ProjektEkonometria.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "source": [
    "<h1>Projket ekonometrtia</h1>\n",
    "<h2>Plik wykonywalny </h2>\n",
    "<h2>Mateusz Jakubczak</h2>\n",
    "<h3>Pobieranie danych"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [],
   "source": [
    "import csv\n",
    "import warnings\n",
    "from operator import itemgetter\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import statsmodels.api as sm\n",
    "from dython.nominal import associations\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "from scipy import stats\n",
    "import math"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "#pare zmiennych globalnych\n",
    "with open(\"linki.csv\",newline=\"\") as csvfile:\n",
    "    reader = csv.reader(csvfile)\n",
    "    links2 = list(reader)\n",
    "\n",
    "#unpack list of list to list\n",
    "links =  [item for sublist in links2[1:] for item in sublist]\n",
    "\n",
    "znaczniki ={\"Cena\":\"_1svub _lf05o _9a071_2MEB_\"}\n",
    "\n",
    "lista_parametrów = [\"Cena\",\"Stan\",\"Typ telewizora\",\"Marka\",\n",
    "                    \"Technologia 3D\", \"Przekątna ekranu\",\n",
    "                    \"Format HD\",\"Rozdzielczość ekranu\",\n",
    "                    \"Liczba złączy HDMI\", \"Technologia HDR\",\n",
    "                    \"Klasa efektywności\",\"Pobór mocy\",\n",
    "                    \"Waga produktu\"\n",
    "                    ]\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [],
   "source": [
    "# uncoment to download data from links\n",
    "# \n",
    "# with open(os.path.join(os.getcwd(), 'parametry.csv'), 'wb') as csvfile:\n",
    "#      writer = unicodecsv.DictWriter(csvfile, fieldnames = lista_parametrów , delimiter = ';')\n",
    "#      writer.writeheader()\n",
    "# for link in links:\n",
    "#     parmaetry = []\n",
    "#     response = requests.get(link)\n",
    "#    # print(response)\n",
    "#     # znalazeinei ceny \n",
    "#     soup = BeautifulSoup(response.content,'html.parser')\n",
    "#     #nalezalo by zastapic lepszym wyszukaniem\n",
    "#     try:\n",
    "#         cena = soup.find(class_=znaczniki[\"Cena\"] )\n",
    "#         # przejsc do dolu dwa razy po pierwszym elemencie \n",
    "#         cena = cena.contents[0].contents[0]\n",
    "#     except:\n",
    "#         AttributeError\n",
    "#         print(link)\n",
    "#         continue\n",
    "#     parmaetry.append(cena)\n",
    "#     #zanjdowanie tabeli parametrow \n",
    "#     pattern = re.compile(r'Parametry')\n",
    "#     tabela = soup.find(\"h2\",text= pattern)\n",
    "#     #do gory dwa poziomy \n",
    "#     tabela = tabela.parent.parent\n",
    "#    \n",
    "#     for param in lista_parametrów[1:]:\n",
    "#         try:\n",
    "#             #najpier znajdz czy jest pole ktore nazywa sie dokladnie tak\n",
    "#             stan_field = tabela.find(text = param)\n",
    "#             #jesli nie ma takiego pola znajdz pierwsze które nazywa sie podobnie\n",
    "#             if stan_field == None:  stan_field = tabela.find(text = re.compile(rf\"{param}\"))\n",
    "#             #jesli nie znalioziona to wpisac -1\n",
    "#             if stan_field == None: stan_field = -1\n",
    "#             #up two nodes in tree then select content of second one\n",
    "#             stan_field = stan_field.parent.parent.contents[1].contents[0]\n",
    "#             try:\n",
    "#                 if stan_field.contents != None : stan_field = stan_field.contents[0]\n",
    "#             except:\n",
    "#                 AttributeError   #dla pól co zawieraja w sobie linka \n",
    "#         except:\n",
    "#             AttributeError \n",
    "#             print(f\"Nie ma takiego parametru, parametr {param}\") \n",
    "#             stan_field = -1\n",
    "#         parmaetry.append(stan_field)\n",
    "#     with open('parametry.csv', 'ab') as csvfile:\n",
    "#         writer = unicodecsv.DictWriter(csvfile, fieldnames = lista_parametrów , delimiter = ';')\n",
    "#         writer.writerow(dict(zip(lista_parametrów,parmaetry)))\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "outputs": [],
   "source": [
    "def print(*args):\n",
    "    __builtins__.print(*(\"%.2f\" % a if isinstance(a, float) else a\n",
    "                         for a in args))\n",
    "    \n",
    "\n",
    "def save_to_latex(content, name):\n",
    "    try:\n",
    "        latex = content.as_latex()\n",
    "    except:\n",
    "        latex = content\n",
    "    with open(f\"latex/{name}.tex\", \"w\") as f: \n",
    "        f.write(latex)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3>Pre-processing and cleaning "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Cena                    object\n",
      "Stan                    object\n",
      "Typ telewizora          object\n",
      "Marka                   object\n",
      "Technologia 3D          object\n",
      "Przekątna ekranu        object\n",
      "Format HD               object\n",
      "Rozdzielczość ekranu    object\n",
      "Liczba złączy HDMI       int64\n",
      "Technologia HDR         object\n",
      "Klasa efektywności      object\n",
      "Pobór mocy              object\n",
      "Waga produktu           object\n",
      "dtype: object\n",
      "Macierz czestości dla jakosci obrazu\n",
      "Format HD             -1  4K UHD  8K UHD  Full HD  HD Ready  inny\n",
      "Rozdzielczość ekranu                                             \n",
      "1366 x 768             2       0       0        0        12     2\n",
      "1920 x 1080            4       0       0       18         0     0\n",
      "3840 x 2160            2      34       0        0         0     0\n",
      "7680 x 4320            0       0       5        0         0     0\n",
      "inna                   0       0       0        0         2     0\n",
      "0.46  telewizorow nie ma podanej swojej wagi\n"
     ]
    }
   ],
   "source": [
    "df = pd.read_csv(\"parametry.csv\",delimiter=\";\")\n",
    "\n",
    "print(df.dtypes)\n",
    "#Cleaning\n",
    "df[\"Cena\"] = df[\"Cena\"].apply(lambda s :''.join([i for i in s if i.isdigit()]))\n",
    "df[\"Cena\"] = df[\"Cena\"].apply(pd.to_numeric)\n",
    "\n",
    "#dla technoligi 3d w przypdako brakow zakladmy \"nie\" oraz zmiana do postaci binarnej \n",
    "df[\"Technologia 3D\"] = df[\"Technologia 3D\"].apply(lambda s : \"nie\" if s != \"tak\" else s)\n",
    "df[\"Technologia 3D\"] = df[\"Technologia 3D\"].map({'tak': True, 'nie': False})\n",
    "df[\"Technologia 3D\"] = df[\"Technologia 3D\"].astype('category')\n",
    "\n",
    "#zmeinic przekonte na zmienna ciagłą \n",
    "df[\"Przekątna ekranu\"] = df[\"Przekątna ekranu\"].apply(lambda s :''.join([i for i in s if i.isdigit()]))\n",
    "df[\"Przekątna ekranu\"] = df[\"Przekątna ekranu\"].apply(pd.to_numeric)\n",
    "\n",
    "#polaczyc rozdzielczosc i format hd, poniewaz sa to pojecia rownowazne\n",
    "#na niektorych ofertach jest podane tylko jedno \n",
    "print(\"Macierz czestości dla jakosci obrazu\")\n",
    "print(pd.crosstab(df[\"Rozdzielczość ekranu\"],df[\"Format HD\"]))\n",
    "#potwierdza to macierz czestosci \n",
    "slownik =   {'3840 x 2160': '4K UHD',\n",
    "             '7680 x 4320': '8K UHD',\n",
    "             '1920 x 1080': 'Full HD',\n",
    "             '1366 x 768': 'HD Ready'}\n",
    "#first join\n",
    "df['jakosc'] = df[[ 'Format HD','Rozdzielczość ekranu']].agg(';'.join, axis=1)\n",
    "#then do operation on string\n",
    "df['jakosc'] = df['jakosc'].apply(\n",
    "    lambda s: s.split(sep = \";\")[0] if s.split(sep = \";\")[0] != \"-1\" and s.split(sep = \";\")[0] != \"inny\"\n",
    "    else slownik[s.split(sep = \";\")[1]])\n",
    "df[\"jakosc\"] = df[\"jakosc\"].astype('category')\n",
    "\n",
    "#zmienic liczbe zleczy hdmi na categorical \n",
    "#mamy tutaj cztery mozliwe wartosci \n",
    "#df[\"Liczba złączy HDMI\"]=df[\"Liczba złączy HDMI\"].astype(pd.to_numeric)\n",
    "\n",
    "#dla braku technologi HDR dac Nie \n",
    "df[\"Technologia HDR\"] =df[\"Technologia HDR\"].apply(lambda s: s if s != \"-1\" else \"Nie\")\n",
    "df[\"Technologia HDR\"]=df[\"Technologia HDR\"].astype('category')\n",
    "\n",
    "#klasy efektrywnosci -1 zamienic na inna \n",
    "df[\"Klasa efektywności\"] = df[\"Klasa efektywności\"].apply(lambda s: s if s != \"-1\" else \"inna\").astype(\"category\")\n",
    "\n",
    "#przeksztalcic pobor mocy do zmeinnej ciaglej \n",
    "df[\"Pobór mocy\"] = df[\"Pobór mocy\"].apply(lambda s :''.join([i for i in s if i.isdigit()]))\n",
    "df[\"Pobór mocy\"] = df[\"Pobór mocy\"].apply(pd.to_numeric)\n",
    "\n",
    "# usunac wage produktu, pozniewaz brakuje polowych danych \n",
    "print( df[\"Waga produktu\"].value_counts(\"-1\")[0], \" telewizorow nie ma podanej swojej wagi\")\n",
    "df.drop([\"Waga produktu\", \"Rozdzielczość ekranu\",\"Format HD\"], axis=1 , inplace=True)\n",
    "\n",
    "#podstawowe zmiany \n",
    "df[\"Stan\"] = df[\"Stan\"].astype('category')\n",
    "df[\"Typ telewizora\"] = df[\"Typ telewizora\"].astype('category')\n",
    "\n",
    "#marka dla malo papularne zastapic inną,(mało popularna znaczy ze ilosc ofert danej marki jest mniejsza od mediany dla marki)\n",
    "count = df[\"Marka\"].value_counts()\n",
    "popular = count > count.mean()\n",
    "df[\"Marka\"] = df[\"Marka\"].apply(lambda row: row if popular[row] else \"inna\").astype(\"category\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3>Statytyki opisowe "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "SD = <bound method Series.std of 0      9999\n",
      "1      4499\n",
      "2      3299\n",
      "3      3299\n",
      "4      4999\n",
      "      ...  \n",
      "76     7999\n",
      "77    18890\n",
      "78    16490\n",
      "79    14852\n",
      "80    24000\n",
      "Name: Cena, Length: 81, dtype: int64>\n",
      "SD = <bound method Series.std of 0     82\n",
      "1     65\n",
      "2     55\n",
      "3     55\n",
      "4     55\n",
      "      ..\n",
      "76    75\n",
      "77    75\n",
      "78    75\n",
      "79    75\n",
      "80    82\n",
      "Name: Przekątna ekranu, Length: 81, dtype: int64>\n",
      "SD = <bound method Series.std of 0     4\n",
      "1     4\n",
      "2     4\n",
      "3     4\n",
      "4     4\n",
      "     ..\n",
      "76    3\n",
      "77    4\n",
      "78    4\n",
      "79    4\n",
      "80    4\n",
      "Name: Liczba złączy HDMI, Length: 81, dtype: int64>\n",
      "SD = <bound method Series.std of 0     228\n",
      "1     118\n",
      "2     111\n",
      "3     111\n",
      "4     140\n",
      "     ... \n",
      "76    116\n",
      "77    410\n",
      "78    409\n",
      "79    272\n",
      "80    579\n",
      "Name: Pobór mocy, Length: 81, dtype: int64>\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAbkAAAEXCAYAAAAwQjq2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3df5xVVb3/8ddblBkQxAroInid8Wf5g5Tolylh2E0kMwXLHxV2fWRXqLDHxcS8lWmlooUP+xpFapoW1QioZVR2a7LQigF/DJCRAl7AoihBDAZy/Hz/WHviMJyBGWfmnDN73s/H4zz2Pnuvvc5a5wBv1t7rnK2IwMzMLI/2KXcDzMzMuotDzszMcsshZ2ZmueWQMzOz3HLImZlZbjnkzMwstxxyZr2ApBpJIWnfcrfFrJQccmYVRtIaSTskDW61/bEsqGrK0zKznschZ1aZVgPntTyRdBzQ7+VU5NGb9WYOObPKdBfwwYLnk4FvtTyRNEHSo5Kel7RW0lUF+1pOTV4k6f+An7euXNLEbMR4rKTDJP1c0t8kbZT0bUkHdmPfzErGIWdWmX4DHCDptZL6AO8D7i7Y/w9SCB4ITAAukfSeVnW8DXgt8M7CjZI+BFwPnBoRywAB1wIHZeUPBq7q6g6ZlYNDzqxytYzm3gE8Caxv2RER9RHRGBEvRcQTwFxSqBW6KiL+ERHbCrZdClwGjI2Ip7K6noqIByNie0T8FfhykbrMeiSfqzerXHcBDwG1FJyqBJD0JuA64FigL1AF1LU6fm2ROi8Dro6IdQV1DQVuBk4GBpL+8/tc13TBrLw8kjOrUBHxDGkCyunA/Fa7vwPcDxwcEYOAr5FOO+5SRZFq/wP4H0kTC7Zdm5UdGREHAO8vUpdZj+SQM6tsFwFvj4h/tNo+EPh7RDRJeiNwfjvrWw6cBtwi6d0Fdb0AbJI0nDTaM8sFh5xZBYuIpyOiociuKcDVkrYAnwG+34E6HwfeBXxD0njgc8AoYDPwALuPGs16LPmmqWZmllceyZmZWW455MzMLLcccmZmllsOOTMzyy1/GbwTBg8eHDU1NeVuhplZj7JkyZKNETGkFK/lkOuEmpoaGhqKze42M7O2SHqmVK/l05VmZpZbDjkzM8sth5yZmeWWQ87MzHLLIWdmZrnlkDMzs9xyyJmZWW75e3Kd0Lh+MzUzHih3MyrKmusmlLsJZmb/4pGcmZnllkPOym7RIhg5EqqqYNQoWLq0eLl774XDD4fqahg7Flav3rlv7Vo480zYf38YNAguuKAkTTezCueQs7JqaoKJE2HLFpg1CzZsgEmToLl513J//jOcey4ccADccAMsWQKTJ6d9EXDWWfDgg3DZZTBzJgwpya/imVml8zU5K6uFC1OwzZwJU6akMLvmGqivh3HjdpabOxe2b4crroBzzoHFi+Guu+Dpp+GZZ1LoXXklzJiRRoRS2bpkZhXEIzkrq5ZTjsOHp+WIEWm5alX7y61YkdbnzYP+/dNo7+abu6/NZtZz7DXkJL1QsH66pD9K+vfubVbRdpwhaYWkZZK+0GrfVZJC0uEF2z6RbRu9l3ovldS/u9ptHRORlnsbiRWW2749re+3HyxYALW1cOmlsHJl97XTzHqGdo/kJI0DvgKcFhH/12pft532LKj7JmBCRBwL3FqkaCNwbsHzScCKdrzEpYBDrkxqa9Ny3bq0XL9+5/amJtixY+/lWm7pN2FCmnwyYUIKwcKJKWbWO7UrnCSdDHwDOD0ins623QH8HTgBWCppC/BCRNyY7V8GvCsi1kj6NHABsBbYCCyJiBslHQbcAgwBtgIfjognW9cN/DewAxgBrI6IYv983QucCXxe0qHAZuCfBX2YDbwB6AfcExGflfRx4CDgF5I2ArOAq7ND+gF9I6K2Pe+RvTzjx8PQoTB7NgwcCLfdlkJr7FjYd1845hhYtixNOpkxA66/Pl3DW7AATjoJDjsMDjoo1TFvXpp9ec89MGAAnHBCuXtnZuXWnpFcFXAf8J6IeLLVviOBUyPiv9s6ODtdOJEUWGcDhacP5wAfi4jXA9OBrxarW9I+wO+B2yW1FTrPA2slHQucB3yv1f4rI2I0MBJ4m6SREXEz8CxwSkScEhH3R8TxEXE88DhwY5H+XCypQVJD89bNbXXb2qm6GurqUihNm5bCqq4O+vTZtdywYWnyyaZNMH16CrA77kj7+vVLwVZVBVOnputy8+enusysd2vPSO6fwMPARcC0VvvqIqJ590N2cRJwX0RsA5D0g2w5ADgRqNPOCzBVbdT9MWA5KQR/IOkUoBa4LCLOKTjmu6RTlu8ExgEfKtj3XkkXk/o8DDgaeKJYgyV9EtgWEbe03hcRc0jhTNWwI2Ivfbd2GDMGGht33x6t3t2zz06PYk4+uXgdZta7tSfkXgLeC/xM0qci4osF+/5RsP4iu44Mq7NlW1MI9gE2ZaOmYgrrficwMyLqJV0NPAD8jt1Haz8AbgAaIuL5lvDMRn/TgTdExHPZ6dBqisiuPZ4DjGmjXWZm1kO0a+JJRGwF3gVcIOmiNoqtAUYBSBpFGmkB/Bo4Q1J1NnqbkNX5PLBa0jnZMZL0ujbqfhR4v6R9IuL7wB+B80lhV9jObcDlwBdaHX8AKTQ3S3o1ML5g3xZgYNaGQ0ijxfe2jDzNzKznavesyIj4u6TTgIeySRqtzQM+KOkxYDGwMjtusaT7Sde4ngEaSJNCIE1GmS3pf4D9SKcbHy9S9xdIMzuXSdoG/BL4OvAdSRNbtfO7Rdr+uKRHSac8VwGLCnbPARZK+hNQD7wKWJCNAp+NiNP3+MaYmVnFUrS+8NEdLyINiIgXsu+jPQRcHBFt/EJhz1E17IgYNvmmcjejovguBGa2N5KWZBMBu12pftZrjqSjSdfB7sxDwJmZWeUrSchFxPmleB0zM7NC/u1KMzPLLYecmZnllm+10wnHDR9EgydamJlVLI/kzMwstxxyZmaWWw45MzPLLYecmZnllkPOzMxyyyFnZma55ZAzM7PccsiZmVluOeTMzCy3HHJmZpZbDjkzM8sth5yZmeWWQ87MzHLLIWdmZrnlkDMzs9xyyJmZWW455MzMLLd8Z/BOaFy/mZoZD5S7GRVlje+UbmYVxCM5MzPLLYecld2iRTByJFRVwahRsHRp8XL33guHHw7V1TB2LKxevXPf2rVw5pmw//4waBBccEFJmm5mFc4hZ2XV1AQTJ8KWLTBrFmzYAJMmQXPzruX+/Gc491w44AC44QZYsgQmT077IuCss+DBB+Gyy2DmTBgypPR9MbPK42tyVlYLF6ZgmzkTpkxJYXbNNVBfD+PG7Sw3dy5s3w5XXAHnnAOLF8Ndd8HTT8Mzz6TQu/JKmDEjjQilsnXJzCqIR3JWVi2nHIcPT8sRI9Jy1ar2l1uxIq3Pmwf9+6fR3s03d1+bzazn6BEhJ6lZ0mOSlkmqk9RfUo2kZW2Uv1rSqdl6vaTR2fqPJB1YyrZbx0Sk5d5GYoXltm9P6/vtBwsWQG0tXHoprFzZfe00s56hR4QcsC0ijo+IY4EdwH/tqXBEfCYiflZk++kRsam7GmkdV1ubluvWpeX69Tu3NzXBjh17L1dTk9YnTEiTTyZMSCFYODHFzHqnnnhN7lfAyGy9j6RvACcC64EzI2KbpDuAH0bEPYUHSloDjAYGAD8GfgucAKwEPhgRWyVdB7wbeBH4aURM7/4u9V7jx8PQoTB7NgwcCLfdlkJr7FjYd1845hhYtixNOpkxA66/Pl3DW7AATjoJDjsMDjoo1TFvXpp9ec89MGAAnHBCuXtnZuXWU0ZyAEjaFxgPNGabjgBuiYhjgE3AxA5UdxQwJyJGAs8DUyS9EjgLOCbb/vkibbhYUoOkhuatmzvRG4P0dYC6uhRK06alsKqrgz59di03bFiafLJpE0yfngLsjjvSvn79UrBVVcHUqem63Pz5qS4z6916ykiun6THsvVfAbcBBwGrI6Jl+xKgpgN1ro2IRdn63cDHgZuAJuBWSQ8AP2x9UETMAeYAVA07IjrYDytizBhobNx9e7R6d88+Oz2KOfnk4nWYWe/WU0JuW0QcX7hBaWbC9oJNzUC/DtTZOqAiIl6U9EZgHHAu8FHg7R1vrpmZVYKeEnLd4d8lvSUiHgHOA34taQDQPyJ+JOk3wFPlbaKZmXVGbw653wOTJX0d+CMwGxgE3CepGhDwiTK2z8zMOqlHhFxEDCiybQ1wbMHzGwvWLyxYH1uwXgOQjdheiojWX0XYCryxa1ptZmbl1qNmV5qZmXVEjxjJdbXWo0AzM8snj+TMzCy3euVIrqscN3wQDb4TtplZxfJIzszMcsshZ2ZmueWQMzOz3HLImZlZbjnkzMwstxxyZmaWWw45MzPLLYecmZnllkPOzMxyyyFnZma55ZAzM7PccsiZmVluOeTMzCy3HHJmZpZbDjkzM8sth5yZmeWWQ87MzHLLdwbvhMb1m6mZ8UC5m1FR1vhO6WZWQTySMzOz3HLIWdktWgQjR0JVFYwaBUuXFi93771w+OFQXQ1jx8Lq1Tv3rV0LZ54J++8PgwbBBReUpOlmVuEcclZWTU0wcSJs2QKzZsGGDTBpEjQ371ruz3+Gc8+FAw6AG26AJUtg8uS0LwLOOgsefBAuuwxmzoQhQ0rfFzOrPL4mZ2W1cGEKtpkzYcqUFGbXXAP19TBu3M5yc+fC9u1wxRVwzjmweDHcdRc8/TQ880wKvSuvhBkz0ohQKluXzKyCeCRnZdVyynH48LQcMSItV61qf7kVK9L6vHnQv38a7d18c/e12cx6jtyEnKQrJS2X9ISkxyS9qdxtso6LSMu9jcQKy23fntb32w8WLIDaWrj0Uli5svvaaWY9Qy5OV0p6C/AuYFREbJc0GOhb5mZZO9TWpuW6dWm5fv3O7U1NsM8+0Lfvnstt3pzWJ0xIk09+8xtobEyjvyOPLE0/zKwy5SLkgGHAxojYDhARGwEkjQNuJPVzMXBJFoJrgDuBM4D9gHOAlcAfgBMj4q+S9sm2vbmlPut648fD0KEwezYMHAi33QY1NWn25L77wjHHwLJladLJjBlw/fXpGt6CBXDSSXDYYXDQQamOefPS7Mt77oEBA+CEE8rdOzMrt7ycrvwpcLCklZK+KultkqqBO4D3RcRxpKC7pOCYjRExCpgNTI+Il4C7gZbJ56cCj7cOOEkXS2qQ1NC8dXM3dyv/qquhri6F0rRpKazq6qBPn13LDRuWJp9s2gTTp6cAu+OOtK9fvxRsVVUwdWq6Ljd/fqrLzHo3RcvFjR5OUh/gZOAU4CPAtcDEiBiT7R8HTI2Is7OR3FsjYn127e4LEXGqpIOB+yJilKTvAndHxA/bes2qYUfEsMk3dXPPehb/4omZ7Y2kJRExuhSvlZfTlUREM1AP1EtqBCbv5ZBsugLNZO9DRKyVtEHS24E3sXNUZ2ZmPVAuTldKOkrSEQWbjgc2ADWSDs+2fQD4ZTuqu5V02vL7WXCamVkPlYuQAwYAd0paIekJ4GhgBvAhoC4b2b0EfK0ddd2f1ffN7mqsmZmVRi5OV0bEEuDEIrv+F9htjl1E1BSsNwBjC3a/jjTh5MmubaWZmZVaLkKuq0iaQZqB6WtxZmY5kJfTlV0iIq6LiEMi4tflbouZmXWeQ87MzHLLpys74bjhg2jw98LMzCqWR3JmZpZbDjkzM8sth5yZmeWWQ87MzHLLIWdmZrnlkDMzs9xyyJmZWW455MzMLLcccmZmllsOOTMzyy2HnJmZ5ZZDzszMcsshZ2ZmueWQMzOz3HLImZlZbjnkzMwstxxyZmaWW74zeCc0rt9MzYwHyt0Ms71a4zvYWy/lkZyZmeWWQ87Milq0CEaOhKoqGDUKli4tXm7SJHjFK0CCj350133Sro/3vKf7221WyCFnZrtpaoKJE2HLFpg1CzZsSGHW3Lx72aoqOOustuuaOBHmzk2P6dO7r81mxTjkzGw3CxemYJsyJT0uughWr4b6+t3Lfvvb8MEPtl3X0UfDGWfAuefCSSd1W5PNinLImdluVq9Oy+HD03LEiLRctarjdX3+8zBgABxyCPzwh13TPrP26nUhJ+mFNra/X9ITkpZLelzSrZIOLHX7zCpRRFpKHTvu8sth/nyYMweeew7OOw+2bu369pm1xV8hACSdBnwCGB8R6yX1ASYDrwY2lbVxZmVQW5uW69al5fr1O7c3NcE++0Dfvnuv57rrdq7/+Mcp8NauhaOO6tr2mrXFIZdcCUyPiPUAEdEM3F7eJpmVz/jxMHQozJ4NAwfCbbdBTQ2MHQv77gvHHAPLlqWy3/seNDSk9RUr4NZbYcIEePRRuPvudMxzz6XrfEOG7AxQs1JwyCXHAG1MkN6VpIuBiwH6HDCkO9tkVjbV1VBXB1OnwrRpKdS+8Q3o02f3spdfDs88k9Z/8Yudj0MOgT/9CT75yTQrc/Ro+NKX2jcCNOsqDrlWJB0H3AUMBD4VEd8r3B8Rc4A5AFXDjojSt9CsNMaMgcbG3bdHqz/1a9a0XccvftGlTTLrsF438aQNy4FRABHRGBHHAwuBfmVtlZmZdYpDLrkWuFHSiIJtDjgzsx6uN56u7C9pXcHzL0fElyUNARZmMys3AcuAn5SlhWZm1iV6XchFRNHRa0TcCdxZ4uaYmVk38ulKMzPLLYecmZnllkPOzMxyq9ddk+tKxw0fRIPvuGxmVrE8kjMzs9xyyJmZWW455MzMLLcccmZmllsOOTMzyy2HnJmZ5ZZDzszMcsshZ2ZmueWQMzOz3HLImZlZbjnkzMwstxxyZmaWWw45MzPLLYecmZnllkPOzMxyyyFnZma55ZAzM7Pc8p3BO6Fx/WZqZjxQ7maYGbDmugnlboJVII/kzMwstxxyZtZrLVoEI0dCVRWMGgVLl+5e5pFH4MQT4cAD02PiRPjrX0vfVnt5HHJm1is1NaXA2rIFZs2CDRtg0iRobt613MqVMHgwXH89nH46zJ8Pn/xkedpsHeeQM7NeaeHCFGxTpqTHRRfB6tVQX79rufPOg/vvh498BL7+9bRt+fKSN9deJoecmfVKq1en5fDhaTliRFquWrVrub59d67/5CdpOWZM97bNuk7uQk7Sw+Vug5n1PBFpKRXfv2gR/Od/wutfD1ddVbJmWSflLuQi4sRyt8HMKl9tbVquW5eW69fv3N7UBDt27Cz70ENw2mlw2GFpNDdgQGnbai9f7kJO0gvZcqykekn3SHpS0rel9H80SWskfU7SUkmNkl6TbX+jpIclPZotjypnX8ys+4wfD0OHwuzZ6XHbbVBTA2PHQr9+abYlpBmX48enCSkf/jA8+CD84AflbLl1RO5CrpUTgEuBo4FDgbcW7NsYEaOA2cD0bNuTwJiIOAH4DPDF1hVKulhSg6SG5q2bu7XxZtZ9qquhri6NyqZNS4FXVwd9+uxa7oknYOtW2LYNpk5NE1E+9rHytNk6Lu+/ePK7iFgHIOkxoAb4dbZvfrZcApydrQ8C7pR0BBDAfq0rjIg5wByAqmFHRLe13My63Zgx0Ni4+/Yo+Jt94YXpYT1T3kdy2wvWm9k11LcX2X4N8IuIOBY4A6ju9haamVm3yXvIddQgILv8zIVlbIeZmXUBh9yuZgLXSloE9NlbYTMzq2y5uyYXEQOyZT1QX7D9owXrNQXrDcDYbP0R4MiC6j7dnW01M7Pu5ZGcmZnllkPOzMxyyyFnZma55ZAzM7Pcyt3Ek1I6bvggGq6bUO5mmJlZGzySMzOz3HLImZlZbjnkzMwstxxyZmaWWw45MzPLLYecmZnllkPOzMxyyyFnZma55ZAzM7PccsiZmVluOeTMzCy3HHJmZpZbDjkzM8sth5yZmeWWQ87MzHLLIWdmZrnlkDMzs9zyncE7oXH9ZmpmPFDuZpiZldSa6yaUuwnt5pGcmZnllkPOzMy63aJFMHIkVFUBLH+txKjWZSTeIvGwxKbsMU9iSLZPEtdKPCvRJPGkxPv29roOOTMz61ZNTTBxImzZArNmAQzeD7hHok+rokcCG4HLgR8BZwMzs32nAjOAPwGXAcOBOyT229NrO+TMzKxbLVwIGzbAlCnpAd/ZCNQCY1sVnRvBuyP4OvCRbNsx2bIlr54GHgQ2A1uAl/b02g45MzPrVqtXp+Xw4S1b1u3IVg4tLBfBjoKn78yWD2XLnwK3AOcAvwdeBZwfQfOeXrvkISfpSknLJT0h6TFJbyp1G8zMrJzUshJF94q3ArcDS4Crss1HAe8nhd3ZwAbS6cr99/RKJQ05SW8B3gWMioiRpHOsa0vZBjMzK63a2rRct65ly/C+2cpqiWqJludIjAF+TDot+c4IXsh2vRsYBNwVwQLgZ6Trckfv6bVLPZIbBmyMiO0AEbExIp6V9BlJiyUtkzRHkgAk1UuaJekhSb+X9AZJ8yX9UdLnszL7S3pA0uPZ8e/Ltq+RNDhbHy2pPlu/StLtWd2rJH28pXGSPi3pSUkPSporaXpp3x4zs/wZPx6GDoXZs9MDzh8MrAHqgW3AUoBsxuVCoA/wDeAdEmdk1TydLS+R+DBpwLQDWL2n1y51yP0UOFjSSklflfS2bPv/i4g3RMSxQD9S41vsiIgxwNeA+4CpwLHAhZJeBZwGPBsRr8uO/3E72vEa0vneNwKflbSfpNHAROAE0lB4dLEDJV0sqUFSQ/PWzR3svplZ71NdDXV1MGAATJsG8LcXgXOKXE8bCfQn5cAtwFzgK9m++aSZljXZtr8D749g455eu6QhFxEvAK8HLgb+CnxP0oXAKZJ+K6kReDs7Z9MA3J8tG4HlEfGnbCS4Cjg4236qpOslnRwR7UmeByJie0RsBP4CvBo4CbgvIrZFxBbgB230YU5EjI6I0X36D+rgO2Bm1juNGQONjbBjB8DRKyJoAIhAERybrd+RPS981GT7IoLLIxgeQXUER0dQt7fXLfnPekVEM2mIWp+F2kdI6T06ItZKugqoLjhke7Z8qWC95fm+EbFS0uuB04FrJf00Iq4GXmRniBfWR6t6mknvgzAzs1wp9cSToyQdUbDpeOAP2fpGSQOASR2s8yBga0TcDdwI//oW/RrSqBHSaci9+TVwhqTqrB0958fZzMysqFKP5AYAX5F0IGmk9RTp1OUm0mnHNcDiDtZ5HHCDpJeAfwKXZNs/B9wm6VPAb/dWSUQslnQ/8DjwDNBA+rKhmZn1UIoo+jWFXknSgIh4QVJ/0hcQL46IpW2Vrxp2RAybfFPpGmhmVgE6excCSUsioujkvq7mW+3sao6ko0nX8O7cU8CZmVnlc8gViIjzy90GMzPrOv7tSjMzyy2P5DrhuOGDaOhBd8g1M+ttPJIzM7PccsiZmVluOeTMzCy3HHJmZpZbDjkzM8sth5yZmeWWQ87MzHLLIWdmZrnlH2juBElb2HmroLwaDHu+824OuI/54D72HIdExJBSvJB/8aRz/lCqX9IuF0kN7mPP5z7mQ2/oY1fz6UozM8sth5yZmeWWQ65z5pS7ASXgPuaD+5gPvaGPXcoTT8zMLLc8kjMzs9xyyJmZWW455Nog6TRJf5D0lKQZRfZL0s3Z/ickjWrvsZWik31cI6lR0mOSGkrb8vZrRx9fI+kRSdslTe/IsZWik33My+d4QfZn9AlJD0t6XXuPrRSd7GOP+BzLIiL8aPUA+gBPA4cCfYHHgaNblTkdWAgIeDPw2/YeWwmPzvQx27cGGFzufnRBH4cCbwC+AEzvyLGV8OhMH3P2OZ4IvCJbH5/Tv49F+9hTPsdyPTySK+6NwFMRsSoidgDfBc5sVeZM4FuR/AY4UNKwdh5bCTrTx55ir32MiL9ExGLgnx09tkJ0po89RXv6+HBEPJc9/Q0wor3HVojO9NH2wCFX3HBgbcHzddm29pRpz7GVoDN9BAjgp5KWSLq421rZOZ35LPL0Oe5JHj/Hi0hnIF7OseXSmT5Cz/gcy8I/61Wcimxr/V2Ltsq059hK0Jk+Arw1Ip6VNBR4UNKTEfFQl7aw8zrzWeTpc9yTXH2Okk4hBcBJHT22zDrTR+gZn2NZeCRX3Drg4ILnI4Bn21mmPcdWgs70kYhoWf4FWEA63VJpOvNZ5OlzbFOePkdJI4FbgTMj4m8dObYCdKaPPeVzLAuHXHGLgSMk1UrqC5wL3N+qzP3AB7MZiG8GNkfEn9p5bCV42X2UtL+kgQCS9gf+A1hWysa3U2c+izx9jkXl6XOU9O/AfOADEbGyI8dWiJfdxx70OZaFT1cWEREvSvoo8BPSrKfbI2K5pP/K9n8N+BFp9uFTwFbgQ3s6tgzd2KPO9BF4NbBAEqQ/Q9+JiB+XuAt71Z4+Svo3oAE4AHhJ0qWkWW3P5+VzbKuPpNu25OJzBD4DvAr4atafFyNidM7+PhbtIz3k72O5+Ge9zMwst3y60szMcsshZ2ZmueWQMzOz3HLImZlZbjnkzMwstxxyZvYvSnedOK/Vtj6SpkqqLle7zF4uh5xZBZD0b5K+K+lpSSsk/UjSkV1Q74WSDmpn2bcB2yNibqtdXwKWR0RTZ9tjVmr+npxZmSl9i/dh4M7sS79IOh4YGBG/6mTd9aTb6/geY9YreSRnVn6nAP9sCTiAiHgMeEd2E8zHJK2X9E0ASe+X9Lts+9ez04l9JN0haVl288xPSJoEjAa+nZXtJ2mcpEezMrdLqsrqvC4bQT4h6cZs26slLZD0ePY4sfRvjVnneCRnVmaSPg7URsQn2tg/CPgV6WfVtgIzgbMj4p+Svkq6t9hy4LqIeEd2zIERsalwJJddU/sjMC4iVkr6FrAU+BbwCPCaiIiCY78HPBIRN0nqAwyIiM3d+FaYdTmP5MwqWHYq89vArIhYAowDXg8slvRY9vxQYBVwqKSvSDoNeL5IdUcBqwt+3PdOYExWtgm4VdLZpCAFeDswGyAimh1w1hM55MzKbzkpuIq5ClgXEd/Mnot07e747HFURFyV3TH6dUA9MJV0O5bWit2zjIh4kXRrlnnAewD/uK/lhkPOrPx+DlRJ+nDLBklvkPRZ4B3AxwvK/i8wSenmmEh6paRDJA0G9omIecCngVFZ+S3AwGz9SaBG0uHZ8w8Av5Q0ABgUET8CLgWOL1qFBrkAAAC3SURBVHitS7LX6SPpgC7ttVkJ+JqcWQXIpvnfRBrRNQFrgP5ALbApK3Z/RHxG0vuAK4BqoC9wPrAN+CY7/+N6RUQslDQR+GK2/y3AicCNpFuyLCaF2CuB+7L6BNwYEXdKejUwh3Q6tBm4JCIe6a73wKw7OOTMeihJnyQF35PlbotZpfLpSrOeawdwY1d8adwsrzySMzOz3PJIzszMcsshZ2ZmueWQMzOz3HLImZlZbjnkzMwst/4/bjbuT9QhHLwAAAAASUVORK5CYII=\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEXCAYAAACzhgONAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAVf0lEQVR4nO3de7hddX3n8fcHIgQMIBiQmybxhooCAjrehsKgHQF5FEkVxKqtTx1B62W0CjOO0vEy8VJx2irWK6gMUkYRK2C1KtVBUBIauVi03JTgBeRmIgZJ+M4fa6XZOeb8shPOOXsH3q/n2c++rO9e63tWcs5n/9Zv77VTVUiSNJktRt2AJGm8GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKHS/k+S0JO+ahvXOT1JJZk3Buq5KcvAUtCVNO4NCI5dkxcDl3iS/Hbh/3Kj7mw5VtXdVXbixz0syN8lFSW5NckeSi5M8c2D5yUnuSbK8v/w4yd8m2W1KfwA9oBgUGrmqmrPmAvwUOHLgsTNG3d+YWQH8KbAzsCPwXuAfJoxyzqqq7YCdgKOAXYElhoU2lUGhsZVkiyQnJrm2fwX990l2Glj+rCTf7V9Z35jkFQNP3zHJef2r6u8ledTA8yrJq5P8W5Lbk3w4SQa2+bYkP0lyc5LPJNlhkv52T/LlJLcluSbJnw0s2ybJ6f36/zXJW5IsG1h+Q5Jn97ef2o8M7kjy834EsNX6tllVK6vqR1V1LxBgNV1g7LSe2nuq6irgxcAtwJuG2O3S7zEoNM5eB7wA+ANgd+B24MMASR4BXAD8Dd2r6/2ApQPPPRb4S7o/otcA756w7ucBTwH2BV4E/Of+8Vf0l0OARwJzgL+dpL8zgWV9bwuB9yQ5tF/2DmB+v47nAC9t/JyrgTcCc4GnA4cCJzTqSXI5sBL4MvCJqrp5stqqWg2cC/zH1jqlyRgUGmf/BfjvVbWsqu4GTgYW9odZjgP+qarO7F8531pVg0Hxxar6flWtAs6gC5JBi6rqjqr6KfCtgeXHAR+squuqagVwEnDMxAnsJA8HngW8tX+VvxT4BPDHfcmLgPdU1e1VtQz468l+yKpaUlWXVNWqqroB+Du6cJxUVe0DbA+8BPh/rdrez1jPqEMaxn1+94Y0jeYB5yS5d+Cx1cDDgIcD1zae+4uB23fRjQyGWb478JOBZT+h+z152ITn7w7cVlXLJ9QeOLD8xoFlg7fXkeSxwAf7527bb2/JZPVrVNVK4Mz+0NbSqvpBo3wP4LYNrVNaH0cUGmc3AodV1UMGLrOr6qZ+2aM28PxN8TO6gFrjEcAq4JfrqdspyXYTam/qb/8c2HNg2cMb2zwVuBp4TFVtD/w3uvmHYT2I7hDXeiXZAjgS+M5GrFP6dwaFxtlHgXcnmQeQZOckz++XnQE8O8mLksxK8tAkEw8vbYozgTcmWZBkDvAeuncRrRosqqobge8C/yvJ7CT7AK/s+wL4e+CkJDsm2QN4bWOb2wG/BlYkeRxw/GSFSZ7WT+Jv1U+Yv5VutPO99dQ+KMnj+59pV7pRi7TRDAqNs/9NN1n7tSTLgUuA/wDQzy0cTvdOntvoJrL3nYJtfgr4LPBt4Hq6CeM/n6T2WLoJ658B5wDvqKqv98v+J91E9/XAPwH/F7h7kvW8mW6uYTnwceCsRn9b003o30o3ejkcOKKqfjZQ8+IkK4A76PbfrcABE2qkocUvLpKmX5LjgWOqqjlJLY0jRxTSNEiyW5Jn9p/L2Itu5HPOqPuSNoXvepKmx1Z0b3NdQHcI6PPAR0bakbSJPPQkSWry0JMkqWmsDz3NnTu35s+fP+o2JGmzsmTJkl9V1c5Ttb6xDor58+ezePHiUbchSZuVJD/ZcNXwPPQkSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVKTQSFJajIoJElNBoUkqcmgkCQ1GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKCRJTbNG3UDLFTfdyfwTz5vSdd6w6IgpXZ8k3d85opAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVKTQSFJajIoJElNBoUkqcmgkCQ1GRSSpCaDQpLUZFD0LroI9tkHtt4a9t8fLrts/XVf+hI8+tEwezYcfDBcf/3aZcm6lxe8YEZal6RpZVAAK1fC0UfD8uVwyinwy1/CwoWwevW6db/4BRxzDGy/Pbz//bBkCbz85evWHH00nHlmd3nzm2fuZ5Ck6TJr1A2Mgwsu6MLhfe+DE07oAuGd74QLL4RDD11bd+aZcPfdcNJJ8Ed/BJdeCp/9LFx7LTzqUV3NE54ARx4JD37wSH4USZpyjihYe/hojz266z337K6vu27j6971LpgzB+bNg698ZXr6laSZtFEjiiQPBb7R390VWA3c0t9/alX9bgp7G5mq7jrZuLq3vhWe9jS45RZ405vg2GO7kcq2205fr5I03TYqKKrqVmA/gCQnAyuq6gNrlieZVVWrprTDGbBgQXe9bFl3fdNNax9fuRK22AK22qpdB7Bo0dp1fvWr8MUvwo03wl57TW//kjSd7vMcRZLTgNuAJwOXJVnOQIAkuRJ4XlXdkOSlwOuArYDvASdU1er1r3nmHHYY7LILnHoqbLcdfPKTMH9+966mWbNg773hyiu7iewTT4T3vrcbKZxzDjzrWd38xPnnw+c+1z3n9tu7eY+dd14bIpK0uZqqOYrHAs+uqjdNVpDk8cCLgWdW1X50h62OW0/dq5IsTrJ49V13TlF7bbNnw9lnd3MLr399Fxpnnw1bbrlu3W67dRPad9zRvaPpyU+G007rls2bBz//ObzlLd08xYEHwnnndSMRSdqcTdW7ns4eYmRwKHAAcGm6g/rbADdPLKqqjwEfA9h6t8fUFPW3QQcdBFdc8fuP14QOXvjC7jLR3nvDt741Pb1J0ihNVVD8ZuD2KtYdqczurwOcXlUnTdE2JUkzYDreHnsDsD9Akv2BNUfpvwEsTLJLv2ynJPOmYfuSpCk0HUHxBWCnJEuB44EfA1TVD4G3AV9LcjnwdWC3adi+JGkKbfKhp6o6eZLHfwv84STLzgLO2tRtSpJmnp/MliQ1GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVLTrFE30PKkPXZg8aIjRt2GJD2gOaKQJDUZFJKkJoNCktRkUEiSmgwKSVKTQSFJajIoJElNBoUkqcmgkCQ1GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVKTQSFJajIoJElNBoUkqcmgkCQ1GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVLTrFE30HLFTXcy/8TzRt2GJM2oGxYdMeoW1uGIQpLUZFBIkpoMCklSk0EhSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVKTQSFJm4mLLoJ99oGtt4b994fLLvv9mpNPBqgDEmrwsmZ5wsMTzk34TcKdCWdsaLsGhSRtBlauhKOPhuXL4ZRT4Je/hIULYfXqdesWLgR41XXAscBr+4f/BSAhwDnAc4D3A28BbtnQtg0KSdoMXHBBFw4nnNBdXvlKuP56uPDCdeue+ESAj99exeeBbfqHP9pfHwIcAHwQWAR8rIo3bGjbBoUkbQauv7673mOP7nrPPbvr665bf30/engV8Gvg//QPP6G/Phq4C/h1wus2tO0NBkWS1UmWDlzmN2pXbGh9kqT7rvpZh2TSkkOAxwCfq2LN3+at++t7gKOA64EPJTy2ta1ZQ/Tz26rab4g6SdI0WbCgu162rLu+6aa1j69cCVtsAVtttc5TXt1ff3TgsRv66/OqODfhacCTgAXAjyfb9jBBsY4kc4BzgR2BBwFvq6pzJ9TsBpwFbN9v4/iq+k6SPwT+ki7VrgX+pKochUjSBhx2GOyyC5x6Kmy3HXzykzB/Phx8MMyaBXvvDVdeuaZ691nAC4CLqrhiYDXnAzcDRydcAywEVtBPdk9mmDmKbQYOO50DrASOqqr96YY2f5X83uDnJcA/9iORfYGlSeYCbwOe3T93MfBfJ24syauSLE6yePVddw7RniTd/82eDWefDXPmwOtf34XG2WfDlluur/o1c+leyA+OJqjit3ThcDfwYbp5ihdWcXNr26mq1nKSrKiqOQP3HwScAhwE3AvsBSyoql+sqU1yEPAp4HPAl6pqaZLnAacB/cCJrYCLq+qVk217690eU7u9/EPN/iTp/uaGRUfcp+cnWVJVB05ROxt/6Ak4DtgZOKCq7klyAzB7sKCqvt2HxRHAZ5O8H7gd+HpVHXsfe5YkzaBNeXvsDsDNfUgcAsybWJBkXl/zceCTwP7AJcAzkzy6r9k2SXOmXZI0epsyojgD+Icki4GlwNXrqTkY+Isk99BNlLysqm5J8grgzCRr3qL1Nhoz7ZKk0dtgUAzOT/T3fwU8vVVbVacDp69n+TeBp2xSp5KkkfCT2ZKkJoNCktRkUEiSmgwKSVKTQSFJajIoJElNBoUkqcmgkCQ1GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaDApJUpNBIUlqMigkSU2zRt1Ay5P22IHFi44YdRuS9IDmiEKS1GRQSJKaDApJUpNBIUlqMigkSU0GhSSpyaCQJDUZFJKkJoNCktRkUEiSmgwKSVKTQSFJajIoJElNBoUkqcmgkCQ1GRSSpCaDQpLUZFBIkpoMCklSk0EhSWoyKCRJTQaFJKnJoJAkNRkUkqQmg0KS1GRQSJKaUlWj7mFSSZYDPxp1H0OYC/xq1E0MwT6nln1OLfucOvOqauepWtmsqVrRNPlRVR046iY2JMli+5w69jm17HNqbS59TiUPPUmSmgwKSVLTuAfFx0bdwJDsc2rZ59Syz6m1ufQ5ZcZ6MluSNHrjPqKQJI2YQSFJahqLoEjy3CQ/SnJNkhPXszxJ/rpffnmS/ce0z8cluTjJ3UnePIoe+z421Odx/X68PMl3k+w7pn0+v+9xaZLFSZ41jn0O1D0lyeokC2eyv4Htb2h/Hpzkzn5/Lk3y9nHss685uO/xqiT/PNM99j1saH/+xcC+vLL/t99pFL1Ou6oa6QXYErgWeCSwFfAD4AkTag4HLgACPA343pj2uQvwFODdwJvHeH8+A9ixv33YGO/POaydR9sHuHoc+xyo+yZwPrBwHPsEDga+Mor/lxvZ50OAHwKP6O/vMo59Tqg/EvjmKPftdF7GYUTxVOCaqrquqn4HfB54/oSa5wOfqc4lwEOS7DZufVbVzVV1KXDPDPc2aJg+v1tVt/d3LwH2nOEeYbg+V1T/Wwg8GBjFOy+G+f8J8OfAF4CbZ7K5AcP2OWrD9PkS4ItV9VPofq9muEfY+P15LHDmjHQ2AuMQFHsANw7cX9Y/trE1020cehjGxvb5SrrR2kwbqs8kRyW5GjgP+NMZ6m3QBvtMsgdwFPDRGexromH/3Z+e5AdJLkiy98y0to5h+nwssGOSC5MsSfKyGeturaF/j5JsCzyX7oXC/dI4nMIj63ls4ivHYWqm2zj0MIyh+0xyCF1QjOLY/1B9VtU5wDlJDgLeCTx7uhubYJg+PwS8tapWJ+srnxHD9HkZ3TmAViQ5HPgS8Jhp72xdw/Q5CzgAOBTYBrg4ySVV9ePpbm7Axvy+HwlcVFW3TWM/IzUOQbEMePjA/T2Bn21CzXQbhx6GMVSfSfYBPgEcVlW3zlBvgzZqf1bVt5M8KsncqprJE7IN0+eBwOf7kJgLHJ5kVVV9aWZaBIbos6p+PXD7/CQfGdP9uQz4VVX9BvhNkm8D+wIzGRQb8//zGO7Hh52AsZjMngVcByxg7aTR3hNqjmDdyezvj2OfA7UnM7rJ7GH25yOAa4BnjPm/+6NZO5m9P3DTmvvj1OeE+tMYzWT2MPtz14H9+VTgp+O4P4HHA9/oa7cFrgSeOG599nU7ALcBD57pf/OZvIx8RFFVq5K8FvhHuncafKqqrkry6n75R+neSXI43R+3u4A/Gcc+k+wKLAa2B+5N8ga6d0r8etIVj6BP4O3AQ4GP9K+CV9UMnw1zyD6PBl6W5B7gt8CLq//tHLM+R27IPhcCxydZRbc/jxnH/VlV/5rkq8DlwL3AJ6rqynHrsy89CvhadaOf+y1P4SFJahqHdz1JksaYQSFJajIoJElNBoUkqcmgkCQ1GRTSDOrPgnzshMe2TPKaJLNH1ZfUYlDoASHJrkk+n+TaJD9Mcn6Sx07Bel+RZPcha/8AuLuqJn6K96+Aq6pq5X3tR5oOfo5C93vpPlX4XeD0NR+USrIfsF1Vfec+rvtCuk/hL77PjUpjyhGFHggOAe4Z/BR1VS0FnjPwxTM3Jfk0QJKXJvl+//jf9YeGtkxyWv8FNVckeWP/BUUHAmf0tdskOTTJv/Q1n0qydb/ORf1I5vIkH+gfe1iSc/qzuf4gyTNmftdIG+aIQvd7SV4HLKiqN06yfAfgO3SnhrkLeB/wwqq6J8lH6L6z4ypgUVU9p3/OQ6rqjsERRT/H8G/AoVX14ySfoTtj62eAi4HHVVUNPPcs4OKq+lCSLYE5VXXnNO4KaZM4otADWn9Y6gzglKpaQndq6wOAS5Ms7e8/ku4EcY9M8jdJngus7/xdewHX19rTYZ8OHNTXrgQ+keSFdGEE8J+AUwGqarUhoXFlUOiB4Cq6P/7rczKwrKo+3d8P3VzGfv1lr6o6ubpvBNwXuBB4Dd0p2ida75dRVNUqurO1fgF4AfDVTf1BpFEwKPRA8E1g6yR/tuaBJE9J8g7gOcDrBmq/ASxMsktft1OSeUnmAltU1ReA/0F32nOA5cB2/e2rgflJHt3f/2Pgn5PMAXaoqvOBNwD7DWzr+H47WybZfkp/ammKOEehB4T+LawfohtZrARuoPuugwXAHX3Zl6vq7UleDJwEzKb7LoKX0J2W+9OsfXF1UlVdkORo4D398qcDzwA+QPd9BpfSBcFOwLn9+gJ8oKpOT/Iw4GN0h7ZWA8dX1cXTtQ+kTWVQSJNI8ha68Lh61L1Io+ShJ2lyvwM+MBUfzJM2Z44oJElNjigkSU0GhSSpyaCQJDUZFJKkJoNCktT0/wGfEuteC2KO8QAAAABJRU5ErkJggg==\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#stystki opisowe\n",
