIWFsXTpUmbNmkVCQsJt11GAE/lnPDw8MBqN1lHwgvtoQEAA7du3Z/fu3Zw+fZquXbvi6+tLcnIyX3zxBTk5ObRs2dKWTReRu6TCJvLAMBgM1nBWtWpVSpcuzb59+9i0aRM1atRgxIgRODo6Wh8czWaztYBC165dcXd3t/VXEHmoFIx2jxs3jjVr1gBQokQJjh07xo8//khYWBgVK1bk6aefJikpia+++ooLFy7QuHFjnJ2dbdl0kSIhLy+PFStWEBkZSc2aNTEYDNYCJ3Xr1sVoNPLbb7+xc+dO1q5dy7fffsvRo0eZOXMmZcqUwWw2a+sdETulECcPlJtvQKGhoURGRnL27FlOnz7N2bNnqVGjBm5ublgsFr777js++OADxo4dS2RkpCpRitxnBX3z5r62ceNG5s6dy/jx4+nfvz8vvPACISEhnD17lhUrVuDn50elSpVo2rQpubm5REVFUbVqVRt/E5GiwcHBgbVr11K6dGnq169PXl4ejo6O5Obm4ujoyCOPPELFihUpU6YMOTk5tG3blh49elCtWrVb9pMTEfujNXHywCl4OLRYLFSsWJF33nmHKVOmsHnzZmshE7PZzNixY+nTpw9PPvkkFotFbxNF7rPU1FT8/Pxu6WuJiYn4+flRuXJl3NzcAGjTpg3FixdnwoQJjBo1iosXL9K7d28GDRpkq6aLFAmHDx/m2rVrNGzYEAcHB5ycnPDx8SEjI8NaYAi4ZaS7fv361K9f/7Zr6Z4pYt80EicPrII3/h4eHjRt2pS0tDTi4uLYuXMn0dHRNG7cmHHjxt3yeRG5P9LS0ujWrRtBQUGULVvWevynn35i9+7d9OzZE2dnZ+sIQOnSpXF2dmbTpk0cP37cOo1SRP4Zk8nEW2+9xcyZMwkODqZ8+fI4OzuzYcMGvvvuO7Zu3cqGDRvYtGkTcXFxHDx4kL1795KSkkJcXByVK1fGaNS7e5GiQiFOHih/3OzXYDBgNpsxGo3Ur18fZ2dnYmJiqFatGnPmzAHQnH6RQnD+/HlycnLo0qULZrOZzMxMXFxc8PX1ZdmyZZw/f56WLVvi6OiIyWTCwcGBU6dOce7cOerWrcuOHTuIjIykZMmStv4qInbJwcGBmjVrkpGRwezZs8nPz6dmzZrs3buXEydO0KBBAxISEkhOTubw4cMcOnSIHTt2sH79ei5evEjXrl1t/RVE5B7SZt9iMzdXtruTmwPdzZ/dvXs3NWvWxNnZGZPJpDeLIoWkoE8OGzaMCxcuMHbsWCpUqMDcuXP55JNPaNKkCe+99x5ubm5cvXqVGTNmkJaWxn/+8x+eeOIJ+vXrR+/evW39NUTsWlpaGp9++ilz5syhQ4cOODk5ER8fz8KFC3FwcMBoNJKcnIzBYODixYvk5OQQGRmJm5vb//e+KyL2Q0+/YhN/LE1+6tQpLl68SJMmTQgPDycoKOiWETkHBwfrOXXr1rVeQwFOpPAUFB7y8vJi//79TJgwgTfeeIOePXuSSdCLOQAAIABJREFUm5vLihUraNq0KdWqVePq1askJCQQHR2Nv78/ERERt+3lKCJ/X7FixejduzelSpVi8uTJXL9+nfLly5OXl4eHhwcApUqVwmAwUKpUKet52o9RpGjRdEqxiYKANmnSJBYuXMiRI0e4evUqn376KYcPH8bX15dy5crd8Zw/+7uI3D8Fo3AGg4EmTZrg6OjIzp072bZtG0FBQXTp0oWGDRta96yKjIxkxIgR1KpVi9WrV7N27Vq6d++u6ZQi94DRaKRixYpUq1aNhIQE4uPjqVWrFmXKlLH20z/SCJxI0aLplGIzP/zwA8OGDeP999+nRo0alCxZkpkzZzJr1iwWLVpEWFgYAQEBtm6myEOtYAQ8NzeXvLw80tPTrUFs69atzJw5k2vXrvHiiy/SrVs36znp6els376dw4cP8/XXX9O7d2969uxpw28iUvSYzWaOHz/O22+/zalTpxg1ahStW7fWPowiDwGNxIlNWCwWvvnmGywWC7169aJEiRJcunSJ0aNH06lTJ8qVK8enn35K48aNNf1DxEZMJhOOjo7Ex8czefJkpkyZwjfffMOWLVsICQmhfv36VK9enaNHj7JlyxaSk5OpUaMGrq6unDp1inHjxmE2m+nYsSM9evSw9dcRKXIcHBzw9/cnKiqKc+fOMX/+fHJzc6lTp47unSJFnEbipND8cUH1qFGj2LVrF5s3bwagdevWBAcHM2fOHH799Vc6derEwoULadCgga2aLPLQKpg+mZGRQZs2bQgPDyciIoKAgADWrl3LyZMnmTt3Lg0bNuTixYtMmzaNmJgYevfuzYABAwC4fv06rq6uGhUQ+Qf+WK35z44VSE1NZfr06SQnJxMdHV0YTRQRG1JVCCkUN2/GvWPHDho2bEi1atVYv349cXFxLFy4EIPBwDvvvIOTkxPXrl2jRIkSeHt727jlIg+nggfFjz/+mMDAQMaPH09ISAgAP//8M3l5eRQvXpy9e/dSp04dxo8fT9myZXnxxReB31/aqP+K/DM3FyFJTEzEyckJLy8va5+6U5jz9fVl5MiR1pcmqkQpUrQpxEmhKLjZTJkyhR9++IFVq1bRvn17Pv/8c1555RW8vLyIjo4mMDCQjIwMdu3ahZeXF8HBwTZuucjDqeAhMTk5GR8fH0qXLg3A7Nmz+e9//8vChQuJj49nxowZzJ49m7CwMPr27QuoCp7I3bBYLNb+M2bMGHbs2EFOTg61a9eme/fu1KpVC4PBcMeQ5urqav1vBTiRok09XApVZGQkp0+fZtWqVXh4eDBz5kzq16/PjRs32Lx5M7NmzeK9997j888/Z/To0fj6+pKfn2/rZos8dApm2hsMBtLT0zEYDMTGxvLRRx/x1ltv8cgjj2AymUhOTr4tsCnAifxzBS89J0+ezA8//MAzzzxDy5YtOXz4MO+++y6xsbHA/229IyIPJxU2kUJR8FY/LCyMzMxMVq5cySOPPELlypWpV68eeXl5bN68mQMHDhAQEEC/fv1o1qyZpoOI2EjBg2ReXh5LliyxrrcZMGCAdcrk8ePHOX78OK1bt8bPz8+WzRUpEvLz8zEYDGRlZbF06VJ69uxJ165dadasGYGBgezdu5cdO3ZgNpupWbMmBoMBs9ms+6TIQ0jTKeW+MZlM1s24b5768fjjj7N582bWr19PZGQkgYGBjBo1ipSUFIxGI+7u7jg5OVnPE5H7r6C/Xrt2jdTUVBwdHSlVqhRPP/00e/fuZfXq1ZQuXZqePXuSkZHByZMnmTlzJnXr1iUsLMzWzRcpEhwcHNi/fz/z588nJSXllpcjrVu3pmTJksyYMYMlS5Zw6dIlXnvtNdzd3W3YYhGxFY3EyT2Vnp7Ovn37CAkJwcHBgeTkZN555x38/Pzw9vbG2dmZUqVKATBjxgwiIyMJDQ0FwN3dHVdX11