    "for column in df:\n",
    "    try:\n",
    "        ax = df[column].plot(kind = \"box\" ,vert = False , sym = \"b+\" )\n",
    "        #print(df[column].describe())\n",
    "        plt.savefig(f\"figs/{column}.jpg\")\n",
    "        plt.show()\n",
    "        text = df[column].describe().to_latex()\n",
    "        save_to_latex(text, f\"{column}_stat\")\n",
    "        print(f\"SD = {df[column].std}\")\n",
    "      \n",
    "    except:\n",
    "        TypeError\n",
    "        ax = df[column].value_counts().div(100).plot(kind = \"barh\",title= column )\n",
    "        #print(df[column].value_counts().div(100))\n",
    "        ax.set_xlabel(\"Częstość\")\n",
    "        for i, v in enumerate(df[column].value_counts().div(100)):\n",
    "             ax.text(v + 0.01, i , str(v), color='blue', fontweight='bold')\n",
    "        plt.savefig(f\"figs/{column}.jpg\")\n",
    "        plt.show()"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "outputs": [],
   "source": [
    "#przygotwanie danych do regresji \n",
    "#usuniecie obserwacji odstajacych \n",
    "count = df[\"Cena\"]\n",
    "kwartyl = count.quantile(0.85)\n",
    "drogie  = count > kwartyl\n",
    "df = df.drop(drogie.loc[drogie].index)\n",
    "#df = df.drop(\"Liczba złączy HDMI\" , axis= 1)\n",
    "\n",
    "regresion_data = pd.get_dummies(df, drop_first=True)\n",
    "#usunac brakujace dane \n",
    "regresion_data = regresion_data.drop(\"jakosc_8K UHD\", axis=1)\n",
    "#to save unchanged dataframe\n",
    "df_initial= df"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3>Macierz korelacji "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 864x864 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "corr_matix  = associations(df,nan_strategy= \"drop_features\",plot = False, figsize=(12,12))\n",
    "ax = corr_matix[\"ax\"]\n",
    "\n",
    "plt.savefig(\"figs/macierzkorelacji.jpg\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3>Model liniowy"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "outputs": [],
   "source": [
    "Y = regresion_data[\"Cena\"]\n",
    "X = regresion_data.iloc[:,1:]\n",
    "\n",
    "X = sm.add_constant(X).astype(float)\n",
    "\n",
    "model = sm.OLS(Y,X)\n",
    "result = model.fit()\n",
    "\n",
    "result.summary()\n",
    "save_to_latex(result.summary(),\"first_model\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3> Wstepna analiza modelu "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 720x720 with 0 Axes>"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "plt.figure( figsize = (10,10))\n",
    "plot = sm.graphics.plot_fit(result,1)\n",
    "plt.savefig(\"figs/wtepna_analiza.jpg\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3> Poprawki do modelu</h3>\n",
    "\n",
    "*  Zbadnie wspoliniowosci \n",
    "*  Usuniecie zmiennych nie znaczacych stystycznie metoda wsteczna krokową "
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "outputs": [],
   "source": [
    "def variance_inflation_factor(exog, exog_idx):\n",
    "    \"\"\"variance inflation factor, VIF, for one exogenous variable\n",
    "    https://en.wikipedia.org/wiki/Variance_inflation_factor\n",
    "    \"\"\"\n",
    "    k_vars = exog.shape[1]\n",
    "    x_i = exog.iloc[:, exog_idx]\n",
    "    mask = np.arange(k_vars) != exog_idx\n",
    "    x_noti = exog.iloc[:, mask]\n",
    "    r_squared_i = sm.OLS(x_i, x_noti).fit().rsquared\n",
    "    vif = 1. / (1. - r_squared_i)\n",
    "    return vif"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "outputs": [],
   "source": [
    "#wspoliniowsc \n",
    "vif_list = []\n",
    "for i in range(X.shape[1]):\n",
    "    vif = variance_inflation_factor(X,i)\n",
    "    vif_list.append((X.iloc[:,i].name,vif))\n",
    "vif_list = pd.DataFrame(vif_list)\n",
    "save_to_latex(vif_list.to_latex(),\"VIF\")\n",
    "#nie ma wyskoiej wpołsliniowsci w zmiennych\n",
    "#wartosc 100 dal const wynika z powodu skalowania.\n",
    "#https://stats.stackexchange.com/questions/332428/regression-model-constant-causes-multicollinearity-warning-but-not-in-standardi/332597#332597"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                     0                     1\n",
      "0                 Stan   0.13544747232794202\n",
      "1       Typ telewizora  0.008491797331009644\n",
      "2                Marka   0.40890046107407024\n",
      "3   Klasa efektywności    0.5532931630889933\n",
      "4               jakosc    0.8667612754589771\n",
      "5      Technologia HDR    0.8050348162403695\n",
      "6       Technologia 3D    0.6797816760227706\n",
      "7           Pobór mocy              0.753333\n",
      "8     Przekątna ekranu            9.6261e-07\n",
      "9   Liczba złączy HDMI              0.974711\n",
      "10               const              0.121632\n"
     ]
    }
   ],
   "source": [
    "stan_f = \"(Stan_Używany = Stan_Powystawowy = Stan_Po zwrocie )\"\n",
    "typ_f = \"(Typ telewizora_LED = Typ telewizora_OLED = Typ telewizora_QLED = Typ telewizora_inny = Typ telewizora_plazmowy = 0)\"\n",
    "marka_f = \"(Marka_LG=Marka_Philips=Marka_Samsung= Marka_Sony=Marka_inna =0)\"\n",
    "klasa_f = \"(Klasa efektywności_A+ =Klasa efektywności_B = Klasa efektywności_C = Klasa efektywności_D=Klasa efektywności_inna=0)\"\n",
    "jakosc_f = \"(jakosc_Full HD= jakosc_HD Ready = 0)\"\n",
    "hdr_f = \"(Technologia HDR_Tak = 0)\"\n",
    "tech3D_f = \"(Technologia 3D_True=0)\"\n",
    "\n",
    "testy_f = [stan_f,typ_f,marka_f,klasa_f,jakosc_f,hdr_f,tech3D_f]\n",
    "zmienne_ciagle = [\"Pobór mocy\",\"Przekątna ekranu\",\"Liczba złączy HDMI\",\"const\"]\n",
    "\n",
    "def run_test(zmienne_kategoryczne,zmienne_ciagle , model):\n",
    "    wynik_test = []\n",
    "    #F test dla dummy zmiennej kategorycznej \n",
    "    for test in zmienne_kategoryczne:\n",
    "        try:\n",
    "            wynik = model.f_test(test)\n",
    "            name = test[1:].split(sep = \"_\")[0]\n",
    "            s = f\"{name} nie jest znacząca\"\n",
    "            s1 = f\"{name} jest znaczące\"\n",
    "            wynik_test.append((name, wynik.pvalue))\n",
    "            #print(f\"{s1 if wynik.pvalue < 0.05 else s}\")\n",
    "        except:pass\n",
    "    #t test dla zmiennych ciaglych \n",
    "    for ttest in zmienne_ciagle:\n",
    "        try:\n",
    "            wynik_test.append((ttest,model.pvalues[ttest]))\n",
    "        except:pass\n",
    "    return wynik_test\n",
    "        \n",
    "wynik_test = run_test(testy_f,zmienne_ciagle,result)\n",
    "wynik_pd = pd.DataFrame(wynik_test)\n",
    "print(wynik_pd)\n",
    "save_to_latex(wynik_pd.to_latex(),\"Testy_istotnosci\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "outputs": [],
   "source": [
    "#metoda przyjmuje df oraz liste tupli (nazwa,p-value)\n",
    "def metoda_wsteczna_krokowa(df, pvalue_list):\n",
    "    najwieksza = max(pvalue_list, key=itemgetter(1))\n",
    "    #usuwanie \n",
    "    df = df.drop(najwieksza[0], axis= 1)\n",
    "    return df "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "outputs": [],
   "source": [
    "#safeguard from globals \n",
    "df2 = df"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "usuwana zmienna to Liczba złączy HDMI, z pvalue rowne 0.974711137133669\n",
      "usuwana zmienna to jakosc, z pvalue rowne 0.8627778996805936\n",
      "usuwana zmienna to Technologia 3D, z pvalue rowne 0.792828352613247\n",
      "usuwana zmienna to Pobór mocy, z pvalue rowne 0.7259775248299482\n",
      "usuwana zmienna to Technologia HDR, z pvalue rowne 0.5133421065955824\n",
      "usuwana zmienna to Marka, z pvalue rowne 0.20835637190097164\n",
      "usuwana zmienna to Klasa efektywności, z pvalue rowne 0.16149461615509111\n"
     ]
    }
   ],
   "source": [
    "#while na metode wsteczna krokowa \n",
    "while True:\n",
    "    najwieksza =  max(wynik_test, key=itemgetter(1))\n",
    "    #gdy wszystkie zmeinne sa stsystcznie znaczace koniec metody krokowej\n",
    "    if najwieksza[1] < 0.1: break\n",
    "    print(f\"usuwana zmienna to {najwieksza[0]}, z pvalue rowne {najwieksza[1]}\")\n",
    "    df2 = metoda_wsteczna_krokowa(df2,wynik_test)\n",
    "    regresion_data2 = pd.get_dummies(df2, drop_first=True)\n",
    "    Y = regresion_data2[\"Cena\"]\n",
    "    X = regresion_data2.iloc[:,1:]\n",
    "    X = sm.add_constant(X).astype(float)\n",
    "    result2 = sm.OLS(Y,X).fit()\n",
    "    wynik_test = run_test(testy_f,zmienne_ciagle,result2)   \n",
    "result2.summary()\n",
    "save_to_latex(result2.summary(),\"model_poprawiony\")\n",
    "#wszystkie zmienne sa znaczace "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot = sm.graphics.plot_fit(result,1)\n",
    "plt.savefig(\"figs/poprawion_model.jpg\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "efekt kataliz w parach (i,j) []\n"
     ]
    }
   ],
   "source": [
    "def sign(x):\n",
    "    return math.copysign(1, x)\n",
    "#macierz korelacji zmeinnych egzodoniczych\n",
    "R_corr = associations(df2.iloc[:,1:],nan_strategy= \"drop_features\")\n",
    "R_corr = np.array(R_corr[\"corr\"])\n",
    "\n",
    "R = associations(df2,nan_strategy= \"drop_features\")\n",
    "R = np.array(R[\"corr\"].iloc[:,0])[1:]\n",
    "\n",
    "R_sorted = R.argsort()\n",
    "r_0 = R[R_sorted]\n",
    "r = R_corr[R_sorted]\n",
    "\n",
    "pairs = []\n",
    "for i in range(r.shape[0]):\n",
    "    for j in range(r.shape[0]):\n",
    "        if abs(r_0[i]) < abs(r_0[j]):\n",
    "            r_ij = r[i, j] * sign(r_0[i]) * sign(r_0[j])\n",
    "            if r_ij > (abs(r_0[i]) / abs(r_0[j])) or r_ij < 0:\n",
    "                pairs.append((i, j))\n",
    "\n",
    "if pairs == None: print(\"Brak efektu katalizy \")\n",
    "else: print(f\"efekt kataliz w parach (i,j) {pairs}\")\n",
    "#brak katalizatorów "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "rozklad reszt jest normalny\n"
     ]
    }
   ],
   "source": [
    "#Normalność rozkładu składnika losowego.\n",
    "Jarque_Bera_pv = float(result2.summary().tables[2].data[2][3])\n",
    "if Jarque_Bera_pv >0.1: print(\"rozklad reszt nie jest normalny \")\n",
    "else:print(\"rozklad reszt jest normalny\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "outputs": [],
   "source": [
    "def sort_by(data, sorter):\n",
    "      sotred_sorter ,sorted_arr = zip( *sorted(zip(sorter,data)))\n",
    "      return np.array(sorted_arr)\n",
    "def p_value_test_seri(reszty, Y):\n",
    "    reszty = np.array(reszty.resid)\n",
    "    sortedd = sort_by(reszty,Y)\n",
    "    normalize = np.vectorize(lambda a: 1 if a>0 else 0)\n",
    "    sortedd = normalize(sortedd)\n",
    "    series_pvalue = sm.stats.Runs(sortedd).runs_test()\n",
    "    return series_pvalue[1]\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8216985501200023 P-value testu seri\n"
     ]
    }
   ],
   "source": [
    "#test seri (Wald)\n",
    "\n",
    "print(f\"{p_value_test_seri(result2,Y)} P-value testu seri\")\n",
    "#model nie jest liniowy "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "outputs": [
    {
     "data": {
      "text/plain": "<class 'statsmodels.stats.contrast.ContrastResults'>\n<Wald test (chi2): statistic=[[27.71026936]], p-value=9.611506453997187e-07, df_denom=2>"
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_value_RESET = sm.stats.linear_reset(result2)\n",
    "p_value_RESET\n",
    "#Brak podstaw do odrzucenia h_0 pierwotna postac modelu jest poprawna \n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "outputs": [],
   "source": [
    "#transformcaja ceny do log \n",
    "Y = np.log(regresion_data2[\"Cena\"])\n",
    "X = regresion_data2.iloc[:,1:]\n",
    "X = sm.add_constant(X).astype(float)\n",
    "result2_log = sm.OLS(Y,X).fit()\n",
    "result2_log.summary()\n",
    "save_to_latex(result2_log.summary(), \"model_logarytm\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0010945248619090076 test seri dla modelu z log\n"
     ]
    }
   ],
   "source": [
    "print(f\"{p_value_test_seri(result2_log ,Y)} test seri dla modelu z log\")\n",
    "#transformacja modelu jest poprawna "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "outputs": [
    {
     "data": {
      "text/plain": "<class 'statsmodels.stats.contrast.ContrastResults'>\n<Wald test (chi2): statistic=[[0.1804874]], p-value=0.6709547118839503, df_denom=1>"
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#po korecji na log model liniowy nie jest juz najlepszy \n",
    "p_value_RESET = sm.stats.linear_reset(result2_log,2)\n",
    "p_value_RESET"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                   Cena   R-squared:                       0.805\n",
      "Model:                            OLS   Adj. R-squared:                  0.775\n",
      "Method:                 Least Squares   F-statistic:                     27.08\n",
      "Date:                Sun, 05 Jul 2020   Prob (F-statistic):           8.33e-18\n",
      "Time:                        13:18:43   Log-Likelihood:                -25.902\n",
      "No. Observations:                  69   AIC:                             71.80\n",
      "Df Residuals:                      59   BIC:                             94.14\n",
      "Df Model:                           9                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===========================================================================================\n",
      "                              coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-------------------------------------------------------------------------------------------\n",
      "const                       5.3786      0.209     25.754      0.000       4.961       5.797\n",
      "Przekątna ekranu            0.0004    4.7e-05      9.302      0.000       0.000       0.001\n",
      "Stan_Po zwrocie             0.1391      0.282      0.492      0.624      -0.426       0.704\n",
      "Stan_Powystawowy            0.0632      0.290      0.218      0.828      -0.517       0.644\n",
      "Stan_Używany               -0.0465      0.108     -0.430      0.669      -0.263       0.170\n",
      "Typ telewizora_LED          0.6247      0.192      3.259      0.002       0.241       1.008\n",
      "Typ telewizora_OLED         0.6817      0.512      1.331      0.188      -0.343       1.707\n",
      "Typ telewizora_QLED         1.1406      0.256      4.449      0.000       0.628       1.654\n",
      "Typ telewizora_inny        -0.0224      0.418     -0.054      0.957      -0.860       0.815\n",
      "Typ telewizora_plazmowy     0.3887      0.266      1.464      0.149      -0.143       0.920\n",
      "==============================================================================\n",
      "Omnibus:                       28.917   Durbin-Watson:                   1.829\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):               70.967\n",
      "Skew:                          -1.308   Prob(JB):                     3.89e-16\n",
      "Kurtosis:                       7.224   Cond. No.                     3.17e+04\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 3.17e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "0.0010945248619090076 test seri dla modelu z log\n",
      "<Wald test (chi2): statistic=[[5.6455531]], p-value=0.05944067349951398, df_denom=2>\n",
      " Brauch(0.22511425823068912, 0.23095277139604928)\n",
      " White (0.1240566270299608, 0.10427497541450217)\n"
     ]
    }
   ],
   "source": [
    "#model z przekatna do kwadratu \n",
    "Y = np.log(regresion_data2[\"Cena\"])\n",
    "X = regresion_data2.iloc[:,1:]\n",
    "X[\"Przekątna ekranu\"] = X[\"Przekątna ekranu\"]**2\n",
    "X = sm.add_constant(X).astype(float)\n",
    "result2_log_sq = sm.OLS(Y,X).fit()\n",
    "print(result2_log_sq.summary())\n",
    "\n",
    "save_to_latex(result2_log_sq.summary(), \"model_log_kwadrat\")\n",
    "print(f\"{p_value_test_seri(result2_log ,Y)} test seri dla modelu z log\")\n",
    "#transformacja modelu jest poprawna \n",
    "\n",
    "#tearz postac modelu jest juz poprawna \n",
    "p_value_RESET = sm.stats.linear_reset(result2_log_sq,3)\n",
    "print(p_value_RESET)\n",
    "\n",
    "#breuschpagan test \n",
    "breush = sm.stats.diagnostic.het_breuschpagan(result2_log_sq.resid, X)\n",
    "#p-value of lagrange multiplier test\n",
    "lm_pvalue = breush[1]\n",
    "#p-value for the f-statistic\n",
    "f_pvalue = breush[3]\n",
    "print(f\" Brauch{lm_pvalue,f_pvalue}\")\n",
    "#h_0 =homoskedasticity , odrzucam wiec jest heterodastycznosc\n",
    "#white test \n",
    "white = sm.stats.diagnostic.het_white(result2_log_sq.resid,X)\n",
    "lm_pvalue = white[1]\n",
    "#p-value for the f-statistic\n",
    "f_pvalue = white[3]\n",
    "print(f\" White {lm_pvalue,f_pvalue}\")\n",
    "#h_0 =homoskedasticity , odrzucam wiec jest heterodastycznosc\n",
    "#wystepuje niska heterodastycznosc "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "result2_log_sq.resid.plot()\n",
    "\n",
    "plt.savefig(\"figs/reszty_log_sq.jpg\")\n",
    "#wynika to z pojedynczych obserwacji \n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "outputs": [],
   "source": [
    "#poprawka na heterodastycznosc\n",
    "X2 = X\n",
    "X2[\"Przekątna ekranu\"] = X[\"Przekątna ekranu\"] / np.sqrt(np.exp(result2_log_sq.resid)**2)\n",
    "Y2 = np.log(np.exp(Y)/ np.sqrt(np.exp(result2_log_sq.resid)**2))\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "outputs": [],
   "source": [
    "result2_log_sq_het = sm.OLS(Y2,X2).fit()\n",
    "result2_log_sq_het.summary()\n",
    "\n",
    "save_to_latex(result2_log_sq_het.summary(), \"model_log_sq_homo\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Brauch (3.140388461885039e-07, 4.953662458242724e-12)\n",
      " White (2.2945588282568748e-05, 4.777562405746083e-10)\n"
     ]
    }
   ],
   "source": [
    "breush = sm.stats.diagnostic.het_breuschpagan(result2_log_sq_het.resid, X)\n",
    "#p-value of lagrange multiplier test\n",
    "lm_pvalue = breush[1]\n",
    "#p-value for the f-statistic\n",
    "f_pvalue = breush[3]\n",
    "print(f\" Brauch {lm_pvalue,f_pvalue}\")\n",
    "#h_0 =homoskedasticity , odrzucam wiec jest heterodastycznosc\n",
    "\n",
    "\n",
    "#white test \n",
    "white = sm.stats.diagnostic.het_white(result2_log_sq_het.resid,X)\n",
    "lm_pvalue = white[1]\n",
    "#p-value for the f-statistic\n",
    "f_pvalue = white[3]\n",
    "print(f\" White {lm_pvalue,f_pvalue}\")\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "outputs": [
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "result2_log_sq_het.resid.plot()\n",
    "\n",
    "\n",
    "plt.savefig(\"figs/reszty_log_sq_hetero.jpg\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.6386188579970305e-05 test seri\n"
     ]
    }
   ],
   "source": [
    "#model jest liniowy\n",
    "print(f\"{p_value_test_seri(result2_log_sq_het,Y)} test seri\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "outputs": [
    {
     "data": {
      "text/plain": "<class 'statsmodels.stats.contrast.ContrastResults'>\n<Wald test (chi2): statistic=[[8.8332306]], p-value=0.01207503353496347, df_denom=2>"
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#postac modelu jest prawidlowa\n",
    "p_value_RESET = sm.stats.linear_reset(result2_log_sq_het,3)\n",
    "p_value_RESET\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "outputs": [],
   "source": [
    "#test Chowa\n",
    "from sklearn.model_selection import train_test_split\n",
    "def chow_test(X , Y):\n",
    "    n= X.shape[0]\n",
    "    k = X.shape[1]\n",
    "    x_1, x_2 , y_1, y_2 =  train_test_split(X,Y,test_size=0.5)\n",
    "    \n",
    "    ols_1 = sm.OLS( y_1,x_1)\n",
    "    ols_2 = sm.OLS( y_2,x_2)\n",
    "    mod_1 = ols_1.fit()\n",
    "    mod_2 = ols_2.fit()\n",
    "    rsk_1 = mod_1.ssr\n",
    "    rsk_2 = mod_2.ssr\n",
    "    rsk = sm.OLS(Y, X).fit()\n",
    "    rsk = rsk.ssr\n",
    "    r_1 = k - 1\n",
    "    r_2 = n - 2 * (k -1)\n",
    "    f_stat = ((rsk - rsk_1 - rsk_2) / (rsk_1 + rsk_2)) * (r_2 / r_1)\n",
    "    #return p-value of F statistic\n",
    "    return stats.f.sf(f_stat, dfn=r_1, dfd=r_2)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "outputs": [
    {
     "data": {
      "text/plain": "[0.3212254434435945,\n 0.001343873020050195,\n 0.028924044911825673,\n 1.58781733609053e-05,\n 0.02059132002564599,\n 0.0007728029864326307,\n 0.14595913608706587,\n 0.06849378026357827,\n 0.0010656147940581952,\n 0.3470785752328491,\n 0.6325369622698582,\n 0.48514519962295255,\n 0.016373902850327953,\n 0.017171033629302165,\n 0.20000052956752384,\n 0.21699479967269822,\n 0.19254559106980398,\n 0.0024861031461344135,\n 0.06211706899104807,\n 0.09452892067125386]"
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "chow_list = []\n",
    "for i in range(20):\n",
    "    chow_list.append(chow_test(X2,Y2))\n",
    "chow_list\n",
    "#test chowa pokazuje ze model nie jest stabilny, ponieważ jest mało danych i duża wariancja w dancyh."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "outputs": [
    {
     "data": {
      "text/plain": "[('const', 19.751096599853554),\n ('Przekątna ekranu', 1.2967734278112613),\n ('Stan_Po zwrocie', 1.0544684157460662),\n ('Stan_Powystawowy', 1.5948010145355067),\n ('Stan_Używany', 1.3879348950594765),\n ('Typ telewizora_LED', 3.511562041170873),\n ('Typ telewizora_OLED', 1.759289202476175),\n ('Typ telewizora_QLED', 3.1901534859977843),\n ('Typ telewizora_inny', 1.1848862798512998),\n ('Typ telewizora_plazmowy', 1.89202020130632)]"
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#wspołliniowosc \n",
    "vif_list = []\n",
    "for i in range(X2.shape[1]):\n",
    "    vif = variance_inflation_factor(X2,i)\n",
    "    vif_list.append((X2.iloc[:,i].name,vif))\n",
    "vif_list \n",
    "#poza const nie ma wysokiej wpołliniowosci "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "outputs": [],
   "source": [
    "def collinearity_test(X):\n",
    "        collinear = []\n",
    "        for i in range(X.shape[1]):\n",
    "            x = X.drop(X.columns[i], axis=1)\n",
    "            model = sm.OLS(X.iloc[:,i],x)\n",
    "            res = model.fit()\n",
    "            if res.rsquared  > 0.9:\n",
    "                collinear.append(X.columns[i])\n",
    "        return collinear"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "outputs": [
    {
     "data": {
      "text/plain": "['const']"
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coliniar = collinearity_test(X2)\n",
    "coliniar\n",
    "#wysoka wspolniwososc const wynika ze złego scalowania"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "outputs": [],
   "source": [
    "#koincydencja \n",
    "def model_coincidence(X_w,Y):\n",
    "    X = np.array(X_w)\n",
    "    Y = np.array(Y)\n",
    "    r_0 = [np.corrcoef(X[:, i], Y)[0][1] for i in range(X.shape[1])]\n",
    "    x_t_x_inverse = np.linalg.inv(X.T @ X)\n",
    "    a = x_t_x_inverse @(X.T@Y)\n",
    "    results = []\n",
    "    for r in range(X.shape[1]):\n",
    "        if sign(r_0[r]) != sign(a[r]):\n",
    "            results.append(\n",
    "                dict(variable=X_w.columns[r], a=a[r], r=r_0[r])\n",
    "            )\n",
    "    return results\n",
    "\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "outputs": [
    {
     "data": {
      "text/plain": "[{'variable': 'const', 'a': 5.699255350628206, 'r': nan},\n {'variable': 'Typ telewizora_LED',\n  'a': 0.8278866585456797,\n  'r': -0.179241207762127},\n {'variable': 'Typ telewizora_plazmowy',\n  'a': 0.5140352297157307,\n  'r': -0.01789614388591152}]"
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "koincydencja = model_coincidence(X2,Y2)\n",
    "koincydencja\n",
    "#koincydecja w dummy varieble nie jest znaczaca, \n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3> Wybór modelu </h3>\n",
    "Wszyskie tessty zakonczone. \n",
    "Otrzymany model ma postać: \n",
    "log(Cena) = a + a_1 * przekątna_ekranu^2 + a_3 *stan + a_4 * typ\n",
    "Teraz zbiórczo testy dla tego modelu \n",
    "oraz przwidywania z 95% pewnoscia \n",
    "poziom istotnosci modelu to apha = 0.1 \n"
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "markdown",
   "source": [
    "<h3>Testowanie Modelu oraz Pełny opis</h3>\n",
    "\n",
    "*  Współczynnik determinacji.\n",
    "*  Efekt katalizy.\n",
    "*  Normalność rozkładu składnika losowego.\n",
    "*  Istotność zmiennych.\n",
    "*  Testy dodanych (pominiętych zmiennych).\n",
    "*  Obserwacje odstające.\n",
    "*  Test liczby serii (uporządkować obserwacje!!).\n",
    "*  Test RESET.19.Testowanie heteroskeastyczności.\n",
    "*  Test Chowa (wraz z komentarzem na temat podziału próbki na dwie grupy).\n",
    "*  Współliniowość.\n",
    "*  Koincydencja.\n",
    "*  Interpretacja parametrów modelu (tam gdzie trzeba elastyczność).\n",
    "*  predykcja wraz z 95% przedziałem ufności."
   ],
   "metadata": {
    "collapsed": false
   }
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.809\n",
      "Model:                            OLS   Adj. R-squared:                  0.780\n",
      "Method:                 Least Squares   F-statistic:                     27.76\n",
      "Date:                Sun, 05 Jul 2020   Prob (F-statistic):           4.69e-18\n",
      "Time:                        13:18:44   Log-Likelihood:                -17.733\n",
      "No. Observations:                  69   AIC:                             55.47\n",
      "Df Residuals:                      59   BIC:                             77.81\n",
      "Df Model:                           9                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===========================================================================================\n",
      "                              coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-------------------------------------------------------------------------------------------\n",
      "const                       5.6993      0.181     31.483      0.000       5.337       6.061\n",
      "Przekątna ekranu            0.0002   2.36e-05      7.116      0.000       0.000       0.000\n",
      "Stan_Po zwrocie             0.3620      0.249      1.452      0.152      -0.137       0.861\n",
      "Stan_Powystawowy            0.4769      0.252      1.891      0.064      -0.028       0.982\n",
      "Stan_Używany               -0.0655      0.096     -0.681      0.498      -0.258       0.127\n",
      "Typ telewizora_LED          0.8279      0.168      4.921      0.000       0.491       1.165\n",
      "Typ telewizora_OLED         1.0879      0.452      2.406      0.019       0.183       1.992\n",
      "Typ telewizora_QLED         1.6965      0.216      7.853      0.000       1.264       2.129\n",
      "Typ telewizora_inny        -0.1934      0.371     -0.521      0.604      -0.936       0.549\n",
      "Typ telewizora_plazmowy     0.5140      0.240      2.144      0.036       0.034       0.994\n",
      "==============================================================================\n",
      "Omnibus:                        6.372   Durbin-Watson:                   1.969\n",
      "Prob(Omnibus):                  0.041   Jarque-Bera (JB):               10.243\n",
      "Skew:                          -0.117   Prob(JB):                      0.00597\n",
      "Kurtosis:                       4.873   Cond. No.                     3.86e+04\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 3.86e+04. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "F-vaule dla modelu wynosi:4.6933845457305416e-18 \n",
      " Brak podstaw do orzucenia H_0\n",
      " R^2 jest statystycznie znaczące\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "efekt kataliz występuje w parach (i,j) []\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "P-value dla Jarque_Bera wynosi 0.00597 \n",
      "Brak podstaw do odrzucenia H_0,rozkład reszt jest normalny\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "('Stan', array(0.04048028))\n",
      "('Typ telewizora', array(5.88649788e-08))\n",
      "('Przekątna ekranu', 1.723455164413816e-09)\n",
      "('const', 1.3798049965915669e-38)\n",
      "Wszystkie zmienne są statystycznie istotne na podstawie metody wstecznej krokowej\n",
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 2 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": "<Figure size 432x288 with 2 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAFKCAYAAAD2aJMUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdd3xUZbrA8d8zk4RQklCTQKRYECkqCAiKShG7LmJZsVzbIusils9ar7uLrl5dUde1LBZ0sdxdGxdFxL4gIKKL9KIEEZBqEhJIKKkzz/3jHJLJzKRNyiTh+X4+55M557znnOdlmPOc9z3vzBFVxRhjjAnkiXYAxhhjGh9LDsYYY0JYcjDGGBPCkoMxxpgQlhyMMcaEsORgjDEmhCUHY4xpBERkuohkisjaCtaLiDwrIhtFZLWInBSw7lwRSXfX3VcX8VhyMMaYxuE14NxK1p8H9HSnCcALACLiBaa66/sAV4pIn9oGY8nBGGMaAVVdCORUUmQM8IY6vgXaikhn4GRgo6puUtUi4G23bK1YcjDGmKYhDdgWML/dXVbR8lqJqe0OTD1ZN7NZ/q7JnOumRjuEejF0VGG0Q6gXrU7oH+0Q6k2ra6ZKrXdSg8+p9LvstzjdQYdMU9VpNThauHi1kuW1YsnBGGMipD5f9cs6iaAmySDYdqBrwPwRwE4groLltWLdSsYYEylfSfWn2psNXOuOWhoK5KrqLuA7oKeIHCkiccA4t2ytWMvBGGMipP7qn/Sr6sMSkbeAEUBHEdkOPADEAqjqi8DHwPnARuAgcIO7rkREJgGfAV5guqquq1FFwrDkYIwxkapBt1JVVPXKKtYrcEsF6z7GSR51xpKDMcZESIsKoh1CvbHkYIwxEapJt1JTY8nBGGMiVYfdSo2NJQdjjImQ1s0opEbJkoMxxkTKkoMxxphg6rduJWOMMUFstJIxxpgQ1nIwxhgTyu45GGOMCWajlYwxxoSy5GCMMSaYFjfP53iAJQdjjIlYTZ7n0NRYcjDGmAjZPQdjjDGh7If3jDHGBLNuJWOMMaEsOZim6L//PpP5S9fTIak1c565I2S9qvLIP+awYHk68S3ieGzSpfQ9Og2Ahcs38Mj0Ofj9fi4fPZgJlwxv6PAr1emUwfS9axLi8bB11sf89PpbYcsl9enFaa/+neX3P8yuuQsBOHLcJXQdewGCsHXWR2x+a2ZDhl6p2GOH0HrMHYh4KFjyIfnz/1lufcvhV9FiwNnOjMeLN7k7OX++AC0uJOnmqUhMLHhiKFrzJQe/+EcUahDe1xt388RnG/CrcvGANG4c1qPc+n0FJfxx1lp25Rbg8yvXntKdMf27AHD+s4toHefF4xG8HuHN8UOiUIPwtMhGKx2WRCQVeBoYDBQCW4A7VHVDNOOqrktGnsQ15w3l3mdnhF2/cPkGtuzK5vOpd7JqwzYenPYBM6ZMxOfz89DLs3n1gRtJ6ZDIZfc8z6jBx3FM15QGrkEFPB763Xs7/7nlbvIzsjj9jRfIWLiY/Zt/DinX+9YJZH27tHRRwtE96Dr2AhZdOxEtKebkZ6eQuehbDmzb0cCVCEM8tBl7J7kv34E/N5O2t75C0feL8GVuKS2Sv+BN8he8CUBc72HEn34Fmr8PgNxpt0FRPni8JE18gZj0bynZWutHCdeaz6889mk6L1w9gJTEeK5+ZQnDj+3I0Z3alJZ5d+k2jurYmmfG9SfnQBFjn1/M+cenEuv1ADDt2oG0axUXrSpUqC67lUTkXOAZnOdAv6KqjwWtvxu42p2NAXoDnVQ1R0S2APsAH1CiqoNqG4+ntjtorkREgPeB+ap6tKr2Ae4HGskZsmqD+x5JUkKrCtfPXfI9F48YgIjQv1c38g4UkJmTx+qN2+neuQNdU9sTFxvDBaedwNwlPzRg5JVr2/c4DmzbwcEdu9CSEnZ8Po+U4aeGlDvyirHsmreQwpw9pcva9OjO3jXf4y8sRH1+cpavInXkaQ0ZfoViuvbGt3s7/pyd4CuhcNVc4vqeXmH5uP6jKVz5RdmConznrzcG8caAaj1HXD1rd+bStV1LjmjXilivh3P6pjA/PSuk3IEiH6pKfpGPpJaxeD0ShWhrRv2+ak+VEREvMBU4D+gDXCkifcodS/UJVe2vqv2B/wYWqGpOQJGR7vpaJwaw5FCZkUCxqr54aIGqrlTVr0TkbhH5TkRWi8ifAUSkh4j8ICIvi8g6EflcRFq6625yy68SkZkiUvEZuwFl5OSR2jGpdD61QyIZOXlkZOeS2qFseUqHJDJy8qIRYlgtkztSkJFZOl+QuZuWyZ3KlYnv1JHUEafx88wPyy3f99Nm2g84gdikRDwtWpA8bAjxKckNEndVPEmd8OeW1cufm4knsVP4wrEtiOs1lKI188uWiYe2d7xGh8lzKNrwHSXbvq/fgKspM6+QlMT40vmUxHiy9pXvjhk3uCubdx/g7Ke/4vKXvuXuc3rhESc5iMDEf63gqpf/w8zl2xs09ir5fNWfKncysFFVN6lqEfA2MKaS8lcC4ftS64h1K1WsH7AseKGInA30xHkzBZgtImcAW93lV6rqTSLyLnAp8E/gPVV92d3+f4DfAM81SC0qEe7CUkQId73ZuK7hQqPRoMr0ufMWfnhuGvj95Zbv37KVn954m6FTn6DkYD55P/7UiEachPtXDn/1H9fnNEq2rC7tUnKK+tn79PVIfBsSrvsL3pQj8WVsrp9Qayuoqot/yqZXahum/ddJbNuTz+/+tZwB3YbSpkUMr14/mOSEFuQcKOLmfy6nR4fWDOzeLjpxB6nD/ztpwLaA+e1A2Jsr7sXlucCkwFCAz0VEgZdUdVptA7LkUHNnu9MKd74NTlLYCmxW1ZXu8mVAD/d1PzcptHXLfxZuxyIyAZgA8NIDv2XC5WfVR/ylUjsk8svu3NL5X7LzSG6XQHGJj1+yy5ZnZOeS3D6xXmOpifzMrHJX+/HJHSnI2l2uTNvex3LSo38CIK5tEsnDhuAv8ZGx4Gu2ffAJ2z74BIBeE39DQWZoF0c0+HMz8SSV1cuTlIw/b3fYsi1OPJPClf8Ou04L9lP803Lieg0lvxEkh+TEFmTklT33ICOvgE5tWpQrM3vVTm4Y1gMRoVv7VqS1bcmW3Qfol5ZEcoJTtn3rOEYd14l1O/MaUXLwV13IFfj5dk0LOIlX/8oALgK+DupSGqaqO0UkGfhCRNar6sJqBxeGdStVbB0wMMxyAf5yqO9PVY9R1UPDQgLbyj7Kku9rwCRVPR74MxBPGKo6TVUHqeqg+k4MAKMG92bW/BWoKivTt5LQKp7k9okcf0waW3btZltGDkXFJXy0aDWjBveu93iqK/f79bTumkbLLqlITAxpZ48iY+E35crMG3M18351FfN+dRW75i5g7ZRnyFjwNQBx7doCEJ+STOdRp7Pzs3kNXodwSravx9vxCDztOoM3hhYnnknR94tCykl8a2KPGkDhuq/KlrVui8S7N3hj4ojrOZiSrJ9Dto2Gvl0S2ZqTz449+RT7/Hy2LoMRx5bvLktNimfJZudcl72/kC3ZB0lr15L8Ih8HCp0vmuUX+fhmUw5Hd2rd4HWoiBYVV38K+Hy7U+DV/Xaga8D8EcDOCg47jqAuJVXd6f7NxLlXenJt62Yth4rNAx4VkZsCuoQGA3nAjSLyL1XdLyJpQHEV+0oAdolILM5ogwYZGvP7p95mydrN7Nl3gDPGP8at40ZT4jaDrzxnCMMH9mLB8nTOmvhXWraI5dFJlwIQ4/UyefyvGP/Qq/j8yqVnDqRnt8ZzH159ftY98RxDnpuCeL1sm/0J+zdtodulFwGwNeg+Q7CBjz9IXFIiWuJjzZRnKN63vyHCrprfx/4P/kbS+KfA46Xguzn4MjYTP/RiAAq+nQVAXN/hFG9YAsVlV+OehA4kXPFH8HhAPBSunkfxD4ujUo1gMR4P957bi4lvrsCvypgTu3B0chtmLHPuH1w+8AhuOv0oHpi9jstf/AYFbh91DO1axbF9z0F+/+5qwBn1dF6/VIYd0zGKtSlPfXV20/87oKeIHIlzfhgHXBVcSESSgOHANQHLWgMeVd3nvj4beKi2AUlwX60pIyJdcIayDgQKcIeyAucD491i+3HeKB8wR1X7udveBbRR1QdF5HfAPcDPwBogQVWvr/Tg62Y2yzdmznVTox1CvRg6qnmOd291Qv9oh1BvWl0ztda30rLuPKXan9NOf/2m0uOJyPk45xsvMF1VHxGRmwEODYwRkeuBc1V1XMB2R+G0FsC54H9TVR+pST3CxmPJoZGy5NCkWHJoeuoiOWTeMaTan9Pkp//TuMZ1VMG6lYwxJkLqb5bXcIAlB2OMiZi/yJKDMcaYIHV4Q7rRseRgjDER8lf/aw5NjiUHY4yJkDaWL9fXA0sOxhgTIUsOxhhjQli3kjHGmBC+4ib11YUaseRgjDERspaDMcaYEHbPwRhjTAi/37qVjDHGBKni6Z9NmiUHY4yJkLUcjDHGhPCVWHIwxhgTxFoOxhhjQlhyMMYYE8JnycEYY0wwazkYY4wJUeJrvsnBE+0AjDGmqfKpVHuqioicKyLpIrJRRO4Ls36EiOSKyEp3mlzdbSNhLQdjjIlQXXUriYgXmAqcBWwHvhOR2ar6fVDRr1T1wgi3rRFrORhjTITqsOVwMrBRVTepahHwNjCmmmHUZtsKWcuhkZpz3dRoh1AvLnz9lmiHUC8eurJ5vl/eT9dHO4R684drar8PfzW6iw4RkQnAhIBF01R1mvs6DdgWsG47MCTMbk4RkVXATuAuVV1Xg21rxJKDMcZEqDr3Eg5xE8G0ClaH25EGzS8HuqvqfhE5H5gF9KzmtjVm3UrGGBOhYr9Ue6rCdqBrwPwROK2DUqqap6r73dcfA7Ei0rE620bCWg7GGBOhmrQcqvAd0FNEjgR2AOOAqwILiEgqkKGqKiIn41zcZwN7q9o2EpYcjDEmQr5ad944VLVERCYBnwFeYLqqrhORm931LwKXAb8TkRIgHxinqgqE3ba2MVlyMMaYCPnCdvdHxu0q+jho2YsBr/8O/L2629aWJQdjjIlQXbUcGiNLDsYYE6Fm/CA4Sw7GGBOpojrsVmpsLDkYY0yEfNp8+5UsORhjTISsW8kYY0wISw7GGGNCWHIwxhgToqj2P2HUaFlyMMaYCNkNaWOMMSGsW8kYY0wIn3UrGWOMCWbJwRhjTAjrVjLGGBOiSP3RDqHeWHIwxpgIWbeSMcaYEH4bymqMMSaYtRxMk9TplMH0vWsS4vGwddbH/PT6W2HLJfXpxWmv/p3l9z/MrrkLAThy3CV0HXsBgrB11kdsfmtmQ4Zeqf/++0zmL11Ph6TWzHnmjpD1qsoj/5jDguXpxLeI47FJl9L36DQAFi7fwCPT5+D3+7l89GAmXDK8ocOv1NHDBnPOvZPweDyseO9jvp5e/j07dsSpjJx0A+pX/D4fnz0+lW0r1gJw2ydvUnjwIOrz4/f5eOXK30WjCmEdNWwwZ997C+LxsPK9j/lm+tvl1h874lTOmHQD+J3YP3/8eba79WqR0JoLHryLTsf0AFXmTH6SHau/j0ItQllyMIjIH3Ae2u0D/MBvgVOAaap6MJqxheXx0O/e2/nPLXeTn5HF6W+8QMbCxezf/HNIud63TiDr26WlixKO7kHXsRew6NqJaEkxJz87hcxF33Jg244GrkR4l4w8iWvOG8q9z84Iu37h8g1s2ZXN51PvZNWGbTw47QNmTJmIz+fnoZdn8+oDN5LSIZHL7nmeUYOP45iuKQ1cg/DE4+G8+2/nnxPuJi8ji/FvvUD6/MXs3lT2nm3+z3I2zF8MQHLPo7jsyck8P+b60vVv/Ob35O/Na+jQKyUeD+fefxtvTriHvIwsbnzreX6c/02l9Rr75J94acwNAJx97yQ2ff0d7935ZzwxMcS2bBGVeoRTl9+QFpFzgWdwngP9iqo+FrT+auBed3Y/8DtVXeWu2wLswzk/lajqoNrG46ntDg4HInIKcCFwkqqeAIwGtgF3AK2iGVtF2vY9jgPbdnBwxy60pIQdn88jZfipIeWOvGIsu+YtpDBnT+myNj26s3fN9/gLC1Gfn5zlq0gdeVpDhl+pwX2PJCmh4n/2uUu+5+IRAxAR+vfqRt6BAjJz8li9cTvdO3ega2p74mJjuOC0E5i75IcGjLxyaf2OY8/WHezdsQt/SQnrPp1Hr5Hl37Pi/ILS13Et49Em0Ofdpd9x5ATU6/tPv+TYSuoV2zIe3HrFtW5Ft4HHs/I95/HI/pISCvcdaLjgq1Cs/mpPlRERLzAVOA/oA1wpIn2Cim0GhrvnoIeBaUHrR6pq/7pIDGAth+rqDOxW1UIAVd0tIrcBXYAvRWS3qo4UkReAwUBL4P9U9QEozeqvAxcBscDlqrq+PgNumdyRgozM0vmCzN2069e7XJn4Th1JHXEa3/zuTtr26VW6fN9Pm+k18UZikxLxFRSSPGwIe3/YUJ/h1qmMnDxSOyaVzqd2SCQjJ4+M7FxSO5QtT+mQxOoft0UjxLASUjqSG/Ce5WXsJu343iHleo06jTNvH0/r9m1565b7S5cryjUvPYGqsnzGhyyf+VGDxF2VhJSO7MvIKp3Py8iqoF7DGOHW651b/gBAuyM6czAnlwsfvoeUY4/ilx9+5PMpU8slk2iqw26lk4GNqroJQETeBsYApf1nqro4oPy3wBF1dfBwrOVQPZ8DXUVkg4g8LyLDVfVZYCdOth7plvuDm7VPAIaLyAkB+9itqicBLwB3hTuIiEwQkaUisvTTrJ21DDn08YXBV5l97ryFH56bBv7yVzX7t2zlpzfeZujUJxjy3BTyfvwJ9TWdr/uEu5gWkbAf48b1kMcw0YSpTPq8RTw/5nreuWMyIybdULr81Wtv4+UrfsubE+9j0LiL6TbwhJBtG4twLZ70eV/z0pgbmHHHZIZPuh4Aj9dLau+eLH93Nv+44maK8gs49cZxDRxtxXxotafAz7c7TQjYVRpOb8Qh291lFfkN8EnAvAKfi8iyoP1GzFoO1aCq+0VkIHA6MBJ4R0TuC1P01+4bE4PT2ugDrHbXvef+XQZcUsFxpuE2FecMGlWrS5L8zCziU5JL5+OTO1KQtbtcmba9j+WkR/8EQFzbJJKHDcFf4iNjwdds++ATtn3g/N/rNfE3FGRm0VSkdkjkl925pfO/ZOeR3C6B4hIfv2SXLc/IziW5fWI0QgxrX0YWSQHvWWJKR/YFvWeBti5bTbuuXWjZNpH8vXnsz8oG4GDOXtLnLSKt33FsXba6wu0byr6M3SSkdCqdT0zpVBprONuWrSmtV15GFnkZWexc4zS013+xsFElh5oMZQ38fIcR7jol7M5FZCROcgjs6x2mqjtFJBn4QkTWq+rCagcXhrUcqklVfao63+0qmgRcGrheRI7EaRGc6fYJfgTEBxQpdP/6aICknPv9elp3TaNll1QkJoa0s0eRsfCbcmXmjbmaeb+6inm/uopdcxewdsozZCz4GoC4dm0BiE9JpvOo09n52bz6DrnOjBrcm1nzV6CqrEzfSkKreJLbJ3L8MWls2bWbbRk5FBWX8NGi1YwaHNq9ES071q2nffc02qal4omJoe+5o9gwv/x71q5rl9LXqb174o2JJX9vHrEt44lr1RJw+uyPOmUQmRs3N2j8Fdnp1ivJrVefc0eW3nw+pKJ6HcjeQ15GFu17OD0oPYYMIGtT0KCKKKpJy6EK24GuAfNH4PRMlOP2RrwCjFHV0gyrqjvdv5nA+zjdVLViLYdqEJFegF9Vf3QX9Qd+BnoACcBuIBE4AOSKSArOjaX5DR6sS31+1j3xHEOem4J4vWyb/Qn7N22h26UXAbB15oeVbj/w8QeJS0pES3ysmfIMxfv2N0TY1fL7p95mydrN7Nl3gDPGP8at40ZT4nZ7XXnOEIYP7MWC5emcNfGvtGwRy6OTnDwe4/UyefyvGP/Qq/j8yqVnDqRnt8YxUgmc9+yTR5/j6hec92zlrE/I+mkLAy933rNlMz6k9+gzOOGis/GXlFBSWMjMex4CoHX7dvz6aee1x+tl7Sdz+enr76JWl0Dq8/PZo89x5QtT8Hg9rJr1Cbt/+pmTLr8QgOUz5nDc6DM4/qKz8JeUUFxYxHv3PFy6/ed/eY6L/3I/nthY9m7fxZw/PR6tqoSo6kZzDXwH9HQvMncA43BGR5YSkW44PRD/paobApa3Bjyqus99fTbwUG0DkqYw2iHa3C6l54C2QAmwEZgAXAncAuxyb0i/BgwBNuG0FGar6mvuDelB7o3sQcCTqjqismPWtlupsbrw9VuiHUK9eOjKqdEOoV54G9ldmbr0h9Vza125i3sNqPbndFb6ikqPJyLnA0/jDGWdrqqPiMjNAKr6ooi8gtNjcajpVKKqg0TkKJzWAjgX/G+q6iM1rEoIazlUg6ouA0LHgToJ47mActdXsH2PgNdLgRF1GqAxJirq8uczVPVj4OOgZS8GvB4PjA+z3SbgxDoLxGXJwRhjImTfkDbGGBPCbz/ZbYwxJpjfWg7GGGOC1eFopUbHkoMxxkTInudgjDEmRPNtN1hyMMaYiFnLwRhjTAi7IW2MMSZEid2QNsYYE8xaDsYYY0L4m29usORgjDGRspaDMcaYEJYcjDHGhGjGI1ktORhjTKTsV1mNMcaEsG4lY4wxIZpvarDkYIwxEWvOycET7QCMMaap8qPVnqoiIueKSLqIbBSR+8KsFxF51l2/WkROqu62kbDkYIwxEdIaTJURES8wFTgP6ANcKSJ9goqdB/R0pwnACzXYtsYsORhjTIT8NZiqcDKwUVU3qWoR8DYwJqjMGOANdXwLtBWRztXctsbsnkMjNXRUYbRDqBcPXTk12iHUi8lv3RLtEOrFitufjHYIjVod3nNIA7YFzG8HhlSjTFo1t60xazkYY0wDEJEJIrI0YJoQuDrMJsG5p6Iy1dm2xqzlYIwxEQt3Xg5PVacB0ypYvR3oGjB/BLCzmmXiqrFtjVnLwRhjIiY1mCr1HdBTRI4UkThgHDA7qMxs4Fp31NJQIFdVd1Vz2xqzloMxxkRKqt9yqIyqlojIJOAzwAtMV9V1InKzu/5F4GPgfGAjcBC4obJtaxuTJQdjjIlY3XW+qOrHOAkgcNmLAa8VCDvyIdy2tWXJwRhjIiQ1uOfQ1FhyMMaYSNVRt1JjZMnBGGMiZC0HY4wxYTTfAZ+WHIwxJkLi8UY7hHpjycEYYyIk1nIwxhgTTOyGtDHGmBBiLQdjjDFBxJKDMcaYYHbPwRhjTAiPJzbaIdQbSw7GGBMh61YyxhgTwpKDMcaYECL2JThjjDFBrOVgjDEmhMcTF+0Q6o0lB2OMiZDHWg6mKYo9dgitx9yBiIeCJR+SP/+f5da3HH4VLQac7cx4vHiTu5Pz5wvQ4kKSbp6KxMSCJ4aiNV9y8It/RKEGFTt62GDOuXcSHo+HFe99zNfT3yq3/tgRpzJy0g2oX/H7fHz2+FS2rVgLwG2fvEnhwYOoz4/f5+OVK38XjSqE9d9/n8n8pevpkNSaOc/cEbJeVXnkH3NYsDyd+BZxPDbpUvoenQbAwuUbeGT6HPx+P5ePHsyES4Y3dPgVShg8lCMm3oF4vGR/MpuMt/+33Po2Jw7gqIcep3DXTgByFy3gl39Op8UR3ejxx4dLy7XonMau118m6713GjT+itg9hwiJSAdgrjubCviALHf+ZFUtquH++gNd3EfiVVZuBHCXql5Ys4jL7aML8KyqXhbpPqJKPLQZeye5L9+BPzeTtre+QtH3i/Blbiktkr/gTfIXvAlAXO9hxJ9+BZq/D4DcabdBUT54vCRNfIGY9G8p2Vrrx9LWCfF4OO/+2/nnhLvJy8hi/FsvkD5/Mbs3/VxaZvN/lrNh/mIAknsexWVPTub5MdeXrn/jN78nf29eQ4depUtGnsQ15w3l3mdnhF2/cPkGtuzK5vOpd7JqwzYenPYBM6ZMxOfz89DLs3n1gRtJ6ZDIZfc8z6jBx3FM15QGrkEYHg9db72TjffeTnFWJr2mTid38VcUbN1Srtj+NavY9Me7yi0r3L6V9JuvK91Pv7dns3fRggYKvGoNlRxEpD3wDtAD2AL8WlX3BJXpCryBc671A9NU9Rl33YPATZSdf++v6jxar20iVc1W1f6q2h94EfjbofmaJgZXf5wHbNc7Vd1Z28QgUbysiOnaG9/u7fhzdoKvhMJVc4nre3qF5eP6j6Zw5RdlC4rynb/eGMQbA6r1HHH1pfU7jj1bd7B3xy78JSWs+3QevUaeWq5McX5B6eu4lvFoI4q/MoP7HklSQqsK189d8j0XjxiAiNC/VzfyDhSQmZPH6o3b6d65A11T2xMXG8MFp53A3CU/NGDkFWvVqw+FO7dTtGsnWlLCnvn/JmnYGTXeT8KAQRTu3EFx5i/1EGVkRLzVnmrpPmCuqvbEueC+L0yZEuBOVe0NDAVuEZE+AesDz79VPm+6oTvMWorIZhGJBRCRRBHZIiKxIjJfRJ4WkcUislZETg7cUETigIeAK0RkpYhcISKtRWS6iHwnIitEZEzwASsqIyIfi8gJ7usVIjLZff2wiIwXkR4istZd9op7zJUikiUiD4jjCTfWNSJyhVt2hIh8KSJvAmvcZbNEZJmIrBORCfX2rxvAk9QJf25m6bw/NxNPYqfwhWNbENdrKEVr5pctEw9t73iNDpPnULThO0q2fV+/AddAQkpHcjPK6paXsZuE5NC69Rp1GhM/eI0rpz7Kh5OfKF2uKNe89ATj336Rky69oEFirisZOXmkdkwqnU/tkEhGTh4Z2bmkdihbntIhiYycxtEyiuvYiaLMsverKCuT2A6h71frPv047qU3OPrRp4jvfmTI+nYjz2LPl1+ELI8m8XirPdXSGOB19/XrwMXBBVR1l6oud1/vA34A0iI9YEPfc8gH5gMXALOAccBMVS12f/q2taqeKiJnANOBfoc2VNUi9wQ+SFUnAYjIo8A8Vb1RRNoCS0Tk30HH/EMFZRYCp4vIFpyMO8wtfxpQrnNeVce7x+sOfAa8BlyC05I5EegIfCciC91NTgb6qepmd/5GVc0RkZZuuZmqmh3Bv18NhKOT7WwAACAASURBVPsp4fBXz3F9TqNky+rSLiWnqJ+9T1+PxLch4bq/4E05El/G5rDbN7wwdQvTMkift4j0eYvoNvAERky6gX9OuBuAV6+9jf1Z2bRq35ZrXnqC3Vu2sXXZ6voOuk6EawCJSNh3ttH8mHTYn7UuH/HBH9NZd9VY/AX5JJ58Ckf+eQo/XP/rsl3ExJB0ymnsfOX5eg62ZrwNN1opRVV3gZMERCS5ssIi0gMYAPwnYPEkEbkWWIrTwtgTZtNS0bjV/gpwg/v6BuDVgHVvAajqQiDRPZlX5mzgPhFZiZN04oFu1SzzFXAGTjL4CGgjIq2AHqqaHnwgEYkHZgCTVPVnd7u3VNWnqhnAAmCwW3xJQGIAuE1EVgHfAl2BnuEqIyITRGSpiCx9Y1Xtms7+3Ew8SWX/fzxJyfjzdoct2+LEMylcGZxTHVqwn+KflhPXa2it4qlL+zKySEopq1tiSkf2ZYWvG8DWZatp17ULLdsmArA/y8nLB3P2kj5vEWn9jqvfgOtQaodEftmdWzr/S3Yeye0SSO2QxC/ZZcszsnNJbp8YjRBDFGVlEpdc9n7FdUqmOLv8++U/eBB/gdOVmbfkGyQmBm9iWUso8eRTOPhjOiV7Kz2fNbiadCsFfr7daUL5fcm/3Z6I4CmkR6TymKQNMBO4Q1UPNR9fAI7GuaDdBfy1qv00eHJQ1a+BHiIyHPCq6trA1cHFq9idAJcG9KN1U9XgjtaKynwHDAJOx2lFrMC5YbOsgmO9CLynqofOopVdmB0oPbhzc3w0cIqqnugeJz7cRqo6TVUHqeqga09MrWT3VSvZvh5vxyPwtOsM3hhanHgmRd8vCikn8a2JPWoAheu+KlvWui0S38aZiYkjrudgSrJ+Dtk2WnasW0/77mm0TUvFExND33NHsWH+N+XKtOvapfR1au+eeGNiyd+bR2zLeOJatQQgtmU8R50yiMyNjaVFVLVRg3sza/4KVJWV6VtJaBVPcvtEjj8mjS27drMtI4ei4hI+WrSaUYN7RztcAA6m/0CLtK7EpXZGYmJoN2I0uYu/Klcmpl370tetevVBPIIvryzZNcYuJQCRmGpPgZ9vd5oWuC9VHa2q/cJMHwAZItLZOaZ0BjLDxyOxOInhX6r6XsC+M9wLWT/wMk7vRqWiNZT1DZxWwsNBy68AvhSR04BcVc0NWr8PSAiY/wy4VURuVVUVkQGquiJom7Bl3G6qbcCv3Tg6AU+6UzkicguQoKqPBSxeCPxWRF4H2uO0Qu4Ggi9Dk4A9qnpQRI7DuVFU//w+9n/wN5LGPwUeLwXfzcGXsZn4oU5XZcG3swCI6zuc4g1LoLjsBq4noQMJV/wRPB4QD4Wr51H8w+IGCbs61Ofnk0ef4+oXpiBeLytnfULWT1sYePlFACyb8SG9R5/BCRedjb+khJLCQmbe8xAArdu349dPO689Xi9rP5nLT19/F7W6BPv9U2+zZO1m9uw7wBnjH+PWcaMp8fkAuPKcIQwf2IsFy9M5a+JfadkilkcnXQpAjNfL5PG/YvxDr+LzK5eeOZCe3RrBSCUAv4/tz/2Vox97GvF4yP50DgU/b6bDhWMByJ7zPm3PGEXHi8aCz4e/qJAt/zO5dHNp0YKEgSez9ekp0apBhTwNN+ZkNnAd8Jj794PgAuL0zf8D+EFVnwpa1/lQtxQwFlgbvH3I/hpqFIc7lGq/qj4pIqnAZqCzqu51188HvgGGA4k4/fRLgvbRHudkHwv8Becf7GngVJwr+S2qemHgUFa3nz+kjLu/h4Ez3fscXYAdwEBVXe722c1R1X4ishkoBg66obwIvAQ8DpyH08L5H1V9J3gYrYi0wLm/kgak4yShB1V1fmX/XrvvGdY0htfU0POftoh2CPVi8lu3RDuEerHi9pBrpWZjwL+/qfVtmdOHTaz25/Srr5+P+HjifC3gXZwu8a3A5e59zC7AK6p6vntR/RXOQBi/u+n9qvqxiPwvTpeS4gyF/W1Asgh/zGgM8RORy4AxqvpfAcvm45xUlzZ4QI2QJYemxZJD01MXyeGM026r9ud04aJnG80Ygepo8G4lEXkO52q7Qb6vYIwx9cXjaZ4XOxCF5KCqt1awfEQDh2KMMbXSgPccGpz9tpIxxkRIPM33FNp8a2aMMfVMpPmeQptvzYwxpp7Zr7IaY4wJ4fGG/T5rs2DJwRhjImTdSsYYY0JZcjDGGBPMRisZY4wJYd1KxhhjQlnLwRhjTDDx2s9nGGOMCWbdSsYYY4KpdSsZY4wJ4bFvSBtjjAlmycEYY0wwf0xctEOoN5YcjDEmUtZyMMYYE0w9ngY5joi0B94BeuA8A/rXqronTLktwD7AB5So6qCabB+oYWpmjDHNkHq81Z5q6T5grqr2BOa68xUZqar9DyWGCLYHLDkYY0zE/F5PtadaGgO87r5+Hbi4vre3bqVGqtUJ/aMdQr3wfro+2iHUixW3PxntEOrFgGfuinYIjVpDdSsBKaq6C0BVd4lIckUhAZ+LiAIvqeq0Gm5fypKDMcZEyBdT/VOoiEwAJgQsmhZw8kZE/g2khtn0DzUIaZiq7nRP/l+IyHpVXViD7UtZcjDGmAhpDbqL3EQwrZL1oytaJyIZItLZvervDGRWsI+d7t9MEXkfOBlYCFRr+0B2z8EYYyKkHqn2VEuzgevc19cBHwQXEJHWIpJw6DVwNrC2utsHs5aDMcZEyO+t9Um/uh4D3hWR3wBbgcsBRKQL8Iqqng+kAO+LCDjn9jdV9dPKtq+MJQdjjIlQHbQIqncc1WzgzDDLdwLnu683ASfWZPvKWHIwxpgINVRyiAZLDsYYEyF/rCUHY4wxwZrxkB5LDsYYE6nm+7t7lhyMMSZi1nIwxhgTwpKDMcaYYJ4YjXYI9caSgzHGREis5WCMMSZYM34QnCUHY4yJVMP9YnfDs+RgjDEREo/dczDGGBPEWg7GGGNC1OBZP01OM66aMcbUL2s5GGOMCWHJwRhjTAiv3ZA2xhgTzFoOpkn6euNunvhsA35VLh6Qxo3DepRbv6+ghD/OWsuu3AJ8fuXaU7ozpn8XAM5/dhGt47x4PILXI7w5fkgUalCxo4YN5ux7b0E8Hla+9zHfTH+73PpjR5zKGZNuAL8fv8/H548/z/YVzuN0WyS05oIH76LTMT1AlTmTn2TH6u+jUItQCYOHcsTEOxCPl+xPZpPx9v+WW9/mxAEc9dDjFO7aCUDuogX88s/ptDiiGz3++HBpuRad09j1+stkvfdOg8Zfkf/++0zmL11Ph6TWzHnmjpD1qsoj/5jDguXpxLeI47FJl9L36DQAFi7fwCPT5+D3+7l89GAmXDK8ocOv0GGdHETEB6xxy/4AXKeqB2tzUBHZAgxS1d3VLH+/qj5am2NW4xjXuzFNqs/jNBSfX3ns03ReuHoAKYnxXP3KEoYf25GjO7UpLfPu0m0c1bE1z4zrT86BIsY+v5jzj08l1uv8j5927UDatYqLVhUqJB4P595/G29OuIe8jCxufOt5fpz/Dbs3/VxaZvN/lrNh/mIAknsexdgn/8RLY24A4Ox7J7Hp6+94784/44mJIbZli6jUI4THQ9db72TjvbdTnJVJr6nTyV38FQVbt5Qrtn/NKjb98a5yywq3byX95utK99Pv7dnsXbSggQKv2iUjT+Ka84Zy77Mzwq5fuHwDW3Zl8/nUO1m1YRsPTvuAGVMm4vP5eejl2bz6wI2kdEjksnueZ9Tg4zima0oD1yC82Ab6hrSItAfeAXoAW4Bfq+qeoDK93DKHHAVMVtWnReRB4CYgy113v6p+XNkxq5P38lW1v6r2A4qAm4MCaoh/nvsb4BhVEpEm09JauzOXru1ackS7VsR6PZzTN4X56Vkh5Q4U+VBV8ot8JLWMxdsEHnvYpd9x5Gzdwd4du/CXlPD9p19y7MhTy5Upzi8ofR3bMh7U6RuOa92KbgOPZ+V7zufCX1JC4b4DDRd8JVr16kPhzu0U7dqJlpSwZ/6/SRp2Ro33kzBgEIU7d1Cc+Us9RBmZwX2PJCmhVYXr5y75notHDEBE6N+rG3kHCsjMyWP1xu1079yBrqntiYuN4YLTTmDukh8aMPLKeT3Vn2rpPmCuqvYE5rrz5ahqunuu7g8MBA4C7wcU+duh9VUlBqj5D85+BRwjIiNE5EsReRNYIyI3i8hKd9osIl8CiMjZIvKNiCwXkRki0iZwZyLSUkQ+FZGb3PlZIrJMRNaJyAR32WNAS3ff/xKRHiLyg4i87Jb7XERaumVvEpHvRGSViMwUkZD/jSLSWkSmu+VWiMiYMGUucOPuKCKvichTbp2miMjJIrLY3Xaxm60RketF5D23Pj+KyOMB+9sf8PoyEXmthv/uNZaZV0hKYnzpfEpiPFn7CsuVGTe4K5t3H+Dsp7/i8pe+5e5zeuERceOEif9awVUv/4eZy7fXd7g1kpDSkX0ZZYkuLyOLhOSOIeV6jRrGbz94lSumPsKcyU8C0O6IzhzMyeXCh+/hN++8yAUP3ukkj0YgrmMnijIzS+eLsjKJ7dAppFzrPv047qU3OPrRp4jvfmTI+nYjz2LPl1/Ua6x1LSMnj9SOSaXzqR0SycjJIyM7l9QOZctTOiSRkZMXjRDD8niqP9XSGOB19/XrwMVVlD8T+ElVf66iXIWqHbJ71XweThcTwMnAH1S1j6q+6GarwcB24CkR6Qj8ERitqicBS4HfB+yyDfAh8Kaqvuwuu1FVBwKDgNtEpIOq3kdZ6+Vqt1xPYKqq9gX2Ape6y99T1cGqeiJOF9hvwlTlD8A8VR0MjASeEJHWAfUci5OVzw/o9jrWrcedwHrgDFUdAEwGAru7+gNXAMcDV4hI1yr+WRtWUKNg8U/Z9Eptw+d3nM7bE4bw2Kfr2V9YAsCr1w/mrZuG8PerBvDOd9tZ9vOeMDtsPFRDR42kz/ual8bcwIw7JjN80vUAeLxeUnv3ZPm7s/nHFTdTlF/AqTeOa+BoKyDhWm3l63Xwx3TWXTWW9b+9lqxZMzjyz1PK7yImhqRTTmPvgrn1GGjdC/P2ISKEGwvUmNq2DdhySFHVXQDu3+Qqyo8D3gpaNklEVrsXx+2qOmB1Qm4pIitxTu5bgX+4y5eo6uagss/gnHg/BIYCfYCv3e2vA7oHlP0AeFVV3whYdpuIrAK+BbriJIFwNqvqSvf1Mpx+OIB+IvKViKwBrgb6htn2bOA+N6b5QDzQzV03ErgXuCCoP2+Gqvrc10nADBFZC/wt6BhzVTVXVQuA74PqWyURmSAiS0Vk6fR562qyaYjkxBZk5JV1rWTkFdCpTfm+9dmrdjLquGREhG7tW5HWtiVbdjtdLMkJTtn2reMYdVwn1u1sPFdr+zJ2k5BSdkWdmNKJ/VnZFZbftmwN7bp2oWXbRPIyssjLyGLnmvUArP9iIam9K/pv1rCKsjKJSy77zMd1SqY4u/xtOf/Bg/gL8gHIW/INEhODN7Hsyjrx5FM4+GM6JXsbdzIPltohkV9255bO/5KdR3K7BFI7JPFLdtnyjOxcktsnRiPEsGqSHAI/3+40IXBfIvJvEVkbZgrp3aiMiMQBvwICb/C8AByNcwG7C/hrVfupyT2H/qp6q6oWucvLddS6N3S7A38+tAj4ImDbPqoaeCX/NXCeiHO5JCIjgNHAKe6V/wqcE3c4gf0jPspurL8GTFLV4904wm0vwKUBcXVT1UOdmJuABJyWQqDAuj4MfOneg7ko6BgVxRV4AVRhH4aqTlPVQao66MZR4fJa9fXtksjWnHx27Mmn2Ofns3UZjDi2fBdFalI8SzbnAJC9v5At2QdJa9eS/CIfB9wWRH6Rj2825XB0p9Yhx4iWnevW0757GklpqXhiYuhz7sjSm8+HtOvapfR1au+eeGNiyd+bx4HsPeRlZNG+xxEA9BgygKxNEbe869TB9B9okdaVuNTOSEwM7UaMJnfxV+XKxLRrX/q6Va8+iEfw5ZWdPJtilxLAqMG9mTV/BarKyvStJLSKJ7l9Iscfk8aWXbvZlpFDUXEJHy1azajBvaMdbqnYmOpPgZ9vd5oWuC9VHa2q/cJMHwAZItIZwP2bGS4e13nAclXNCNh3hqr6VNUPvIzT81OpOrnBKiIDgbuA092Dg3P1P1VEjlHVjW7//xGqusFdPxn4E/A88DucK/I9qnpQRI7DaXkcUiwisapaXEUoCcAuEYnFaTnsCFPmM+BWEblVVVVEBqjqCnfdz2493heRy1U13OV7UsB+r68inkMyRKQ3kA6MBfZVc7uIxXg83HtuLya+uQK/KmNO7MLRyW2Yscy5f3D5wCO46fSjeGD2Oi5/8RsUuH3UMbRrFcf2PQf5/burAWfU03n9Uhl2TGiffrSoz89njz7HlS9MweP1sGrWJ+z+6WdOuvxCAJbPmMNxo8/g+IvOwl9SQnFhEe/dUzbM8/O/PMfFf7kfT2wse7fvYs6fHq/oUA3L72P7c3/l6MeeRjwesj+dQ8HPm+lw4VgAsue8T9szRtHxorHg8+EvKmTL/0wu3VxatCBh4MlsfXpKRUeImt8/9TZL1m5mz74DnDH+MW4dN5oSn9MYv/KcIQwf2IsFy9M5a+JfadkilkcnOT3FMV4vk8f/ivEPvYrPr1x65kB6dmscI5UAvA3XxzUbp/flMffvB5WUvZKgLiUR6XyoWwrnHLS2qgNKuL7aoJ3uV9XgG8kjgLtU9UJ3/lXgHMqy2VJVHS8io4ApwKH+jD+q6uxDQ1mBbGA6zvCqPwGzgDSck2gn4EFVnS8iU3CaSctx7hnMca/cEZG7gDaq+qCI/A64B+ckvwZIUNXrg2JvCTwNnIrTitiiqhcGDmUVkQHAv3BaBn9yj/d/7van4NwQygLmAf+lqj2Ch8KKyBzgSTf+y9x/h204b0qb4LiCHfznLc3yq5d/e3x9tEOoF+cn12p0d6M14Jm7qi7UVPW9tNan9kkfLq/25/TvF50U8fFEpAPwLk4X+FbgclXNEZEuwCuqer5brhXOeeYoVc0N2P5/cbqUFGco7G8DkkX4Y1aVHKJBRJKAR1X1lmjHEi2WHJoWSw5NUB0kh9s/qn5yeOaCyJNDNDTW7/e1AgaLSLP4QpoxpnmK8Ui1p6amUX6py23uVHnDxBhjoqkOhqg2Wo0yORhjTFMQ10A/nxENlhyMMSZC1nIwxhgToin8FlmkLDkYY0yErOVgjDEmRAN+Ca7BWXIwxpgIWbeSMcaYEHExlhyMMcYEsXsOxhhjQnjCPoOjebDkYIwxEbKWgzHGmBB2Q9oYY0wIuyFtjDEmhLUcjDHGhLB7DsYYY0J4rOVgjDEmmHUrGWOMCdGcu5WacdWMMaZ+xcZ4qj3VhohcLiLrRMQvIoMqKXeuiKSLyEYRuS9geXsR+UJEfnT/tqvqmJYcjDEmQl5P9adaWgtcAiysqICIeIGpwHlAH+BKEenjrr4PmKuqPYG57nylLDkYY0yEvB6p9lQbqvqDqqZXUexkYKOqblLVIuBtYIy7bgzwuvv6deDiqo5p9xwaqVbXTG2wO10iMkFVpzXEsf5wTUMcxdGQ9WpozbVuTa1eZ3XpWe3PqYhMACYELJpWx3VNA7YFzG8HhrivU1R1F4Cq7hKR5Kp2Zi0HA+X/wzYnzbVe0Hzr1lzrhapOU9VBAVO5xCAi/xaRtWGmMRXtM0i4RKWRxmstB2OMaQRUdXQtd7Ed6BowfwSw032dISKd3VZDZyCzqp1Zy8EYY5qH74CeInKkiMQB44DZ7rrZwHXu6+uAD6ramSUHA9Bk+nhrqLnWC5pv3ZprvWpFRMaKyHbgFOAjEfnMXd5FRD4GUNUSYBLwGfAD8K6qrnN38Rhwloj8CJzlzld+TNWIu6SMMcY0U9ZyMMYYE8KSgzHGmBCWHIwxxoSwoazGNBHu7+H0BOIPLVPVCn9OoSkQkVcJMxZfVW+MQjgmgCWHw5CIxAO/AfpS/kTT5D+QInIscDfQnYD/36o6KmpB1QERGQ/cjjN2fSUwFPgGaNL1AuYEvI4HxlI2Nt9EkSWHw9P/AuuBc4CHgKtxhr41BzOAF4GXAV+UY6lLtwODgW9VdaSIHAf8Ocox1ZqqzgycF5G3gH9HKRwTwJLD4ekYVb1cRMao6usi8ibO2OjmoERVX4h2EPWgQFULRAQRaaGq60WkV7SDqgc9gW7RDsJYcjhcFbt/94pIP+AXoEf0wqlTH4rIROB9oPDQQlXNiV5IdWK7iLQFZgFfiMgemkH3i4jsw7nnIO7fX4B7oxqUAexLcIclt/96JnAC8CrQBpisqi9GNbA6ICKbwyxWVT2qwYOpJyIyHEgCPnV/mtmYOmfJwZhGTkQ8wGpV7RftWOqDiKQROoCgSY/Cag6sW+kwJCItgEtxupICP5APRSumuuR2lfWh/EisN6IXUe2oql9EVolIN1XdGu146pKITAGuAL6nbACBUskTz0zDsORwePoAyAWWEdAv3xyIyAPACJzk8DHOIxMXAU02Obg6A+tEZAlw4NBCVf1V9EKqExcDvVS1Wf0/bA4sORyejlDVc6MdRD25DDgRWKGqN4hICvBKlGOqC01+2GoFNgGxNLOLlObAksPhabGIHK+qa6IdSD3Id7thSkQkEeehJk3+ZrSqLnAT3WB30RJVrfKBLU3AQWCliMyl/Oiy26IXkgFLDoer04Dr3ZE9hbjDCFX1hOiGVSeWukM+X8bpNtsPLIluSLUnIr8GngDm47xfz4nI3ar6f1ENrPZmU/ZAGtOI2Gilw5CIdA+3XFV/buhY6pOI9AASVXV1lEOpNRFZBZx1qLUgIp2Af6vqidGNzDRX9qushyE3CXQFRrmvD9JM/i+43RMAqOoWVV0duKwJ8wR1I2XTDN4zEekpIv8nIt+LyKZDU7TjMtatdFhyR/QMAnrhfAkuFvgnMCyacdWG+2OCrYCO7q+XirsqEegStcDqzqfuoyHfcuevAD6JYjx15VXgAeBvwEjgBsreOxNF1q10GBKRlcAAYLmqDnCXrW7K9xxE5HbgDpxEsCNg1T7gZVX9e1QCq0MicgnO/SIBFqrq+1EOqdZEZJmqDhSRNap6vLvsK1U9PdqxHe6s5XB4KlJVFREFEJHW0Q6oDiwG3gUuU9XnROQ6nC/6bQHejGZgdUFE/gS8pqrvBSyboKrTohhWXShwvwH+o4hMwknsyVGOydAM+ixNRN4VkZeAtiJyE85PJL8c5Zhq6yWg0E0MZwB/AV7H+bJfUz+BAtwKfCYiIwOW3RytYOrQHTjdgbcBA4FrgOuiGpEBrOVwWBGRY4AUVX1SRM4C8nDuO3yC823ipswb8MurVwDT3GcFzHS70Zq6HcAYYIaI/J+qPkET75sXES/wa1W9G2fI8Q1RDskEsJbD4eVpnD54VPULVb1bVe/CSQxPRzWy2vOKyKGLnTOBeQHrmsVFkPu7SsOBPiIyA2gZ5ZBqRVV9wEARadJJrrlqFh8aU209wo35V9Wl7ncCmrK3gAUishvIB76C0tZSbjQDqyNLAVS1ALhBRG7B6YZp6lYAH7jJLvA3o96reBPTEGy00mFERDaq6jE1XddUiMhQnB+o+1xVD7jLjgXaqOryqAZnwhKRV8Ms1ubwPPOmzpLDYcR9Pu88VX05aPlvgLNV9YroRGYqIiLvquqvRWQNzk9Zl9OUhx8DiEj74Kf0iciRqhruoU2mAVlyOIy4P9z2PlCE87tD4HwZLg4Yq6q/RCs2E56IdFbVXc31J09E5GvgPFXNc+d7AzOa64ONmhJLDochdzjkoQ/fOlWdV1l5E30iciPwlar+GO1Y6pKIXADcA1yAM3LuDeBqVW0OI8yaNEsOxjQBIvIQzreju+O0+r7CSRZN/iQqIhfjJIgE4JLmlgCbKksOxjQhItISuAm4C0hTVW+UQ4qIiDxH+Xsoo3Ae/LMF7HkOjYENZTWmCRCRP+L8MGIbnOGfd+EO122ilgbNLwtbykSNtRyMaQJEZDlQAnwELAC+db/z0OS5raFuqpoe7VhMGfuGtDFNgKqehPPN7yXAWcAaEVkU3ahqT0QuAlYCn7rz/UXEngzXCFi3kjFNgIj0A07H+fmMQcA2mna30iEPAifjPP4UVV0pIkdGMyDjsORgTNMwBVgIPAt8p6rFUY6nrpSoam7QzytZX3cjYMnBmCZAVS8I6JtvLokBYK2IXIXzw4k9cX66e3GUYzLYPQdjmoRm3Dd/K9AXKMR5KFMuzjMeTJTZaCVjmgARWYbzXYD5zeXRrqZxs5aDMU1Diao2h58eN02E3XMwpmmwvnnToKzlYEzTENg3/xbOI16tb97UG7vn8P/t3Wmo5XMcx/H3x5AxmbGFZN+XEBqRtShCzROEJmVPPLA8QbINWcsDamTKUpMQyhqSbCOaLHFtZUpqQhKTm334ePD/n5ljzrjCvfd7/vd+XnW65/zOPfV5dL7nt0dEGUkzgbNpCt/MXnsu+6mXYaWIISbpKcZY92973iTGmQiLgU+AY4EFwHzg49JEAaTnEDHUJB051vu2X5msLBNB0ru29++tvJK0HvC87aOqs0136TlEDLH+L/8pekBdb0PfivaIkK+AHeriRE8mpCM6YApvglskaRPgKuBJ4CPg1tpIARlWiuiEbIKLyZZhpYhuWNsBdZ0naX3gRJqhpFXfR7YXVGWKRopDRDdM1U1wT9Ccp/Q2zR6OGBIZVoroAEmzgCuBY9qm54HrbXf6C1XSB7b3rs4Rg1IcIjpA0sm2H/mntq6RtAi40/ZIdZb4qxSHiA6Q9E57VeiYbV0j6SNgF+AzmmElAc5Ee73MOUQMMUnHAccDW0u6o++tOcDKmlTj6rjqALF2KQ4Rw+0L4C1gHs2kbc8ocElJovF1g+3T+xskLQZO/5v/j0mSYaWIDpC03hS7HhQYHBqTNAMYsb1XYawgO6QjOmGqFQZJV0gaBfaV9H37GAW+plneGsXSc4iIv77B/gAAA45JREFUMpJusn1FdY4YlOIQ0SGS5tCs5hmtzjJe2rOVduWv9zm8KukC2wvrkk1vKQ4RHSBpLnAfMJtmuecK4Czbb4/5wSEn6RzgImAbmoMFDwbeoDmAb67tgwvjTWspDhEdIOl94ELbr7WvDwMWdn0/gKQR4EDgTdv7SdoDuA44F5hl+6vSgNNYlrJGdMNorzAA2F7STuB23c+2f5aEpPVtfyJpd9vf09yTHUVSHCK6Yamku4EHaa4NPQV4WdIBALbfqQz3PyyXtDHwOPCCpO9o9nZEsQwrRXSApJfGeNtT4VrN9krUjYBnp9rS3S5KcYjoAEkzbP9enWO8STrb9j1rtN1s+/KqTNHIJriIblgm6TZJe1YHGWcnSZrfeyFpIbB5YZ5opecQ0QGSZgOnAmfS/Ki7F3ionbjtLEkb0NwdfS/NIXzf2r64NlVAikPEUJO0ru2Va7QdQTMxvTHwKM2lP8sq8v1XkjbtezmbZkL6deBqANvfVuSK1VIcIoZY72C69kC6E4CzgO2BxcADwOHAjbZ3K4z5r0n6jGbVlfr+9tj2TiXBYpUsZY3ohk+Bl4BbbL/R1/5o25PoFNs7VmeIsaXnEDHEJC0HbgdmAT/R/MpexfbtFbnGS3s39qXAdrbPk7QrsLvtp4ujTXtZrRQx3GYAG/b9nb3Go+vuA34FDmlfLwduqIsTPRlWihhuX9peUB1iAu1s+xRJpwHY/kmS/ulDMfHSc4gYblP9i/LXdjmrASTtDPxSGykgPYeIYXd0dYAJdi3wHLCtpAeAQ4EzKgNFIxPSEVFK0mY09ziI5ujub4ojBRlWiohCkl4EDrL9jO2nbX8jaVF1rkhxiIhaOwKXSbqmr21uVZhYLcUhIiqtoJlX2VLSU5I2qg4UjRSHiKgk2yttXwA8BiwBtijOFGS1UkTUuqv3xPb97Z3SFxbmiVZWK0VECUnrAO/b3rs6SwzKsFJElLD9B/CepO2qs8SgDCtFRKWtgA8lLQV+6DXanlcXKSDFISJqXVcdINYuxSEiJp2kmcD5wC7ACHDPmjfeRa1MSEfEpJP0MPAb8BrN3dGf276oNlX0S3GIiEknacT2Pu3zdYGltg8ojhV9slopIir81nuS4aThlJ5DREw6Sb+zenWSgA2AH9vntj2nKls0UhwiImJAhpUiImJAikNERAxIcYiIiAEpDhERMSDFISIiBqQ4RETEgD8BEo0UpTER85kAAAAASUVORK5CYII=\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#wczytanie danych i stworzenie dummy variable\n",
    "Dane = pd.get_dummies(df2, drop_first=True)\n",
    "#Czyste przypisanie zmeinyych endo i egzo\n",
    "Y = Dane[\"Cena\"]\n",
    "X = Dane.iloc[:,1:]\n",
    "X = sm.add_constant(X).astype(float)\n",
    "\n",
    "#tranformacja danych \n",
    "Y = np.log(Y)\n",
    "X[\"Przekątna ekranu\"] = X[\"Przekątna ekranu\"]**2\n",
    "#początkowy fit\n",
    "final_model = sm.OLS(Y,X).fit()\n",
    "\n",
    "#korekcja heterodastyczności \n",
    "X[\"Przekątna ekranu\"] = X[\"Przekątna ekranu\"] / np.sqrt(np.exp(final_model.resid)**2)\n",
    "Y = np.log(np.exp(Y)/ np.sqrt(np.exp(final_model.resid)**2))\n",
    "\n",
    "final_model = sm.OLS(Y,X).fit()\n",
    "print(final_model.summary())\n",
    "#zapisane  modelu \n",
    "save_to_latex(final_model.summary(), \"final_model\")\n",
    "print(\"\\n\\n\\n\")\n",
    "#Współczynik Determinacji \n",
    "print(f\"F-vaule dla modelu wynosi:{final_model.f_pvalue} \\n Brak podstaw do orzucenia H_0\\n R^2 jest statystycznie znaczące\")\n",
    "print(\"\\n\\n\\n\")\n",
    "#Badanie efektu katalizy\n",
    "R_corr = associations(df2.iloc[:,1:],nan_strategy= \"drop_features\")\n",
    "R_corr = np.array(R_corr[\"corr\"])\n",
    "\n",
    "R = associations(df2,nan_strategy= \"drop_features\")\n",
    "R = np.array(R[\"corr\"].iloc[:,0])[1:]\n",
    "\n",
    "R_sorted = R.argsort()\n",
    "r_0 = R[R_sorted]\n",
    "r = R_corr[R_sorted]\n",
    "\n",
    "pairs = []\n",
    "for i in range(r.shape[0]):\n",
    "    for j in range(r.shape[0]):\n",
    "        if abs(r_0[i]) < abs(r_0[j]):\n",
    "            r_ij = r[i, j] * sign(r_0[i]) * sign(r_0[j])\n",
    "            if r_ij > (abs(r_0[i]) / abs(r_0[j])) or r_ij < 0:\n",
    "                pairs.append((i, j))\n",
    "print(f\"efekt kataliz występuje w parach (i,j) {pairs}\")\n",
    "print(\"\\n\\n\\n\")\n",
    "#Normalność reszt \n",
    "Jarque_Bera_pv = float(final_model.summary().tables[2].data[2][3])\n",
    "print(f\"P-value dla Jarque_Bera wynosi {Jarque_Bera_pv} \\nBrak podstaw do odrzucenia H_0,rozkład reszt jest normalny\")\n",
    "print(\"\\n\\n\\n\")\n",
    "\n",
    "#test istotnosci zmiennych oraz testy danych usuniętych\n",
    "stan_f = \"(Stan_Używany = Stan_Powystawowy = Stan_Po zwrocie )\"\n",
    "typ_f = \"(Typ telewizora_LED = Typ telewizora_OLED = Typ telewizora_QLED = Typ telewizora_inny = Typ telewizora_plazmowy )\"\n",
    "testy_f = [stan_f,typ_f]\n",
    "zmienne_ciagle = [\"Przekątna ekranu\",\"const\"]\n",
    "\n",
    "wynik_test = run_test(testy_f,zmienne_ciagle, final_model)\n",
    "wynik_pd = pd.DataFrame(wynik_test)\n",
    "for wyn in wynik_test:\n",
    "    print(wyn)\n",
    "print(\"Wszystkie zmienne są statystycznie istotne na podstawie metody wstecznej krokowej\")\n",
    "save_to_latex(wynik_pd.to_latex(),\"Testy_istotnosci\")\n",
    "print(\"\\n\\n\\n\")"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%% \n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.731\n",
      "Model:                            OLS   Adj. R-squared:                  0.655\n",
      "Method:                 Least Squares   F-statistic:                     9.591\n",
      "Date:                Sun, 05 Jul 2020   Prob (F-statistic):           2.50e-10\n",
      "Time:                        13:18:44   Log-Likelihood:                -29.570\n",
      "No. Observations:                  69   AIC:                             91.14\n",
      "Df Residuals:                      53   BIC:                             126.9\n",
      "Df Model:                          15                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "===========================================================================================\n",
      "                              coef    std err          t      P>|t|      [0.025      0.975]\n",
      "-------------------------------------------------------------------------------------------\n",
      "const                       6.1042      0.345     17.692      0.000       5.412       6.796\n",
      "Liczba złączy HDMI          0.2156      0.078      2.777      0.008       0.060       0.371\n",
      "Pobór mocy                  0.0024      0.001      3.068      0.003       0.001       0.004\n",
      "Marka_LG                    0.0967      0.253      0.382      0.704      -0.410       0.604\n",
      "Marka_Philips              -0.2888      0.297     -0.974      0.334      -0.884       0.306\n",
      "Marka_Samsung               0.0084      0.256      0.033      0.974      -0.504       0.521\n",
      "Marka_Sony                  0.0890      0.270      0.329      0.743      -0.453       0.631\n",
      "Marka_inna                 -0.0659      0.215     -0.306      0.761      -0.498       0.366\n",
      "Technologia 3D_True        -0.1506      0.282     -0.534      0.595      -0.716       0.415\n",
      "Technologia HDR_Tak         0.4236      0.225      1.882      0.065      -0.028       0.875\n",
      "Klasa efektywności_A+       0.3399      0.126      2.697      0.009       0.087       0.593\n",
      "Klasa efektywności_B        0.0591      0.443      0.134      0.894      -0.829       0.947\n",
      "Klasa efektywności_C        0.1037      0.532      0.195      0.846      -0.963       1.171\n",
      "Klasa efektywności_D    -5.141e-16   1.04e-15     -0.495      0.623    -2.6e-15    1.57e-15\n",
      "Klasa efektywności_inna     0.1939      0.204      0.948      0.347      -0.216       0.604\n",
      "jakosc_8K UHD                    0          0        nan        nan           0           0\n",
      "jakosc_Full HD             -0.3088      0.215     -1.438      0.156      -0.739       0.122\n",
      "jakosc_HD Ready            -0.4872      0.225     -2.162      0.035      -0.939      -0.035\n",
      "==============================================================================\n",
      "Omnibus:                        9.631   Durbin-Watson:                   2.108\n",
      "Prob(Omnibus):                  0.008   Jarque-Bera (JB):                9.395\n",
      "Skew:                          -0.777   Prob(JB):                      0.00912\n",
      "Kurtosis:                       3.924   Cond. No.                     3.54e+20\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The smallest eigenvalue is 7.46e-36. This might indicate that there are\n",
      "strong multicollinearity problems or that the design matrix is singular.\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "[('Marka', array(0.68461803)), ('Klasa efektywności', array(0.11816963)), ('jakosc', array(0.1064694)), ('Technologia HDR', array(0.0653743)), ('Technologia 3D', array(0.59543978)), ('Pobór mocy', 0.00339234003328616), ('Liczba złączy HDMI', 0.007571310892696465), ('const', 6.775218135671438e-24)]\n",
      "usuwana zmienna to Marka, z pvalue rowne 0.6846180285084702\n",
      "usuwana zmienna to Technologia 3D, z pvalue rowne 0.7432098756758226\n",
      "usuwana zmienna to Klasa efektywności, z pvalue rowne 0.14748468790117536\n",
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                      y   R-squared:                       0.625\n",
      "Model:                            OLS   Adj. R-squared:                  0.602\n",
      "Method:                 Least Squares   F-statistic:                     26.68\n",
      "Date:                Sun, 05 Jul 2020   Prob (F-statistic):           4.87e-13\n",
      "Time:                        13:18:44   Log-Likelihood:                -40.994\n",
      "No. Observations:                  69   AIC:                             91.99\n",
      "Df Residuals:                      64   BIC:                             103.2\n",
      "Df Model:                           4                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "=======================================================================================\n",
      "                          coef    std err          t      P>|t|      [0.025      0.975]\n",
      "---------------------------------------------------------------------------------------\n",
      "const                   6.9767      0.188     37.146      0.000       6.602       7.352\n",
      "Pobór mocy              0.0024      0.001      3.270      0.002       0.001       0.004\n",
      "Technologia HDR_Tak     0.4020      0.190      2.110      0.039       0.021       0.783\n",
      "jakosc_8K UHD        2.785e-15   3.29e-16      8.474      0.000    2.13e-15    3.44e-15\n",
      "jakosc_Full HD         -0.5283      0.193     -2.731      0.008      -0.915      -0.142\n",
      "jakosc_HD Ready        -0.8429      0.204     -4.140      0.000      -1.250      -0.436\n",
      "==============================================================================\n",
      "Omnibus:                        1.429   Durbin-Watson:                   2.077\n",
      "Prob(Omnibus):                  0.489   Jarque-Bera (JB):                1.448\n",
      "Skew:                          -0.277   Prob(JB):                        0.485\n",
      "Kurtosis:                       2.557   Cond. No.                     2.09e+19\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The smallest eigenvalue is 2.15e-33. This might indicate that there are\n",
      "strong multicollinearity problems or that the design matrix is singular.\n"
     ]
    }
   ],
   "source": [
    "#test zmeinnych odrzuconych \n",
    "df_initial = df_initial.drop(list(df2), axis = 1)\n",
    "Data_odrzucone = pd.get_dummies(df_initial, drop_first=True)\n",
    "X_2 = Data_odrzucone\n",
    "X_2 = sm.add_constant(X_2).astype(float)\n",
    "Y_2 = Y\n",
    "\n",
    "model_odrzucone = sm.OLS(Y_2,X_2).fit()\n",
    "print(model_odrzucone.summary())\n",
    "save_to_latex(model_odrzucone.summary(), \"model_odzrucony\")\n",
    "print(\"\\n\\n\\n\")\n",
    "marka_f = \"(Marka_LG=Marka_Philips=Marka_Samsung= Marka_Sony=Marka_inna =0)\"\n",
    "klasa_f = \"(Klasa efektywności_A+ =Klasa efektywności_B = Klasa efektywności_C = Klasa efektywności_D=Klasa efektywności_inna=0)\"\n",
    "jakosc_f = \"(jakosc_Full HD= jakosc_HD Ready = 0)\"\n",
    "hdr_f = \"(Technologia HDR_Tak = 0)\"\n",
    "tech3D_f = \"(Technologia 3D_True=0)\"\n",
    "testy_f = [marka_f,klasa_f,jakosc_f,hdr_f,tech3D_f]\n",
    "zmienne_ciagle = [\"Pobór mocy\",\"Liczba złączy HDMI\",\"const\"]\n",
    "testy_odrzucone = run_test(testy_f,zmienne_ciagle,model_odrzucone)\n",
    "print(testy_odrzucone)\n",
    "save_to_latex(pd.DataFrame(testy_odrzucone).to_latex(), \"p_value_odrzucone_sdtmodel\")\n",
    "\n",
    "while True:\n",
    "    najwieksza =  max(testy_odrzucone, key=itemgetter(1))\n",
    "    #gdy wszystkie zmeinne sa stsystcznie znaczace koniec metody krokowej\n",
    "    if najwieksza[1] < 0.1: break\n",
    "    print(f\"usuwana zmienna to {najwieksza[0]}, z pvalue rowne {najwieksza[1]}\")\n",
    "    df_initial = metoda_wsteczna_krokowa(df_initial,testy_odrzucone)\n",
    "    Data_odrzucone = pd.get_dummies(df_initial, drop_first=True)\n",
    "    X_Odrz = Data_odrzucone.iloc[:,1:]\n",
    "    X_Odrz = sm.add_constant(X_Odrz).astype(float)\n",
    "    model_odrzucone = sm.OLS(Y_2,X_Odrz).fit()\n",
    "    testy_odrzucone = run_test(testy_f,zmienne_ciagle,model_odrzucone)   \n",
    "print(model_odrzucone.summary())\n",
    "save_to_latex(model_odrzucone.summary(),\"model_odrzucony\")\n",
    "#model ten jest gorszy od modelu ze zmeinnymi z wyjsciowego modelu."
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.00\n",
      "Brak podstaw do odrzucenia H_0 na poziomie 0.1\n",
      " Model jest linowy \n",
      "\n",
      "\n",
      "\n",
      "\n",
      "<Wald test (chi2): statistic=[[7.99217329]], p-value=0.004697998878662177, df_denom=1>\n",
      "Brak podstaw do odrzucenia H_0 na poziomie 0.1\n",
      " Postać modelu jest poprawna \n",
      "\n",
      "\n",
      "\n",
      "\n",
      " P-value dla Breusch-Pagan =3.140388461885039e-07, F-value = 4.953662458242724e-12\n",
      "Brak podstaw do odrzucenia H_0 na poziomie 0.1\n",
      " Model jest homodekatyczny na podstawie testu Breusch-Pagan\n",
      "\n",
      "\n",
      "\n",
      "\n",
      " P-value dla White'a =2.2945588282568748e-05, F-value = 4.777562405746083e-10\n",
      "Brak podstaw do odrzucenia H_0 na poziomie 0.1\n",
      " Model jest homodekatyczny na podstawie testu White'a\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "[0.0004486101659010369, 0.06844771017783449, 0.012720313782877201, 0.8099242983366114, 0.5097953900923814, 0.0931610521242215, 2.5601245143718073e-06, 0.3061418350168913, 0.11853665177876153, 0.10629255437156347, 3.5115798186298576e-05, 1.4174463986392216e-06, 0.05413056982038244, 0.00023465570293964446, 0.000642512462765969, 0.02375313465555952, 0.03912067444613241, 1.1626454220688114e-05, 1.5451636967533333e-05, 0.4216993390487784]\n",
      "Model nie jest stabliny\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "[('const', 19.751096599853554), ('Przekątna ekranu', 1.2967734278112613), ('Stan_Po zwrocie', 1.0544684157460662), ('Stan_Powystawowy', 1.5948010145355067), ('Stan_Używany', 1.3879348950594765), ('Typ telewizora_LED', 3.511562041170873), ('Typ telewizora_OLED', 1.759289202476175), ('Typ telewizora_QLED', 3.1901534859977843), ('Typ telewizora_inny', 1.1848862798512998), ('Typ telewizora_plazmowy', 1.89202020130632)]\n",
      "Nie ma wpółlinowości między zmienymi(dla wszytkich zmeinnych vif < 5, ingorujemy const\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "['const']\n",
      "Nie ma wyskoiej współliniowiści, ingnorujemy const\n",
      "\n",
      "\n",
      "\n",
      "\n",
      "[{'variable': 'const', 'a': 5.699255350628206, 'r': nan}, {'variable': 'Typ telewizora_LED', 'a': 0.8278866585456797, 'r': -0.179241207762127}, {'variable': 'Typ telewizora_plazmowy', 'a': 0.5140352297157307, 'r': -0.01789614388591152}]\n"
     ]
    }
   ],
   "source": [
    "#test seri \n",
    "print(p_value_test_seri(final_model ,Y))\n",
    "print(\"Brak podstaw do odrzucenia H_0 na poziomie 0.1\\n Model jest linowy \")\n",
    "print(\"\\n\\n\\n\")\n",
    "\n",
    "#test RESET\n",
    "print(sm.stats.linear_reset(final_model,2))\n",
    "print(\"Brak podstaw do odrzucenia H_0 na poziomie 0.1\\n Postać modelu jest poprawna \")\n",
    "print(\"\\n\\n\\n\")\n",
    "\n",
    "#Testowanie heteroskeastyczności\n",
    "breush = sm.stats.diagnostic.het_breuschpagan(final_model.resid, X)\n",
    "lm_pvalue = breush[1]\n",
    "f_pvalue = breush[3]\n",
    "print(f\" P-value dla Breusch-Pagan ={lm_pvalue}, F-value = {f_pvalue}\")\n",
    "print(\"Brak podstaw do odrzucenia H_0 na poziomie 0.1\\n Model jest homodekatyczny na podstawie testu Breusch-Pagan\")\n",
    "print(\"\\n\\n\\n\")\n",
    "white = sm.stats.diagnostic.het_white(final_model.resid,X)\n",
    "lm_pvalue = white[1]\n",
    "#p-value for the f-statistic\n",
    "f_pvalue = white[3]\n",
    "print(f\" P-value dla White'a ={lm_pvalue}, F-value = {f_pvalue}\")\n",
    "print(\"Brak podstaw do odrzucenia H_0 na poziomie 0.1\\n Model jest homodekatyczny na podstawie testu White'a\")\n",
    "print(\"\\n\\n\\n\")\n",
    "#Test Chowa \n",
    "chow_list = []\n",
    "for i in range(20):\n",
    "    chow_list.append(chow_test(X,Y))\n",
    "chow_pd = pd.DataFrame(chow_list)\n",
    "save_to_latex(chow_pd.to_latex(), \"chow_final\")\n",
    "print(chow_list)\n",
    "print(\"Model nie jest stabliny\")\n",
    "print(\"\\n\\n\\n\")\n",
    "#Współlinowość \n",
    "vif_list = []\n",
    "for i in range(X.shape[1]):\n",
    "    vif = variance_inflation_factor(X,i)\n",
    "    vif_list.append((X.iloc[:,i].name,vif))\n",
    "vif_list_pd = pd.DataFrame(vif_list)\n",
    "#save to latex\n",
    "print(vif_list)\n",
    "print(\"Nie ma wpółlinowości między zmienymi(dla wszytkich zmeinnych vif < 5, ingorujemy const\")\n",
    "print(\"\\n\\n\\n\")\n",
    "coliniar = collinearity_test(X)\n",
    "coliniar_pd = pd.DataFrame(coliniar)\n",
    "\n",
    "print(coliniar)\n",
    "print(\"Nie ma wyskoiej współliniowiści, ingnorujemy const\")\n",
    "print(\"\\n\\n\\n\")\n",
    "#Koincydencja \n",
    "koincydencja = model_coincidence(X2,Y2)\n",
    "koincydencja_pd = pd.DataFrame(koincydencja)\n",
    "print(koincydencja)\n",
    "\n",
    "#Interpretacja Parametrów\n",
    "#w PDF "
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    Średnia wartosc        Dolna        Górna  Prawdziwa\n",
      "0       2477.346083  1174.160943  4944.613420     3299.0\n",
      "1       2477.346083  1174.160943  4944.613420     3299.0\n",
      "2        615.199526   378.295259  1513.229797      399.0\n",
      "3       1755.097975   980.177779  4175.716966     2178.0\n",
      "4        720.531276   410.408866  1636.248821      599.0\n",
      "..              ...          ...          ...        ...\n",
      "64       706.851938   354.604947  1423.254348     1549.0\n",
      "65       615.836034   265.391430  1243.152119      950.0\n",
      "66       952.694245   429.182524  1712.206178     1400.0\n",
      "67       988.969597   424.452393  1693.559656     1720.0\n",
      "68      1170.000553   472.174385  1883.748679     1899.0\n",
      "\n",
      "[69 rows x 4 columns]\n"
     ]
    }
   ],
   "source": [
    "#Predykcje z 95% przedziałem ufności \n",
    "#default alpha is 0.05\n",