umYinEidx/ZrMZo9HI1atX6d+/P9HR0cTGxnLmzBkiIyOpV68eubm5bN++nXXr1rF69Wo2bdpEUFAQH330EfB/IwgicndOnDjB0qVLOX36NDVr1qRSpUrW+2KpUqWoU6cOCQkJrFq1Ch8fH2rWrGnjFouILSjEyT01aNAgzp07R7NmzTAYDKxYsYI1a9bw+eefc+DAAU6cOEHFihWpUqUKFy5c4NixY9SpUwc3Nzc9AIrYSMFUrBdffBGDwUDnzp0xmUzs2LGDw4cPU7t2bZ566imaN29Obm4u4eHhtGnThkGDBuHo6GjdT05E/r4/VposXbo0UVFRnDhxgvXr1xMaGkrZsmVxcHDAYrFQrFgxGjZsiIeHB926ddO9U+QhpX3i5J7as2cPFStWpFixYvz44480a9aM3Nxctm3bxurVqzl58iRXrlyhffv2JCUlER8fz5gxY2jWrJmtmy7y0Ln54fHUqVOMHj2aCRMmEB4eDsAnn3zCypUrKV68OK+99hqNGze+7Rpatyryz93cf5KSknB2dsbX1xdnZ2eSkpJ455132LFjB2+88QadO3fG3d39ttCnarAiDyeNxMk9FRwcjKurK19++SVvvvkm165do2bNmlStWpWmTZvSpk0bSpQowZEjR0hPT+fkyZMcOHCATp064ejoqDeKIoWooL/Nnj2bmJgY0tPTee6553BzcwOgTp06eHt7s3v3bnbt2oXZbCYyMvKO1xCRv+fm/VNnzZrF9OnTmTNnDmfPniUgIICKFSvSuHFjMjIymDNnDllZWVSpUuW2NXB6iSLycFKIk/vC39+fa9eusX37do4ePUqFChUoWbIkXl5e1KxZkwYNGtC8eXOcnZ3p1KkTERERehgUsQGTycSePXtYs2YN165do3HjxgQGBlr7Y6VKlShXrhz79u0jNjaWevXqERgYaONWixQNBoOBhQsXMnv2bBo3bky9evVYvnw5586dw9fXl/DwcBo2bIiXlxdz587l8OHDPPHEE9YNvUXk4aUQJ/fEzUUNTCYTXl5eNGnShNTUVOLi4ti3bx/BwcHWDYPd3Nys8/rLly8P3L4uQETuPwcHB6pXr054eDhxcXH8/PPPVK9enYCAAOtnQkJCqFatGuHh4TRv3tyGrRUpOgwGA4cOHWL8+PGMGjWKV199lXr16nHgwAF27drFwYMHKVasGOXKlePRRx8lKCiI4sWLU79+fVs3XUQeAApxctcKihpcuHCBlStXsnHjRrKysggPD6dBgwYYDAbrdCwPDw/rqJvJZLplGogCnIhtODk5Ua5cOWrUqMHGjRtZu3Yt5cqVs1aZtVgs+Pv7U7VqVUCVKEX+iTvt5/bjjz+SmppKv379cHd3Z9OmTRw9epQBAwawd+9efvzxRwwGAx4eHjRq1Ig6deoA6oMiohAnd6mgNHlmZiYvvPACcXFxHD9+nNWrV5Ofn09UVBQ1atTA39+f/fv3s2vXLlJSUoiKitI8fpFC9r9Gux0cHAgODiYqKorDhw+zePFiSpQoQVhYmHXfxgJ6eBT5+1566SUqVKhwy3Tk//73v8TExNC7d2+MRiMDBgygfPny9OvXjzJlyrB8+XIOHz7MpUuXaNWqlfU89UERUYiTu1IQxLp164arqyvTpk2jXbt2HDt2jJiYGE6ePEmrVq2oUKECERER/PTTTyQmJvL0009bNwIXkfvrq6++4sCBA9aiJImJifj4+Nzxs35+ftSvX5+0tDQ+/vhjjEYjjz76qB4aRe7CuXPnOH78OK1bt8bNzY3c3FwcHR1xcXEhOzubZs2aMXPmTH755ReWLFkCwIEDBzhz5gx9+/alT58+uLi4aNmBiFjpKVr+kZtvJN9++y3p6elMnTqV8PBwzpw5Q4kSJWjatCmbN2+mc+fOLFiwgBo1avDxxx9jMplwdXVVaXKRQuLg4MCECRO4dOkSaWlpXL9+nbfeegtPT887fj4wMJChQ4dSokQJoqKi9NAocpeCg4OZOHEiRqORd955B7PZzBtvvEFkZCSBgYG4urpy5MgR6zYeubm5JCUl4efnR9u2bfH09NQ9U0RuoX8N5G9JTEwkPT39loe6rKws8vLyrA+Eq1atIiUlhaFDh9K9e3cOHjxIy5YtmTt3LsWLF6dUqVK3lFYWkfurcePGjBo1iujoaJYtW8bjjz/+pwGugLe3N/379+fRRx9F24mK3D2j0Uhubi7Ozs4sW7aMcePGcfHiRev0SldXV3755RdiY2NZuXIlixYt4plnnrH2Vd0zReRm+hdB/rJjx47Rrl07Fi9ezMWLF63HTSYT169fx8fHh0OHDhEdHU2fPn2oUKECLVq0wNvbm8qVK2M0Gq03Ib3ZFyk8JUqUoFq1apjNZtzc3JgzZw4JCQn/3/PUX0Xu3s0vQZydnXnjjTd499132bp1K2+88Qa//vorAM8++yzx8fEMHz6c2bNn0717d5566ilbNVtEHnAGi16xyt/Qt29ftm/fznPPPccLL7xAWFgY8PuUyqeeeophw4aRkpLCggULyMnJITY2lqVLlzJ9+nSCg4MBbSUgUlhu7msJCQmcOnUKR0dHpkyZQmpqKpMnT6ZBgwY2bqVI0WUymTAajaSnp5OUlERubi4VK1bE3d2duLg4xo0bh8ViYcKECTRo0ICEhASSk5MpVqwY1apVA9A0ShG5IxU2kb+koDRyu3btyM3NZdGiRVy6dImAgACCgoKIiIgAYO3atRgMBtq1a8fFixeZP38+fn5+PPfcc9ZrKcCJ3H83lzO/cuUK+fn51KpVi7Jly1K9enVOnjzJ3LlzrVsHZGdnWx8eReTuFVRvzsjIYODAgSxatIht27ZRpUoVSpcuTXBwMPXq1WP37t0sWbKEEiVK0KBBA8qUKWPdp1EBTkT+jEKc/CUODg7Wh8J69eoREhLCggUL+O233yhRogRlypTBwcGBCxcusHTpUrZv387XX39NcnIyCxYswNnZWfvaiBSS/Px8HB0dARg7diwzZ84kOjqasLAwwsLCCAwMJDIykvT0dObMmcO5c+fYsmULa9eupU2bNqocK3IPFISvF154gfz8fAYNGkSLFi2oX78+WVlZXL58mXLlytGyZUuSkpKYO3cunp6e1KxZ03oN3TNF5M9oOqX8LQWLsgH279/Pq6++iqurK/3796d169ZkZ2ezcuVKfvrpJ8qWLcvLL79MaGiodUqJiBSe0aNHs2PHDrp06UJISAj16tXDy8vLWljIZDIxf/58VqxYgdlsZsaMGdZCJnp4FLl7K1as4NNPP2X+/PkEBQUB8Pnnn7N06VJSUlJo0qQJkydPJiMjg/nz59OmTRsqVqxo41aLiD1QiJM/tXXrVmshhOLFixMYGIizs7N1jxtnZ2euX79Or169SExMpGfPnjz33HPW/acKPqPpICKFLy4ujmHDhjFlyhSioqIA2LNnDzNmzCAjI4MqVaowdOhQfHx8uHbtGunp6YSGhqq/itxDX375JYsXL+bzzz/nzJkzLF68mO+//54mTZpQvHhxfvjhB0aPHk379u0xm804Ojpa/xQR+V80NCJ39Ouvv9KnTx/r30uVKsW1a9fw8/MjODgYf39/ypUrR+nSpRk8eDBvvvkmy5YtIzU1lQ4dOhAREWEdsdMDoUjhy8/Px8fHh9KlS5OUlMRXX31FdHQ0ERER+Pj4sGXLFurWrUuHDh3w8/PDz88PUH8VuZc8PT25fv06/fv359SpUxgMBt5//306dOjA5cuXiYuLIzMzE8Aa3BTgROSvUIiTOypZsiQdOnQgLi4OgIYNG1K5cmUOHTrEuXPn+OWXX9i0aZN1426TyUR2djaLFi2iYsWK1kInImIbjo6OnDx5ksmTJ3Ps2DHOnz/Pyy+/TL9+/fD29qZNmzbW0uYicvduHknLysrC09OTNm3akJSUxP79++nYsSMtW7akdu3aAKSlpeHi4oKrq6uNWy4i9kjTKeVP5efnM3PmTH744QdKlCjB8OHDqVy5MgA3btwgOzub9PR04uPjMRgMHD16FCcnJ3r16mXjlos8XP5sCuTSpUtZvHgxVapUoUmTJnTs2BGAy5cv8+qrr9KiRQv69u1b2M0VKXIK1n2npaUxffp04uLiqFOnDsOHD8fb25sbN27g7u5OSkoKSUlJZGZmMnfuXHJzc1mxYoWtmy8idkghTv6/vvjiC5YsWYKTkxPdunWzPgj+LypkIlI4bl4/c+jQIU6fPk1eXh6NGjXC39+fnJwcXFxcMJlMmEwmjh8/zrfffktMTAwxMTH4+fmpkInIXbj5JUr79u1xcHCgSpUqhIWF0aVLF9zd3a39sGvXruzfvx8/Pz9CQkKYN28eHh4eWgcnIn+bQpz8JT/++COzZ8/m+vXrPPPMM/Tu3Rv4v7Cmh0CRwndzvxs7diy7du0iOTkZb29vsrKyePnll3n22WcpXbo0Y8aMYeXKlQQEBODq6srkyZOpUaOGHh5F7pH33nuPXbt2ER0dbd3nbcOGDWzevBmDwcC///1v/P392bt3L6VKlSIsLAxfX1+99BSRf0T/ashf0qxZMwIDA5k6dSrLly/n0qVLDBo0yFquXAFOpPAV9LtPP/2U2NhY3nrrLSpXroyvry8DBw7k008/pV27duTm5tKrVy+Cg4MpV64c5cuXJzw8/Jb95ETk7ly5coU6deoQEBDAiRMnWLFiBUuXLiUoKIjs7GySk5OZM2cOTz31lPUci8WiACci/4hG4uRvuXTpElOmTGHPnj1Ur16dQYMGUb58eVs3S+ShlZGRweuvv0758uUZMmQILi4unDx5kmeffZYBAwYQHBzMhQsX6Natm62bKlKk3PwCMycnh/79+5OSkkKLFi1Yv349CQkJDB48mC5durBx40ZGjhzJmjVrCA8Pt3HLRaQoUC1p+VsCAgKYOHEibdu2ZcuWLezZs8fWTRJ56OTn51v/29PTk4sXL1or3WVmZtK9e3datGhB165dOXPmDJ988gnJyck2bLFI0VIQ4H799VdOnz6Ni4sLb775JtevX7dOp5wxYwY9evTA3d0dHx8fypYti9lstnXTRaSI0Bi+/G3Ozs4MGTKEOnXq0KRJE1s3R+ShUrCGLTMzk/379/PYY49RqlQpEhMTMZvN9OjRg6CgIMaMGWPdq9HLywsPDw8bt1yk6DAYDOTm5vLWW2+xf/9+oqOjadSoERs3buTUqVOEhYUBv/fXxMREZs+eTalSpahUqZKNWy4iRYWmU8pd+7Py5iJyb93c1zp27EiJEiWYNm0ax44ds77x9/LyYsWKFfj5+ZGRkcH48eO5cuUK8+bNs4Y6Ebk3tmzZwmeffUZcXBwjR47k5ZdfBn5fH9enTx8sFgs5OTkYjUZWrlyJs7OzigmJyD3hOH78+PG2boTYNxU1ESkcBX1t/fr1HDt2jNGjR+Pv74+/vz8eHh6cPHkSs9lM7dq1+fnnn1mzZg3r1q1j+vTplCxZkvz8fPVXkX/oTkW8ypYtS4UKFbhx4wYLFy4kLS2Nxo0bYzQaSU5OxmKx0Lx5c4YOHYq3t7cqUYrIPaOROBERO/Ldd98xc+ZMcnNzWbZsmbWUeXp6Otu2bWPp0qUcO3YMZ2dnKlWqRI8ePWjSpIne/ovcI8ePHyc0NBRXV1frsaSkJJYtW8aiRYto2bIlH330kXXKZcEIuGatiMi9pJE4ERE7cuDAARISEjhx4gSenp7UrVsXABcXFypWrMgzzzxD48aNeemll2jfvr11DY4eHkXu3oYNG3j55Zfx8/O7JcgVK1aM8PBwbty4wTfffMMPP/xAo0aN8PX1tZ6rUXARuZcU4kREHmB/nAJZtWpVypQpQ0pKCitXriQzM5PHHnsM+L2IgoODA/7+/nh7e+Pi4qIHR5F7KDQ0lJ07d/LNN98AUKZMGby9vQHw8PAgKCiIbdu2kZWVBUC9evVs1lYRKdoU4kREHlAmkwlHR0eysrI4evQohw4dIiwsjODgYCpXrkx2djYbNmzgl19+oXHjxri4uFiDHOjNv8jdurk/we8j2s8++yznzp1jyZIlpKenU7p0aUqUKAHAb7/9xunTp5k0aRJt2rSxVbNF5CGgECci8gAym83WAgh9+/Zl/vz5rFmzho0bNxIVFUVYWBjVqlUjNzeXLVu2sH37diIjIylevLiNWy5SNBSsIzWbzaxZs4bdu3dz7tw5KlasSPPmzXFzc2PBggUkJibi5ubGjRs3WLx4MdnZ2fTs2RO4czEUEZF7QYVNREQeYH379uX06dP06NEDR0dH3n77bby9vXn33Xdp0KABWVlZrFixgvnz5+Pv78/y5ctxdnbWg6PIXbi5CEnfvn3Zv38/OTk5uLm50bZtW0aPHg3A5s2bGT58ONnZ2Tg7O+Pt7c3SpUsJCgpSIRMRua8U4kREHlBff/01c+bM4cMPPyQyMpJff/2VDz74gMTERC5dusTo0aPp0qULADExMYSEhBAZGWnjVosUHWPHjmXLli1MnDiR4sWL88EHH3DgwAEaNWrExIkT8fPzIykpie3bt+Pt7U316tUpU6aMqsGKyH2nzUpERB4wBVOwEhMTKVmyJGFhYQBs27aN7Oxsxo0bx6effsr48eM5ffo0Tz75JG3btrVxq0WKln379hEXF8c777xDo0aNyMzM/H/t3XlcVdX6+PHPYQYRFUXIATXkgIIM4myIKGpqDmk4pDjkrdTyIpopXSsUs9LSQnOeJQwculigoagY9HWeNYecUMERREBk3L8/+J19OQI5ZKL2vF+vXtez9zprr73P2dz9nLXWs7C1tcXFxYWjR48yevRoPvvsMxwcHBg4cKD6PkVRJIATQvztpJ9fCCGeMbqhkDk5OaSnp1OpUiVOnz7N3Llz6devHz4+Pvj5+QGwfPlyfv3114psrhAvjJKDk3JzcykqKuKll14CYMuWLZw+fZqxY8fy5ptvcujQIYYOHcrcuXMpKChQ3ydDmYUQT4P0xAkhxDOqZ8+e6kPlV199RatWrejduzcA+fn5+Pn5ERISombGE0L8NSUDsLy8PFJSUjAxMSErK4svvviCQYMG0bJlSxo1asSSJUuoXbs2Fy9eVJMQCSHE0yJ/dYQQooKVnD+Tn59Pfn4+Go0GV1dXdbHuzMxMdc2pzMxMDh06BKBmo5QkCkI8np07d5KUlMTly5dp3rw5Xbt2xdbWFh8fH0JDQ6lZsyZLly6latWqjBkzhsLCQo4dO0bt2rWZMGECzZo1A+QeFEI8XRLECSFEBSoqKlIDuLlz53Lw4EFSU1OxtLRk2LBh+Pn5UVhYiJmZGRs3bsTKyooDBw6we/dufvzxRzQajTw8CvGYgoODOXbsGNnZ2djb2/PDDz/g5OSEra0tAP7+/gCkp6djYWEBFA9z3rFjByYmJjRq1AgoHoYp96AQ4mmS7JRCCPEMmDJlCnFxcXTs2BFzc3Nu3LhBbGws7733HqNHj+b48eN89tlnnDlzBq1Wy7hx42jRooVkwRPiMX3++eds3ryZ4OBg2rRpg5WVFenp6VSrVg2A1atX0759e+rWrcu6dev49NNPad68OaampuzatYuVK1fi4eEhP6IIISqE9MQJIUQFKLkI8J49e9iyZQtTp07F29sbExMTNm3aRGxsLI6Ojhw+fJimTZsSGRlJcnIylStXplq1anq9eEKIh7djxw5iYmIICQnB19cXQ0ND8vLy1ABu2rRphIeHs2PHDsaPH0/Pnj1JTk5m165dmJmZERYWJgGcEKJCSRAnhBBPUV5enroYt+4BMD09HUVRqFWrFiYmJly9epWPPvqIt956CycnJ8aPH8/AgQPx9/fH3t5erUseHoV4NLofT/bu3Yurqyuenp7qDyG65CSrVq0iMjISf39/Dh06xMSJE5kyZQrjxo2jsLCQ3NxcdWil3INCiIoif32EEOIpycjIYOLEiezdu5fCwkL1AfDu3btkZWVRs2ZNAAYNGkTr1q0JCgrCwsKClJQUbt26VZFNF+KFoNFoUBSFvXv3UqVKFTUxEBQHZImJiUyfPp1p06YRGhrKmDFjuH79OpGRkdy9exdDQ0M1gBNCiIokQZwQQjwlly5dYtOmTYSGhrJz505yc3MB8PHxoWrVqowfP56hQ4dSrVo1Pv30U4yNjcnPz6dmzZpYWlpWcOuFeDHk5uZiamqqru2Wl5en7svLy+ODDz6ga9euAHTu3BkDAwOqV68uwZsQ4pkiQZwQQjwlrq6ubN68GY1Gw0cffcRPP/1Eeno61tbWjBo1ilOnTrFv3z7Gjh2Lra0tKSkprFixgrS0NLp161bRzRfihWBmZka9evX49ddfuXHjBiYmJuq+Dh06MGLECHVbQkICFhYW6jICkgtOCPGsMAwJCQmp6EYIIcQ/RdWqVXnttdc4evQoq1atwszMDK1Wi5ubG8bGxqSkpBAVFcWePXuIjIzk+PHjfPPNNzg4OOgNwRRCPLwFCxYAUKtWLaA4kPvpp5+4efMmrq6uak93yXssNzeXDRs2kJqayogRI7C0tNRbDFwIISqSBHFCCPE3KZmBsqioCI1GQ0FBAebm5nTr1o2MjAwWL15MWloabm5ueHt74+HhgZ2dHRkZGXTq1Im33noLT09PyUQpxF8QHx9PcnIybdq0AaBu3bqcP3+e//73v+Tn52NnZ0eNGjXUAC4tLY24uDjmzJnD5MmTcXd3V+9hIYR4Fsg6cUII8TdJS0ujcuXK6mLdZQkPD+fzzz+nVatWBAUF0bhxYwwMDPQCQCHEXxMbG8vChQv54YcfMDc3V7d//PHHrF27liZNmtClSxeaNWtGSkoKmzdv5uTJk/j7+/P222/L/SiEeOZIECeEEH+DW7du4efnh62tLQCNGjXC0NCQBg0aYGdnh42NDQ4ODpiYmHDkyBHGjRtHs2bNGDJkCM2bN5ckCkI8Yf369aNx48YEBwdjamqqbl+7di0rV67k4sWL6py3V199FR8fH3r06AEgQZwQ4pkj68QJIcTfIDk5maKiIi5cuICZmRkVOHmLAAAgAElEQVRGRkYYGRmxY8cO7t69S1FREQCWlpbUr18fU1NTkpKSOHXqFNOnT6ddu3YVfAZCvBh06zH6+fkRFRXF0aNHadasGQUFBRgZGeHv74+vry/37t3jxo0bWFlZ4eDgUOr9QgjxLJGeOCGE+Jv88ccfTJo0ibt379K1a1eGDh1Kfn4+mZmZJCcnk5aWRmpqKqdPnwbg9OnTGBkZsXr1aqysrCq49UK8WBRFoX///ty5c4eVK1dia2tLfn4+RkZG0ssmhHjuSBAnhBB/A92v92lpaUyYMIETJ07QvXt3hgwZgr29fZnvycrKAop753S9BEKIv66wsBBDQ0NOnDjBBx98QHZ2NkuXLqVhw4ZqGelxE0I8TyQ7pRBC/A00Gg2FhYVUqlSJXr16cenSJWJiYrhy5Qq1atVS58rl5eVhaGhIUVERpqam6vpU8jApxOO7P5Ok7n6qXr06devW5fDhw8ybN4+qVatSp04dzM3N1XtWEgsJIZ4HEsQJIcTfxMDAQH0o9PX1RaPREB0dzZkzZ6hWrRr169dXAzgJ2oT4a3bv3k10dDSNGzfG1NS0zEBMo9FQv359mjRpQlFREQsWLODgwYNcv34dOzs7rKys0Gg0EsAJIZ55MpxSCCH+RrphXDpxcXFMnz4da2trBg8eTO/evSWAE+IvunbtGq+++ioGBgb4+/szePBg6tSpA+hnliz577y8PM6dO8e8efNIS0vj+PHjtGvXjuzsbGbOnEm1atUq7HyEEOJBZMKFEEL8TUou0P3vf/+bvn370rlzZ2xtbQkJCSE0NBQTExNee+21Cm6pEM+3zMxMKlWqxM2bN0lKSuLmzZsMGzYMV1dXNBqN2ttdsofN2NgYZ2dnvv32WwoLCzl06BBQnJBIAjghxLNOeuKEEOIvKtnblpWVhaWlpd4v/uPHj+e3335j7dq11K5dG41Gw9WrV5k1axYhISGyJpwQT8COHTv45JNPsLGx4ebNmzRs2JDhw4fzyiuvAGUnLpGhzEKI55XMiRNCiL+gZG/bhAkTANBqtWoA9+WXX7J582YWL16Mo6MjGo2GgoICrKys6NSpE8bGxuq8OSHEo9P9Fm1tbc21a9dwdnamY8eOxMbGcuTIEUxNTXF2dlZ75Er2xsncNyHE80qeGoQQ4jGcP38eRVHU5CXXrl1j586dmJmZAcXBXUZGBklJSXz00Ue4ubmp771/6YCSc+aEEA9WchCRLhFJlSpVaN++PcuXL6dOnTqsWLGC/Px85s+fz9KlS8nJyVEzTwohxPNOgjghhHhEhw4dYujQoSxcuJCsrCwMDQ2xsLDA2NiYS5cuqeWqVKlCREQEvXr1qsDWCvHiyc3N1XtdUFAAQPv27Rk1ahQzZsygatWqLFiwgJo1a7Jy5Uq+++47rl27Jr1vQogXggynFEKIR2RgYMDOnTvZvXs36enp1K9fHxsbG+Lj47GwsMDb2xso7iEwMTEpMzOeEOLx/P777wQEBKhDI6tXr46xsbF6f1laWrJr1y6ys7Px9fXl1Vdf5fTp08TFxXH69Glatmwp81CFEM89CeKEEOIRVapUiT59+nD69Gk2bdrE1atXcXBwIDU1lU2bNpGcnMzu3bu5ffs2ycnJVKpUiWvXrmFtbV3RTRfiuXbv3j1GjhzJ+fPnSUxM5I8//iA6Ohpra2vMzc2xtLSkRo0aaDQaZsyYQePGjXF0dKRz585cuXIFY2NjevToUdGnIYQQf5lkpxRCiL9g7ty5rF69Gg8PD27dusX58+extbXlzp073Lx5EyMjIwoKCnBxcWHt2rWSwESIv2j27NkkJCSQmZmJu7s7lpaWrF27Fjc3N5ycnHj33XexsbFh7ty53Lp1i3//+9/Y2trq1SFZKYUQzzsJ4oQQ4i9av349s2fP5vbt23h6erJ69WoATp8+zZ07dzh58iS+vr7Url1bhlQK8ZgmT56Ml5cXvXr1Ijw8nNjYWG7dusXEiRPRarWsWLGC3377TV34+86dO1y+fJmgoCDat29f0c0XQognSoI4IYR4Avbv388nn3xCSkoKQUFB9O7dGysrK70yJdeTE0I8HEVRuHHjBn369KFRo0YsXrwYgNjYWCIiIjh79iyTJk2iV69e5OfnExERwW+//UZqaiqnT5+mVq1a/Pzzz5ibm8sPKEKIF4bMiRNCiCegVq1atG/fnsOHD7Nz507u3LmDg4MDlpaWahkZviXEo9NoNFSqVIk6deqwcOFCqlatipubG46Ojtjb23Pt2jUiIyO5desWPj4+eHh40KJFCzp06ICBgQFvvPEGjRs3lgBOCPFCkSBOCCGeECsrK7p3786JEyeIjIzE3d0dR0fHim6WEC+E6tWrc+rUKY4dO0bbtm2pXLkytWrVolGjRmRlZbFp0yYOHDjAK6+8QvXq1bG2tqZdu3bqPShDmYUQLxIJ4oQQ4gkyMjKiS5cu1K1bl9dee62imyPEC8PMzAxra2uWLl1KzZo18fT0BKBatWq4ublRUFBAYmIiCQkJuLi4UKNGDb33SwAnhHiRSBAnhBB/A2dnZ6B4HpwMoxTi0el6zkr2oNnb25OVlcW6devw8vLC1taWwsJCKlWqhJeXF2ZmZsTFxanJTYQQ4kVlVNENEEKIF5kkMhHi0eXk5GBubg5QKpDr1KkT0dHRJCQk0KRJEwwNDSksLMTExIRBgwbx0ksv0aJFC0CGUAohXlzSEyeEEEKIZ8bs2bMZNWoU6enp3L17l4YNG+oFYi+99BJFRUUsXLgQDw8P6tSpg4GBAUVFRWg0Gho0aICJiYmsBSeEeKFJECeEEEKICqcoCgUFBZw5c4bKlSuzZcsWNm3axC+//MKdO3cwMDDgpZdeQlEUGjRowJ49e0hLS8PHxwconf1VeuCEEC8yWSdOCCGEEM+c06dPc+zYMTZv3szx48fJycmhdevW+Pr60rlzZ3bt2kVQUBArV66kWbNmFd1cIYR4qiSIE0IIIUSFysnJ4dq1a5w4cQITExNatWqlt8biwYMHOXv2LBEREaSnp5Odnc3w4cPZsGEDNWvWZM6cOVhbW1fgGQghxNMlQZwQQgghKkxhYSFBQUGcOHGCy5cvY2VlxQcffMDrr7+ORqPByEg/B9uBAwc4cuQI69evJyUlhSZNmrBixYqKabwQQlQQCeKEEEIIUWECAgLIy8ujX79+tG3blgMHDuDi4kK9evX0yuXn52NsbKy+zs7O5o8//sDZ2RlTU1MKCwslG6wQ4h9DgjghhBBCVIjZs2cTGxvLnDlzaNiwodrrVlRUREZGBocOHcLAwABvb281A2XJ/9WRTJRCiH8aWSdOCCGEEE/dnTt3OHz4MH379qVBgwZ6wyZnzZrF5s2buXz5MgBNmzYlLCyMGjVqAKUzUUoAJ4T4p5G/ekIIIYR46goKCjh58iTW1taYmpqSl5dHcnIyb7/9NkuWLMHc3JyPP/6YkJAQrl27xtSpUyksLKzoZgshxDNBeuKEEEII8dRVqVKFRo0asWHDBhwdHdm/fz/R0dFcvHiRnj17MmHCBGxsbADIzMxk48aN5OTk6GWtFEKIfyoJ4oQQQgjx1BkaGuLn58eiRYsYOHAgAHZ2dnzxxRe0atUKa2trNVlJQUEBVapUKZWpUggh/qnkr6EQQgghKsSgQYOwt7cnJSUFY2Nj2rZti62tLQCKomBoaEh6ejrHjx/Hzs4OU1PTCm6xEEI8GySIE0IIIUSF8fb21nudm5uLqakpGo2G69evEx4ezm+//caPP/6IRqNBURQ0Gk0FtVYIIZ4NEsQJIYQQ4pmQkZHBzJkzqV+/PqampiQlJXH48GFmz55NvXr1ZC04IYT4/ySIE0IIIcQz4dKlS8THx5OTk0Nubi6dOnUiNDQUHx8fdXilEEIIWexbCCGEEM+QvLw8Lly4gImJCXXq1FGTmcgwSiGE+B8J4oQQQgghhBDiOSKLfQshhBBCCCHEc0SCOCGEEEIIIYR4jkgQJ4QQQgghhBDPEQnihBBCCCGEEOI5IkGcEEIIIYQQQjxHJIgTQgghhBBCiOeIBHFCPCWTJk3CyclJ779GjRrRtGlT/P39+fHHH5/o8RRFYebMmbRs2RIPDw++//77J1r/iy4gIAAnJydatGhBQUFBmWUKCwtp2bIlTk5OzJkz57GOc+nSpYduT4cOHR7rGOLhzJkzBycnJy5fvvyXyjwpCQkJjBw5knbt2uHq6kqHDh2YOnUqN27cqLA2leXy5cul7oGMjAxGjRqFh4cHzZs358SJEzg5OTFp0qQKaaMQQrxojCq6AUL80wQHB1OtWjWgONDKyspi48aNTJo0ifT0dN56660ncpwdO3awZMkS2rdvj5+fH15eXk+k3n+ajIwM9u3bR6tWrUrt27t3L7dv337sutevX8+UKVM4cuTIA8uOHDmSnJycxz6WeH7k5+cTEhLCunXr8PDwYNCgQVSpUoUTJ04QFRXFli1bWLNmDXXq1KnopgJgbW3NjBkzcHJyUrctWLCAbdu2MWzYMF5++WXq16/PjBkzsLe3r8CWCiHEi0OCOCGeMj8/v1IPX2+88QbdunXju+++Y/DgwZiYmPzl45w6dQqAcePG6T1ciYdXo0YNsrOziY+PLzOI27p1K9bW1qSlpT1W/Xv37iU3N/ehyrZt2/axjiGeP/Pnz2fdunWMHTuWUaNG6e3r2bMnw4YNY8yYMU+89/5xWVhY0KtXL71tp06domrVqgQHB6vb7i8jhBDi8clwSiGeAWZmZnTo0IGsrCzOnDnzROrMz88HoFKlSk+kvn8iU1NT2rZtS3x8fJn7t27dKkMcxRN18+ZNFi1aRMuWLUsFcADNmjWjT58+/P777xw6dKgCWvhw8vPz5W+PEEL8jSSIE+IZodFogOJ5VjoHDx5k+PDheHp64unpyVtvvVVq6F2HDh2YPHkyH330EU2aNKFdu3Z4eHgwd+5cADp27KgXaOzbt49hw4apdQ4ZMoS9e/c+sM60tDR1Ts7atWvp0qULbm5u9O3blyNHjnDjxg0CAwPx9PTE29ub2bNnU1RUpNaZn5/PwoUL6dmzJx4eHri5udGzZ0/WrVund2wnJycWLVrE8uXL8fPzw9XVlR49erBp06ZS1ywhIYHBgwfj6elJ27ZtCQoKKjUvaPv27QwYMAB3d3eaN2/OmDFjOH/+/EN/Ln5+fly5ckXt2dQ5evQoqampdO7cucz3Pei4AQEBak9KyblC5V37subEnT17lsDAQFq2bImXlxcBAQHs27dP3Z+Xl8dnn31Gx44dcXV1xcfHhylTppCRkfHA8z537hyjRo2iWbNmtGzZkmnTphEVFaU392rOnDk0adKELVu20LZtWzw9PVm7di0A6enphISE4O3tjaurK126dGHRokV63+/y5nLdv33OnDk0btyYc+fOERAQgLu7Ox06dGDevHl69UHx8NfQ0FD1uF27dmXlypUoiqJXLjk5mTFjxtC8eXNatmzJ7NmzS5V50PUZMmQIbm5utG/fnm+//Vb94eTcuXM4OTkxY8aMUu+bOXMmrq6u5X4Gv/zyC/n5+fTv37/cY48dO5akpCQ8PDzKLXP8+HHGjBlDmzZtcHFxoXXr1owfP56rV6/qlVuzZg09evTA3d2dli1b8t5775X6IemXX36hb9++eHp64uXlxfDhw9m/f7+6v+ScON2/9+zZw5UrV/S+22XNiXuY+9PJyYlvvvmGkSNH4urqSrdu3cqdpyqEEP8UMpxSiGdAUVERe/bswcTEBAcHBwCSkpJ49913cXZ2JjAwkLy8PDZs2MCgQYNYvnw5zZo1U98fExNDgwYN+M9//sPNmzdxdnbmv//9L1u2bCE4OFgdvhkfH8/777+Pvb29+iv/2rVrGTZsGGFhYXTs2LHcOq2trdU64uLiGDp0KIqiMH/+fMaMGUPlypVxdHRk0qRJxMXFsWDBAurXr8/rr78OFM8F3LRpEwMHDiQgIID09HSioqL4z3/+g729PS1atFCPvWbNGoqKihg0aBBmZmasXLmSoKAgHBwc0Gq1avvGjx+Po6MjY8aMIT8/n2XLlnH06FE2bNiAlZUVGzZs4KOPPqJ169ZMmDCBjIwM1qxZQ79+/YiKiqJBgwYP/Gx8fX0xNDQkPj5eb1jq1q1bqV+/vvp5lfQwxx05ciRFRUXs27ev1Fyh8q59SRcuXKBfv34YGRkxePBgrK2t+eGHHxg+fDjff/89bm5uTJ06lZ9//pkhQ4ZQt25dzpw5w/fff8/FixdZtmxZueeckpLCm2++CcBbb72FkZER33//PT/99FOpsgUFBUyePJkRI0aQl5eHl5cXGRkZDBgwgCtXrjBgwAAaNGhAUlISX3/9NSdOnOCbb7554HW/n6IoDB8+HEdHRyZMmMDu3bv59ttvuXr1KlOnTgXg7t27DB48mNTUVN58803s7OzYtWsX06dP58KFC3z66adAcW/XgAEDyM/PZ+jQoZiZmREREcGdO3ceuj264HnixIns2bOHefPmkZqayhdffMHLL7+Mi4sLmzdv5sMPP9R736ZNm/D29qZKlSpl1nv8+HEA3N3dyz12Wd+Hkk6dOsWbb75JvXr1eOeddzA3N+fAgQNER0dz/fp1Vq9eDcDGjRsJCQmhd+/eBAQEkJaWxsqVKwkICGDLli1UrlyZPXv2EBQURLt27fD39ycnJ4fw8HCGDx9OTEwMdevWLdW2GTNmsGDBAtLT0wkODi53Htyj3J8rV67E3d2dyZMnc+/ePYyM5PFFCPEPpwghnoqJEycqWq1WOX78uHLr1i3l1q1byvXr15WDBw8qgYGBilarVaZPn64oiqIUFhYqHTt2VAYMGKAUFBSodWRnZyudOnVSevXqpW7z9fVVnJ2dlYsXL+odLywsTNFqtcqlS5cURVGU/Px8pV27doqPj4+SmZmplsvIyFC8vb0Vb29vJS8v70/r9PX1VZycnJSTJ0+q27788ktFq9UqY8eO1Wuni4uLMm7cOEVRFOX69euKk5OT8tVXX+nVd/bsWUWr1SqhoaHqNq1Wq3h4eCjXr19Xtx06dEjRarXKrFmz1OvTtm1bpUePHkpOTo5aLikpSdFqtUp4eLiSmZmpNG3aVAkKCtI75vXr15XmzZsro0ePVv7M4MGDFV9fX/Xfffr00dvftWtX5auvvlIuXbqkaLVaJSwsTFEU5ZGOq/tOlFTetS/ZHkVRlMDAQMXNzU25cOGCui0tLU3x8vJS/v3vfyuKoihubm7KlClT9OqZPXu20qdPHyUrK6vccw8ODlYaN26s/PHHH+q2q1evKh4eHnrfKd13THfuOjNnzlS0Wq2yZcsWve0hISGKVqtVduzYofd+XX0692/XvR49erRSVFSklhs/frzi5OSktjMsLExxcXHR+34qiqJ8/fXXilarVX7//XdFURTliy++UJycnJRjx46pZW7evKm0atWqzPaU1bbAwEC97ZMmTVK0Wq167GXLlilarVY5fPiwWubAgQOKVqtVYmJiyq3/7bffVrRarZKbm1tumfLapGv3J598ori7uyvp6el65YKCghStVqtu/9e//qV0795dr8yOHTuUbt26Kfv27VMURVE+/fRTxdPTU++6nzx5UuncubOyadMmRVGUUveAopT+vipK8b09ceJERVEe7T7RarWKl5eXkpGR8dDXRAghXnQynFKIp+z111+ndevWtG7dmldeeYX+/fsTHx9PQEAA48ePB+DEiRNcunQJPz8/MjIySEtLIy0tjXv37uHr68vvv/+uNyzK3t7+gVnfTpw4wdWrVxk0aBCWlpbqdisrKwYPHsy1a9c4duzYA+u0t7fX65HS/VreqVMndZuFhQXVq1dXU6Hb2Niwf/9+Ro8erZZRFEUdEpWdna13DC8vL2xsbNTXjRo1AlDrO3bsGDdu3KBfv36YmZmp5dq0acPatWvp1asXSUlJZGVl4efnp16/tLQ0DA0NadWqFYmJiQ89JMvPz4/jx49z7do1oHi43NmzZ/XOWedJHPdBn2dRUREJCQn4+PhQr149dXu1atWIiIhg8uTJANjZ2REbG8uGDRvUXqaxY8eyfv36cucrKYpCfHw83t7eer2Mtra29OzZs8z3vPLKK3qvt23bhoODA35+fnrbdZ9/eXMMH+Sdd95Rhx0DDB8+HEVR2L59OwBxcXFotVpsbGz0rr2uHbpyO3fupEmTJri4uKh1Va9ene7duz90W0aMGKH3OiAgACge4gvQrVs3DAwM9IYBx8TEYGFhga+vb7n1GhgU/9/y/cNEH0VISAjbtm2jatWq6rasrCxMTU2B4h5LKP5+nDt3jrlz56pDV318fIiJiVGz2drZ2ZGdnc20adM4e/YsUDy88ZdffuHVV1997DY+6n3i7u6OlZXVYx9PCCFeNDIeQYinbObMmdSoUQMofmCzsrLCwcFBfcCC4vk6ADNmzChzXg1AamoqdnZ2QPED6IPoHtLKGkL48ssvA8XD6Dw9Pf+0zvu3GxoaAqWHeBkaGurNMTIxMWHjxo0kJiZy4cIFLl68qAZvyn1zke6vS5etUzfH7sqVKwB6AYyOm5sb8L9rGBQUVOZ5AKSlpVGzZs1y9+t07NiR6dOns23bNgYOHMjWrVuxs7OjSZMmalt0nsRxH/R53r59m7t375Z5/rrhplD8MD927FiCg4P5+OOP8fDwoFOnTvTt25fKlSuXW/ft27epX79+qX2678mD2nv58mW8vb1LlbOxscHKyqrUNXtY9w9d1Z2/rr7k5GTu3btH69aty3x/amqqWr7k0GGd8s6vLPeX1QXduvvM1taW5s2b88svvzBx4kSKiorYvHkzHTt2xNzcvNx6dT9e3Lp167GXENBoNKSnp7Nw4UJOnTpFcnIyKSkp6n2mu4/ee+89Dh06xJw5c5gzZw4NGzakQ4cO+Pv7q+czePBgEhMTCQ8PJzw8nDp16uDr68sbb7yBs7PzY7UPHv0+edAQUiGE+KeRIE6Ip6xp06YPfDjTPWQFBgaWm7yg5EOkLpD6M/cHSmXtMzY2fmCd5c1FKdlDcr+8vDxGjBjB/v37admyJa1bt2bYsGG0aNGC9u3blyqv640oj+76/Fk5XZnQ0NByr3d585LuV6dOHZydnYmPj2fgwIFs2bKFTp06lXnOT+K4D/o8db00D7pOrVu3Zvv27ep/SUlJfP7556xYsYINGzaU+WCs6/0oa5mLkj80lHR/O/7su1ZUVKT3PStLeb1Q979Pd61116uwsBAvLy/ef//9Mt+vCwg0Gk2ZSzv8Wbvvd/9nr3tvyc+uR48eTJ48mcOHD3Pv3j1u3LjBa6+99qf1enp6EhUVxeHDh8v9/hw9epQZM2YwdOjQUr2dULxG5OjRo6lZsyatWrVSFwtPTExk4cKFajk7Ozuio6PZvXs38fHx/Prrr2pSoWXLltGiRQssLS0JDw/n0KFDbN26lZ07d7J69Wq+//57ZsyYQY8ePR76mpX0qPfJw/yNE0KIfxIJ4oR4BtWuXRsoHpbYpk0bvX1HjhwhIyNDbxjho9R57ty5Uvt02eB0PXtPWmxsLHv27OGzzz7jjTfeULfrhic+qpdeegmAixcvllo/LTg4mKZNm6rna21tXeoa7t69m6Kiokdaj8/Pz49FixZx7tw5jh49qg59vd+TPm5ZqlWrhpmZGRcvXiy1b+nSpdy8eZOgoCB+//137Ozs6N69O927d6eoqIjly5czY8YMYmJi1CGAJVWvXh0LCwsuXLhQal9ZxytL7dq1y/ye3bhxg6ysLPXz0wV/eXl5euVu3rxZZr2XLl2iYcOG6mtdG3U9crVr1yY7O7vUdc/IyOD//u//1HJ16tQp8/wuXbr0EGdX7MqVKzg6OqqvdfdQyWGwXbp0YerUqWzbto2cnByqVq36wPX+fHx8MDExYd26deUO74yOjmbPnj0MGTKkzP2hoaHUq1eP9evXY2FhoW6/PzGNLuOqbng3wP79+xk6dCirV6+mRYsWnD9/nszMTDw8PPDw8OCDDz7gjz/+UBMsPW4Q9zTuEyGEeJHJnDghnkGurq7Y2NiwevVqvfliWVlZ6vC4R/1l2sXFBRsbG9asWUNWVpZenREREdjY2ODq6vrEzqGk27dvA+g9gAOsWrUK4JHThbu6umJtbc2GDRv0AoADBw6wYcMG7t69S5s2bTA1NWXJkiVq6ncoDhxHjx7NV1999ae9h/fz8/NTU/ZXqVKF5s2bl1nuUY6rC2JKLsXwMIyMjGjbti0JCQnqEEEoDlaWLl1KcnIyt2/fpn///no9LwYGBjRp0kTv2PczMDCgQ4cO7Ny5Uy+oycjI4Oeff36o9vn6+nLu3Dm2bt2qt33RokUAau+rbujgyZMn1TJZWVnqvLL76bIq6ixfvhwjIyN16YUOHTpw8uRJduzYoVdu/vz5BAYGqqnzO3fuzJkzZ9i5c6daJjMzk+jo6Ic6P4CoqKhSbdFoNHrLQFhZWeHj40NCQgIJCQl06dLlgb2Q1atXZ+jQofz2228sXbq01P7ExEQiIiJwcnIqc0goFN9vtWrV0gvgUlNTiYuLA/7X0xkYGMiHH36o1/PZuHFjjI2N1e/HtGnTGD16tN7foZdffhkrK6sH9gT/mSd9fwohxD+N9MQJ8QwyNjbm448/ZuzYsfTp04c33ngDU1NT1q5dS0pKCl999dUjp9guWWffvn3VHrF169Zx/fp1wsLC/tJD2Z9p06YNRkZGfPjhhwwaNAgjIyO2b99OYmIixsbGpRKbPIiJiQmTJk1i4sSJDBw4kJ49e5Kdnc2qVatwcHDA398fCwsLxo0bx+eff07//v3p2bMnBQUFREREkJuby8SJEx/pmI0aNaJ27dokJibSp0+fcoNoa2vrhz6ubjhjWFiYOsz0YY0fPx5/f3/8/f3VZDVRUVHcvXuXsWPHUrNmTXr06EFERAQ5ORlpf7MAAATdSURBVDl4enpy+/ZtwsPDqVGjBl27di237sDAQBISEujfvz8BAQGYmJjwww8/qMlRHvRw/e677xIXF8fYsWMZOHAg9evXZ9euXcTFxdG5c2d8fHyA4sB42rRpTJ06lStXrmBiYkJUVJRe8FHSjz/+SFZWFk2bNuXXX39l+/btvPfee2qvju6477//PgMGDMDR0ZH9+/cTHR1Nu3btaNeuHVCcEGXjxo2MGTOGoUOHYm1tTWRk5CMNp/zpp5/IysrCzc2NhIQEtm/fzr/+9a9S8xRfe+01AgMDgeKA6GGMGTOGM2fOMGPGDLZu3UqnTp0wNTXl4MGDxMTEUL16db755pty79d27doRGxvLJ598QpMmTbh8+TJRUVHk5OQA/0skNGLECCZPnsywYcN49dVXURSF6OhocnNz1SUmhg8fzttvv82gQYPo3bs3pqambN26leTkZL788suHvl73e5T7RAghRGkSxAnxjOrSpQvLli1j/vz5zJs3DwMDAxwdHZk/f/6fZrd7mDrnzZvHd999h5GREe7u7nz22Wd66849aVqtlrCwMObOncusWbOoVKkSjo6OLF++nIiICPbs2UN+fv4DeylK6tWrF5aWlixcuJCvv/4aKysrfH19GT9+vBoEDBs2DFtbW5YvX87s2bMxMzPDxcWFmTNnqtn3HkXHjh1ZtWpVuQt86zzscQcOHMiuXbtYsmQJR48efaQgzsHBgcjISGbNmsWSJUswMDDAzc2NL7/8Uh3mFxoaSt26dYmJiSEmJgZzc3Nat25NUFDQnyaKsLe3Jzw8nC+//JKFCxdiampK7969MTQ0ZOnSpQ8c5la1alUiIyP55ptviI2N5c6dO9StW5cPP/yQYcOGqeWsra1ZvHgxX3/9NWFhYVSrVo1+/frx8ssvl5nwYu7cuXz33XfExcVRt25dQkND6devX6njhoWFsXnzZiIjI6lVqxajR4/mnXfeUYMeS0tLIiIimDlzJpGRkRQWFtKtWzccHR0fOtBavHgx06ZN4+eff8bW1pbg4GC9c9Px9fXF0tISS0vLh77HTE1N+e6779i4cSPr169n6dKlZGRkYGtrS0BAACNHjvzTzy8kJAQLCwu2bdtGdHQ0dnZ29O7dm06dOqnfucaNG+Pv74+xsTGrVq1i1qxZFBUV4erqyuLFi2nZsiVQnHl0/vz5LFy4kHnz5pGbm4ujoyOzZs16pGyeZXnS96cQQvyTaJRH+elRCCHEC+/WrVtYW1uX6nELDQ1lzZo1HD58+JEC7r9qzpw5zJ07l/j4+MfO2FhR8vLyaNOmDf3792fChAkV3RwhhBAvCJkTJ4QQQk9gYKCaCEUnJyeH7du34+zs/FQDuOddTEwMmZmZ9OnTp6KbIoQQ4gUiwymFEELo6dWrF5MnT+add96hY8eO5ObmsnHjRq5evcqUKVMqunnPhWXLlnHgwAF27tyJr69vqTXuhBBCiL9CgjghhBB6/P39MTU1ZdWqVcycORMDAwNcXV1ZsWIFLVq0qOjmPRcKCwtJTEzE3d39oefZCSGEEA9L5sQJIYQQQgghxHNE5sQJIYQQQgghxHNEgjghhBBCCCGEeI5IECeEEEIIIYQQzxEJ4oQQQgghhBDiOSJBnBBCCCGEEEI8RySIE0IIIYQQQojnyP8Dben2q8GygxsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')\n",
    "### Balkendiagramm\n",
    "df = df_text\n",
    "\n",
    "import scipy as sp\n",
    "from matplotlib import pyplot\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set(style=\"whitegrid\",font_scale=1.5)\n",
    "sns.set_palette(sns.color_palette(\"Set2\"))\n",
    "\n",
    "a4_dims = (11.7, 8.27)\n",
    "fig, ax = pyplot.subplots(figsize=a4_dims)\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"value\", hue=\"variable\", data=df, alpha=1)\n",
    "\n",
    "ax.set_xticklabels(ax.get_xticklabels(), rotation= 50)\n",
    "\n",
    "\n",
    "sns.set_palette(sns.color_palette(\"Set3\"))\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"paper\", hue=\"variable\", data=df, alpha=0.7)\n",
    "handles, _ = ax.get_legend_handles_labels()\n",
    "ax.legend(handles, [\"Precision\", \"Recall\", \"F1\"], bbox_to_anchor=(1.05,1),loc=2, borderaxespad=0.)\n",
    "ax.set(xlabel='Performance Metrics grouped by Classifier')\n",
    "ax.set(ylabel='value')\n",
    "ax.set_title('Classifier comparison for text data')\n",
    "\n",
    "fig.savefig(\"feature_clf_text.pdf\",bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 842.4x595.44 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')\n",
    "### Balkendiagramm\n",
    "df = df_audio\n",
    "\n",
    "import scipy as sp\n",
    "from matplotlib import pyplot\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set(style=\"whitegrid\",font_scale=1.5)\n",
    "sns.set_palette(sns.color_palette(\"Set2\"))\n",
    "\n",
    "a4_dims = (11.7, 8.27)\n",
    "fig, ax = pyplot.subplots(figsize=a4_dims)\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"value\", hue=\"variable\", data=df, alpha=1)\n",
    "\n",
    "ax.set_xticklabels(ax.get_xticklabels(), rotation= 50)\n",
    "\n",
    "\n",
    "sns.set_palette(sns.color_palette(\"Set3\"))\n",
    "sns.barplot(ax = ax, x=\"Classifier_Modality\", y=\"paper\", hue=\"variable\", data=df, alpha=0.7)\n",
    "handles, _ = ax.get_legend_handles_labels()\n",
    "ax.legend(handles, [\"Precision\", \"Recall\", \"F1\"], bbox_to_anchor=(1.05,1),loc=2, borderaxespad=0.)\n",
    "ax.set(xlabel='Performance Metrics grouped by Classifier')\n",
    "ax.set(ylabel='value')\n",
    "ax.set_title('Classifier comparison for audio data')\n",
    "\n",
    "fig.savefig(\"feature_clf_audio.pdf\",bbox_inches=\"tight\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}