    "prstd = final_model.get_prediction()\n",
    "prstd.summary_frame()\n",
    "mean = prstd.summary_frame().loc[:,\"mean\"]\n",
    "#interwal dla obserwacji \n",
    "lower = prstd.summary_frame().loc[:,\"obs_ci_lower\"]\n",
    "upper = prstd.summary_frame().loc[:,\"obs_ci_upper\"]\n",
    "\n",
    "mean = np.array(np.exp(mean)*np.sqrt(np.exp(final_model.resid))).reshape(-1,1)\n",
    "lower = np.array(np.exp(lower)).reshape(-1,1)#*np.sqrt(np.exp(final_model.resid))\n",
    "upper = np.array(np.exp(upper)).reshape(-1,1)#*np.sqrt(np.exp(final_model.resid))\n",
    "true =  np.array(Dane[\"Cena\"]).reshape(-1,1)\n",
    "Predictions = pd.DataFrame(np.concatenate([mean, lower, upper, true], axis=1), \n",
    "                           columns= ['Średnia wartosc','Dolna','Górna','Prawdziwa'])\n",
    "\n",
    "save_to_latex(Predictions.to_latex(), \"Tabela Predykcji\")\n",
    "print(Predictions)"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "outputs": [
    {
     "data": {
      "text/plain": "<function matplotlib.pyplot.show(*args, **kw)>"
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": "<Figure size 1080x1080 with 1 Axes>",
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3oAAANOCAYAAABHuH7DAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdfaxk530f9u+ze5fcXVqiKIqSVVJ8i5U0kpyiMWOoLdqmYVsrqG0ZgQ0wYC2hVUDUMNyiRZpaIBIjQFjYbQK7RiE1TORacgjLgpLUcgO7NiSkRgHFDp3UpSVbNWuFFC1aorgyLWmXu/fuPv3jzPHO3p1777yel5nPB7iYueeemT338M7wfOf3e56n1FoDAADA9jjV9wEAAACwXoIeAADAlhH0AAAAtoygBwAAsGUEPQAAgC2z1/cBLOsNb3hDvf/++/s+DAAAgF78xm/8xldqrXfN+tlog97999+fp59+uu/DAAAA6EUp5bmjfqZ1EwAAYMsIegAAAFtG0AMAANgygh4AAMCWOTHolVJ+qpTy5VLKbx3a/kOllM+VUj5TSvkfpra/v5Ty7ORn3zG1/dtKKc9MfvaTpZQy2X5rKeXnJtt/rZRy//p+PQAAgN0zT0Xvp5O8a3pDKeU/SPLuJH+m1vr2JH97sv1tSR5J8vbJYz5QSjk9edgHkzyW5K2Tr/Y535fkq7XWb0ny40l+bIXfBwAAYOedGPRqrb+a5MKhzT+Q5EdrrZcn+3x5sv3dST5aa71ca/18kmeTfHsp5c1JXltr/XSttSb5SJLvmXrMhyf3P57k4bbaBwAAwOKWHaP3J5P8u5NWy/+zlPLnJtvvTvKFqf1emGy7e3L/8PYbHlNrPUjySpI7Z/2jpZTHSilPl1Kefumll5Y8dAAAgO22bNDbS3JHkncm+W+TfGxShZtViavHbM8JP7txY61P1lofqrU+dNddMxeABwAA2HnLBr0Xkvyj2vj1JNeSvGGy/S1T+92T5IuT7ffM2J7px5RS9pLcnptbRQEAAJjTskHvf0vyF5KklPInk9yS5CtJPpHkkclMmg+kmXTl12utLyb5WinlnZPK33uS/PzkuT6R5L2T+9+b5FOTcXwAAAAsYe+kHUopP5vkzyd5QynlhSQ/kuSnkvzUZMmFK0neOwlnnymlfCzJZ5McJPnBWuvVyVP9QJoZPM8l+cXJV5J8KMnPlFKeTVPJe2Q9vxoAAMBuKmMtnj300EP16aef7vswAAAAelFK+Y1a60OzfrZs6yYAAAADJegBAABsGUEPAABgywh6AAAAW0bQAwAA2DKCHgAAwJYR9AAAALaMoAcAALBlBD0AAIAtI+gBAABsGUEPAABgywh6AAAAW0bQAwAA2DKCHgAAwJYR9AAAALaMoAcAALBlBD0AAIAtI+gBAABsGUEPAABgywh6AAAAW0bQAwAA2DKCHsAA/P7vJ6+80vdRAADbQtADGIC/+BeTv/7X+z4KAGBbCHoAA/CVryQXLvR9FADAthD0AAZgfz85OOj7KACAbSHoAQzA/n7zBQCwDoIewACo6AEA6yToAQyAoAcArJOgB9CzWrVuAgDrJegB9Ozq1eZWRQ8AWBdBD6BnbcAT9ACAdRH0AHrWtmwKegDAugh6AD1rg54xegDAugh6AD1T0QMA1k3QA+iZoAcArJugB9AzrZsAwLoJegA9U9EDANZN0APomaAHAKyboAfQM62bAMC6CXoAPVPRAwDWTdAD6Fkb8AQ9AGBdBD2AnqnoAQDrJugB9MwYPQBg3QQ9gJ6p6AEA6yboAfRM0AMA1k3QA+hZG/RqTa5e7fdYAIDtIOgB9Gx6bJ6qHgCwDoIeQM8EPQBg3QQ9gJ4JegDAugl6AD2bDnqWWAAA1kHQA+jZdBVPRQ8AWAdBD6BnWjcBgHUT9AB6pnUTAFg3QQ+gZyp6AMC6CXoAPRP0AIB1E/QAeiboAQDrJugB9MwYPQBg3QQ9gJ6p6AEA6yboAfRM0AMA1k3QA+iZ1k0AYN0EPYCeTVfxVPQAgHUQ9AB6pnUTAFg3QQ+gZ1o3AYB1E/QAeqaiBwCsm6AH0LP9/aSU5r6gBwCsg6AH0LP9/eTcuea+oAcArIOgB9Cz6aBnjB4AsA6CHkDP9veT8+eb+yp6AMA6CHoAPdO6CQCsm6AH0DOtmwDAugl6AD1T0QMA1k3QA+jZwYExegDAegl6AD1T0QMA1k3QA+jZ/n5y9uz1+wAAqxL0AHqmogcArJugB9Cz/f3k1luTU6cEPQBgPQQ9gJ7t7ydnziR7e1o3AYD1EPQAejYd9FT0AIB1EPQAetYGvTNnBD0AYD0EPYCeqegBAOsm6AH0zBg9AGDdBD2AHl27ltSqogcArJegB9CjtoJnjB4AsE6CHkCP2qC3t6d1EwBYH0EPoEfTFT2tmwDAugh6AD3SugkAbIKgB9AjFT0AYBNODHqllJ8qpXy5lPJbM372V0sptZTyhqlt7y+lPFtK+Vwp5Tumtn9bKeWZyc9+spRSJttvLaX83GT7r5VS7l/PrwYwfIeDnjF6AMA6zFPR++kk7zq8sZTyliT/UZLnp7a9LckjSd4+ecwHSimnJz/+YJLHkrx18tU+5/uSfLXW+i1JfjzJjy3ziwCMkYoeALAJJwa9WuuvJrkw40c/nuSvJalT296d5KO11su11s8neTbJt5dS3pzktbXWT9daa5KPJPmeqcd8eHL/40kebqt9ANvOGD0AYBOWGqNXSvnuJL9fa/3NQz+6O8kXpr5/YbLt7sn9w9tveEyt9SDJK0nuXOa4AMZG6yYAsAl7iz6glHI+yeNJ/uNZP56xrR6z/bjHzPq3H0vT/pl77733xGMFGLq2gtcGvW98o9/jAQC2wzIVvT+R5IEkv1lK+VdJ7knyL0op35ymUveWqX3vSfLFyfZ7ZmzP9GNKKXtJbs/sVtHUWp+stT5Ua33orrvuWuLQAYZF6yYAsAkLB71a6zO11jfWWu+vtd6fJqj92VrrHyT5RJJHJjNpPpBm0pVfr7W+mORrpZR3TsbfvSfJz0+e8hNJ3ju5/71JPjUZxwew9dqgt7endRMAWJ95llf42SSfTvKnSikvlFLed9S+tdbPJPlYks8m+aUkP1hrvTr58Q8k+ftpJmj5/5L84mT7h5LcWUp5Nsl/k+SHl/xdAEbHrJsAwCacOEav1vqXT/j5/Ye+fyLJEzP2ezrJO2ZsfzXJ9510HADbSNADADZhqVk3AVgPY/QAgE0Q9AB6ZHkFAGATBD2AHmndBAA2QdAD6JHWTQBgEwQ9gB5p3QQANkHQA+hRW8HTugkArJOgB9AjrZsAwCYIegA9aoPe3p6KHgCwPoIeQI9mzbpZa7/HBACMn6AH0KPDQS9Jrl7t73gAgO0g6AH06PAYvUT7JgCwOkEPoEdt0Dt9+npFzxILAMCqBD2AHu3vN5W8Uq4HPRU9AGBVgh5Aj9qgl2jdBADWR9AD6NF00FPRAwDWRdAD6NHBwc1Bzxg9AGBVgh5Aj1T0AIBNEPQAerS/fz3gGaMHAKyLoAfQo1kVPa2bAMCqBD2AHmndBAA2QdAD6JHlFQCATRD0AHqkdRMA2ARBD6BHWjcBgE0Q9AB6JOgBAJsg6AH0yBg9AGATBD2AHhmjBwBsgqAH0KODA62bAMD6CXoAPdrfvx7wtG4CAOsi6AH0SOsmALAJgh5Aj8y6CQBsgqAH0CNBDwDYBEEPoEeWVwAANkHQA+iRMXoAwCYIegA90roJAGyCoAfQI62bAMAmCHoAPdK6CQBsgqAH0KODA62bAMD6CXoAPak1uXpV6yYAsH6CHkBP2hbNtpJ3+nRzK+gBAKsS9AB60ga9tpJXShP2jNEDAFYl6AH05HDQS5rqnooeALAqQQ+gJ7OC3pkzgh4AsDpBD6AnR1X0tG4CAKsS9AB6onUTANgUQQ+gJ1o3AYBNEfQAeqJ1EwDYFEEPoCdt5U7rJgCwboIeQE+M0QMANkXQA+hJG/T29q5vM0YPAFgHQQ+gJ8boAQCbIugB9ETrJgCwKYIeQE8srwAAbIqgB9ATrZsAwKYIegA90boJAGyKoAfQE0EPANgUQQ+gJ8boAQCbIugB9KQNdMboAQDrJugB9ETrJgCwKYIeQE/aoLe3d32b1k0AYB0EPYCeWF4BANgUQQ+gJ1o3AYBNEfQAeiLoAQCbIugB9MTyCgDApgh6AD2ZNRmLMXoAwDoIegA92d9vgl0p17dp3QQA1kHQA+jJ/v6NbZuJ1k0AYD0EPYCezAp6WjcBgHUQ9AB6cnAwO+ip6AEAqxL0AHrSjtGbduZMcu1a8wUAsCxBD6AnR7VuJsnVq90fDwCwPQQ9gJ4cF/SM0wMAViHoAfTkuKBnnB4AsApBD6AnRy2vkAh6AMBqBD2AnmjdBAA2RdAD6InWTQBgUwQ9gJ5o3QQANkXQA+iJ1k0AYFMEPYCeHBxo3QQANkPQA+jJ/v71YNcS9ACAdRD0AHpijB4AsCmCHkBPjNEDADZF0APoieUVAIBNEfQAeqJ1EwDYFEEPoCdaNwGATRH0AHqidRMA2BRBD6Angh4AsCmCHkBPjNEDADZF0APoycGBMXoAwGacGPRKKT9VSvlyKeW3prb9j6WU3yml/D+llH9cSnnd1M/eX0p5tpTyuVLKd0xt/7ZSyjOTn/1kKaVMtt9aSvm5yfZfK6Xcv95fEWCY9vevB7uW1k0AYB3mqej9dJJ3Hdr2K0neUWv9M0n+3yTvT5JSytuSPJLk7ZPHfKCUcnrymA8meSzJWydf7XO+L8lXa63fkuTHk/zYsr8MwFjUOruip3UTAFiHE4NerfVXk1w4tO2Xa63tZcg/S3LP5P67k3y01nq51vr5JM8m+fZSypuTvLbW+ulaa03ykSTfM/WYD0/ufzzJw221D2BbtUFO6yYAsAnrGKP3nyf5xcn9u5N8YepnL0y23T25f3j7DY+ZhMdXktw56x8qpTxWSnm6lPL0Sy+9tIZDB+hHG+TMugkAbMJKQa+U8niSgyRPtZtm7FaP2X7cY27eWOuTtdaHaq0P3XXXXYseLsBgCHoAwCYtHfRKKe9N8p1JHp20YyZNpe4tU7vdk+SLk+33zNh+w2NKKXtJbs+hVlGAbXNU0DNGDwBYh6WCXinlXUn+uyTfXWu9OPWjTyR5ZDKT5gNpJl359Vrri0m+Vkp552T83XuS/PzUY947uf+9ST41FRwBttJJFT1j9ACAVeydtEMp5WeT/PkkbyilvJDkR9LMsnlrkl+ZzJvyz2qt/0Wt9TOllI8l+Wyals4frLVenTzVD6SZwfNcmjF97bi+DyX5mVLKs2kqeY+s51cDGC6tmwDAJp0Y9Gqtf3nG5g8ds/8TSZ6Ysf3pJO+Ysf3VJN930nEAbBOtmwDAJq1j1k0AFmR5BQBgkwQ9gB60QW7vUF/Fqcm7sooeALAKQQ+gB0e1bpbSbBP0AIBVCHoAPTgq6CVNlU/rJgCwCkEPoAcnBT0VPQBgFYIeQA8EPQBgkwQ9gB4cF/SM0QMAViXoAfTAGD0AYJMEPYAeaN0EADZJ0APogdZNAGCTBD2AHmjdBAA2SdAD6EFbsdO6CQBsgqAH0IO2Yre3d/PPBD0AYFWCHkAPjNEDADZJ0APogTF6AMAmCXoAPbC8AgCwSYIeQA+0bgIAmyToAfRA6yYAsEmCHkAPtG4CAJsk6AH0YH8/OX06KeXmnwl6AMCqBD2AHuzvz67mJcboAQCrE/QAenBwcHTQM0YPAFiVoAfQg/39JtDNonUTAFiVoAfQA62bAMAmCXoAPTgu6GndBABWJegB9OCkoKeiBwCsQtAD6IGgBwBskqAH0IOTxuhp3QQAViHoAfRARQ8A2CRBD6AHgh4AsEmCHkAPLK8AAGySoAfQg4OD4yt6tSZXr3Z7TADA9hD0AHqwv98Eulna7ap6AMCyBD2AHpzUupkIegDA8gQ9gB6cNBlLuw8AwDIEPYAezBP0VPQAgGUJegA9EPQAgE0S9AB6YIweALBJgh5AD4zRAwA2SdAD6IHWTQBgkwQ9gB5o3QQANknQA+jBwYHWTQBgcwQ9gB7s718PdIdp3QQAViXoAfTAGD0AYJMEPYCO1WqMHgCwWYIeQMeuXm1ujdEDADZF0APoWBvgtG4CAJsi6AF07KSgp3UTAFiVoAfQsXkrelo3AYBlCXoAHdO6CQBsmqAH0DFBDwDYNEEPoGNtgDNGDwDYFEEPoGNtRa+t3B1mjB4AsCpBD6BjWjcBgE0T9AA6ZnkFAGDTBD2AjlleAQDYNEEPoGNaNwGATRP0ADom6AEAmyboAXRs3jF6WjcBgGUJegAdU9EDADZN0APomKAHAGyaoAfQsZOC3qlTzZegBwAsS9AD6Fgb4NrK3Sx7e8boAQDLE/QAOnZSRS9pgp6KHgCwLEEPoGPzBL0zZwQ9AGB5gh5Ax+at6GndBACWJegBdEzrJgCwaYIeQMcEPQBg0wQ9gI4ZowcAbJqgB9AxY/QAgE0T9AA6pnUTANg0QQ+gY/v7yalTzddRtG4CAKsQ9AA6dnBwfDUv0boJAKxG0APo2P5+E+SOo3UTAFiFoAfQsf39+Sp6gh4AsCxBD6Bj8wS9IY/R+53fSb7/+7WWAsCQCXoAHZu3ojfUIPWpTyX/4B8kL7zQ95EAAEcR9AA6NvbWzStXmttvfKPf4wAAjiboAXRs7K2bly83txcv9nscAMDRBD2Ajo29dbMNeip6ADBcgh5Ax7RuAgCbJugBdGzsQU9FDwCGT9AD6NjBwXxj9IbeummMHgAMl6AH0LH9/aZid5whV/S0bgLA8Al6AB3TugkAbJqgB9AxyysAAJsm6AF0bOzLK2jdBIDhE/QAOqZ1EwDYtBODXinlp0opXy6l/NbUtteXUn6llPK7k9s7pn72/lLKs6WUz5VSvmNq+7eVUp6Z/OwnSyllsv3WUsrPTbb/Winl/vX+igDDonUTANi0eSp6P53kXYe2/XCST9Za35rkk5PvU0p5W5JHkrx98pgPlFJOTx7zwSSPJXnr5Kt9zvcl+Wqt9VuS/HiSH1v2lwEYA62bAMCmnRj0aq2/muTCoc3vTvLhyf0PJ/meqe0frbVerrV+PsmzSb69lPLmJK+ttX661lqTfOTQY9rn+niSh9tqH8A2mjfoXb2a1NrNMS1C6yYADN+yY/TeVGt9MUkmt2+cbL87yRem9nthsu3uyf3D2294TK31IMkrSe6c9Y+WUh4rpTxdSnn6pZdeWvLQAfo1b9BLmrA3NIIeAAzfuidjmVWJq8dsP+4xN2+s9cla60O11ofuuuuuJQ8RoF8HB/ON0Wv3HZq2ddMYPQAYrmWD3pcm7ZiZ3H55sv2FJG+Z2u+eJF+cbL9nxvYbHlNK2Utye25uFQXYGvv71yt2R2l/PsRxeip6ADB8ywa9TyR57+T+e5P8/NT2RyYzaT6QZtKVX5+0d36tlPLOyfi79xx6TPtc35vkU5NxfABbaZHWzSFW9AQ9ABi+Ez5TTkopP5vkzyd5QynlhSQ/kuRHk3yslPK+JM8n+b4kqbV+ppTysSSfTXKQ5Adrre0Ikx9IM4PnuSS/OPlKkg8l+ZlSyrNpKnmPrOU3AxioeZdXSIYZ9LRuAsDwnRj0aq1/+YgfPXzE/k8keWLG9qeTvGPG9lczCYoAu2CRip7WTQBgGeuejAWAY7RLJmxD6+arrw5zVlAAQNAD6FRboRtz0LtyJbnllub+pUv9HgsAMJugB9CheYPeUMfoXb3afN1xR/O99k0AGCZBD6BDi1b0hjZGr23bfP3rm1tBDwCGSdAD6NDYWzfbGTcFPQAYNkEPoENjb91sK3pt66YlFgBgmAQ9gA61wW3vhMVtht66aYweAAyboAfQIa2bAEAXBD2ADo096GndBIBxEPQAOrRtY/RU9ABgmAQ9gA6NfXkFrZsAMA6CHkCHtq11U9ADgGES9AA6tC2tm7ff3twaowcAwyToAXRoW1o3z55Nzp9X0QOAoRL0ADq0La2bt96a3HaboAcAQyXoAXRoW4LeLbc0QU/rJgAMk6AH0KE2uLVB7ihtEBxq6+att2rdBIAhE/QAOrQtFT2tmwAwbIIeQIe2Jehp3QSAYRP0ADo09uUVtG4CwDgIegAdGvvyClo3AWAcBD2ADm1L6+benqAHAEMm6AF0aBtaN2+9NSnFGD0AGDJBD6BD8wa906dv3H8oLl9ugl5ijB4ADJmgB9CheYNeKU3YG1pF7/LlZsbNpKnoXb6cXL3a7zEBADcT9AA6NG/QS5pxcEMLem3rZtIEvUT7JgAMkaAH0KE2uJ2a4933zJnhBb3DrZuJ9k0AGCJBD6BD+/tNgCvl5H339oY5Rm+6dTMR9ABgiAQ9gA61QW8eWjcBgGUJegAdWiToDb11U0UPAIZL0APo0KIVvSG3bhqjBwDDJegBdGgbWzcFPQAYHkEPoENjD3qzWjeN0QOA4RH0ADq0DWP0tG4CwPAJegAdGvsYPa2bADAOgh5Ah7RuAgBdEPQAOnRw0AS4eQy9dfPs2WbhdxU9ABgeQQ+gQ9vUullKM05P0AOA4RH0ADq0Ta2bSdO+KegBwPAIegAd2oag17ZuJk3QM0YPAIZH0APo0KLLKwypdbPW5nimK3paNwFgmAQ9gA6NuaJ35Upzq3UTAIZP0APo0JiD3uXLza3WTQAYPkEPoEOLtm4OKejNquhp3QSAYRL0ADo05uUV2oqe1k0AGD5BD6BDWjcBgC4IegAdGnPQMxkLAIyHoAfQoYODJsDNY2jLK8xq3TRGDwCGSdAD6NCYK3pHtW5euTKs4wQABD2ATo056B3VupkYpwcAQyPoAXRozMsrHNW6mWjfBIChEfQAOrQNyyscbt1MBD0AGBpBD6Aj1641X1o3AYBNE/QAOtJW5xZp3WzD4RActWB6oqIHAEMj6AF0ZNGg1y7DMJSq3qzWTWP0AGCYBD2Ajow96GndBIDxEPQAOjL2oKd1EwDGQ9AD6Egb2NoAd5I2EA4t6GndBIDhE/QAOrJsRW8oSywc17op6AHAsAh6AB3ZltbNWevoGaMHAMMi6AF0ZJnlFZJhBb29veTU1P85br21+V5FDwCGRdAD6Mg2tG5Ot20mSSnNOD1BDwCGRdAD6Mg2tG4eDnpJ076pdRMAhkXQA+jINgS96fF5rdtuU9EDgKER9AA6suwYvSG3biZaNwFgiAQ9gI5sQ0XvqNZNQQ8AhkXQA+jINgS9o1o3jdEDgGER9AA60ga2NsCdZGjLK2jdBIDxEPQAOjL25RW0bgLAeAh6AB3RugkAdEXQA+jINgQ9FT0AGAdBD6AjllcAALoi6AF0ZBsqeke1bu7vDyeQAgCCHkBntiHoHdW6mRinBwBDIugBdGTZ1s2hBL3jWjcT7ZsAMCSCHkBHtmF5haNaNxNBDwCGRNAD6IjWTQCgK4IeQEfawNYGuJMMLegd1bqpogcAwyPoAXSkrejNG/SGtLxCrUe3bhqjBwDDI+gBdGR/vwl5pcy3/5AqegcHTdjTugkA4yDoAXRkf3/+8XnJsILelSvNrdZNABgHQQ+gI4sGvSEtr3D5cnOrdRMAxkHQA+jIshW9IYzRa4Oe1k0AGAdBD6Ajiwa9U5N36CFU9LRuAsC4CHoAHVk06JXS7D+EoHdc6+YttySnTwt6ADAkgh5ARxYNeknTvjn01s1SmnF6gh4ADIegB9CRZYPeECp6x7VuJk37pjF6ADAcgh5ARw4O5l8svTWUoHdc62bSBD0VPQAYjpWCXinlvy6lfKaU8lullJ8tpZwtpby+lPIrpZTfndzeMbX/+0spz5ZSPldK+Y6p7d9WSnlm8rOfLGXe5YQBxmOZit7QxugdVdHTugkAw7J00Cul3J3kv0zyUK31HUlOJ3kkyQ8n+WSt9a1JPjn5PqWUt01+/vYk70rygVLK6cnTfTDJY0neOvl617LHBTBUYx6jp3UTAMZl1dbNvSTnSil7Sc4n+WKSdyf58OTnH07yPZP7707y0Vrr5Vrr55M8m+TbSylvTvLaWuuna601yUemHgOwNcY8Rk/rJgCMy9JBr9b6+0n+dpLnk7yY5JVa6y8neVOt9cXJPi8meePkIXcn+cLUU7ww2Xb35P7h7TcppTxWSnm6lPL0Sy+9tOyhA/RC6yYA0JVVWjfvSFOleyDJv5bktlLKf3rcQ2Zsq8dsv3ljrU/WWh+qtT501113LXrIAL3a9tZNQQ8AhmOV1s3/MMnna60v1Vr3k/yjJP92ki9N2jEzuf3yZP8Xkrxl6vH3pGn1fGFy//B2gK2y7a2bxugBwHCsEvSeT/LOUsr5ySyZDyf57SSfSPLeyT7vTfLzk/ufSPJIKeXWUsoDaSZd+fVJe+fXSinvnDzPe6YeA7A1tiHoqegBwDgsuKLTdbXWXyulfDzJv0hykORfJnkyyTcl+Vgp5X1pwuD3Tfb/TCnlY0k+O9n/B2utVydP9wNJfjrJuSS/OPkC2CrLjtEbQ+umMXoAMCxLB70kqbX+SJIfObT5cprq3qz9n0jyxIztTyd5xyrHAjB021DRO6518+Bgud8RAFi/VZdXAGBO+/tNcFvE0ILeca2biaoeAAyFoAfQkYOD8S6vcOVKcurU0UH1/PnmVtADgGEQ9AA6MublFS5fPrptM7le0TPzJgAMg6AH0JGxj9E7qm0z0boJAEMj6AF0ZMxB78oVQQ8AxkTQA+jImJdXOKl10xg9ABgWQQ+gI2Ou6M3bummMHgAMg6AH0IFak6tXxxv0tG4CwLgIegAdaNsvx7q8wphbN3/oh5KPf7zvoxiPv/E3kr/39/o+CgBWJegBdGDZoDek5RXG2rr5kY8kv/RLfR/FeHz0o8knPtH3UQCwKkEPoANtWDtqwfGjaN1c3aVLzRfzcb4AtsOClxywu/7xP06+8IX59j13Lvn+70/Ont3sMTEebVgb6xi9y5evh7lZbrmlOdahBb2DgyZkD7HSOFQXLzpfANtA0IM5vPJK8pf+0mKPedObku/+7s0cD+Ozyhi9obRuvv71x+9z/vzwgl5bmVKhmp+gB7AdBD2Yw9e+1tz++I8n73nP8fv+3u8lf+7PXX8MJKuN0RtCRe+k1s2kqfgNLSC0xzO04zB/OGQAACAASURBVBqqa9eSV18VjAG2gaAHc2gvEu+66+SqhgtLZhl70Dtp1s2kCXoqeuP26qvNrfcvgPEzGQu9+ZmfSR5+uO+jmE970dNOIX+cdh8XSkzbhuUVTqroDbF10wcvi3G+ALaHoEdv/vk/T/7pP20Wkh46QY9VrVLRaxdb79NYWzdV9BbT/vdzvgDGT9CjNxcvNuNBrlzp+0hOtkjQu/XWpJThXfDSr1WCXtJ/VW+srZsqVIuZDsZj+BAOgKMJevRmTBdgiwS9Upr9fCLOtFVaN5NhBL0xtm6q6C1m+v24Ha8HwDgJevSmvfDatqCXNOvojeH3ojurVvT6XmJhrK2bY/pAaQimz5NzBjBugh69GdMF2KJB7/z5cfxedGfMrZvXrjX//phbN9uF0znedOXTexjAuAl69EbQY5e0QW1vwUVthhD0Ll9ubuep6A0t6E0HF+2bJ5t+33K+AMZN0KM329y6aYweh606Rq/PalQ7YdK8Y/SGNImH4LIYrZsA20PQozdjrOidOzff/sbocdiYWzfbit48rZtXrw6rRVIr4mKcL4DtIejRm7EFvbNnk1NzvmK0bnLYNgS9eVo3k2G1b6roLcb5Atgegh69GVvQm7dtMxH0uNmYl1dYpHUzGVbQU6FajNZNgO0h6NGbMa1vJeixqjEvr7BI62YyrL99FarFmLwGYHsIevSi1u2v6LlIYprWzX6o6C1GRQ9gewh69OLKlWZtrmQcFxOLBj2TsXDYmIPevK2bQwx6KnqLEfQAtoegRy/GdjGhdZNVjXl5hXlbN9vXyJD+9i9dun7cgt7JLl5MXvOa5r7zBTBugh69GFs71TJB7/LlZqp5SMZd0Rtz6+bFi8mdd16/z/EuXUpe//rmvvMFMG6CHr3YhYpe4hNxrmuDWhvc5jWEoDfm1s1Ll64HPa/Hk128mNx+e/N3N4b3ZgCOJujRi20Peu3C6i4saY15eYVFWzeHFPRU9BZz8WLz/mVCKYDxE/ToxbYHvSGOVaJf27C8wrwVvSH93avoLebSpeb9yzhjgPET9OjFLozRax8HSRPUTp9OSlnscVo3V3PxYvJN39RUI70eT9a+1507JxgDjJ2gRy/aC67Tp8dx8SXosar9/cWreckwgt68rZtnzjRfQwp6ly5pRVzEdOum9y+AcRP06EV7AXHnncO/mLh6tbnQNRkLq1g26A1peYWTKnrJ8ALCdIVqSMc1VFo3AbaHoEcv2gD0hjcM/2KiPdZlJmMZ+u9Gd8Zc0Zu3dTNp2jdV9MZL6ybA9hD06EUbgMYQ9Nrj07rJKsYc9OZt3UyGFfT295vzpqI3P62bANtD0KMXY2rdFPRYh1VbN8cS9M6fH07QaytSKnrzqfV666ZgDDB+gh69UNFj16xa0etzjN6VK03Im2fG0NtuG87f/fRrV3A52f5+Mya5HaMnGAOMm6BHLy5dai5gb799+BdfywQ9C6Zz2P7+9dC2iKG0bs5TzUuG1bqporeY9r1O6ybAdhD06MX0OJBLl5qWoaFS0WMdDg7GPUZvnolYkmEFPRW9xRw+X4IxwLgJevSindmtDUSvvtrv8RxnmaB39uyNj4UxL69w5cr8QW9IlaDDFSrB5XjTMwwP6b8jAMsR9OjF4aA35AuKZYLeqVMqCNxo7LNujrl1U0VvPtPvdefPX5+1FIBxEvToxfSivMmwL8CWCXrt/ioItMYe9MbcuqmiN5/p82WcMcD4CXr0YnqMXvv9UC0b9FQQmLZs0Dt1qvkaW+vmEMbdqugt5nDrZuKcAYyZoEcvtr11s91/yL8X3Vo26CVNVa/vit4irZtXrzbhsG+HK3oHB/0G5qE7PBnL9DYAxkfQoxeCHrtm7EFvkdbNZBjtm4cretPbuNnhYJw4XwBjJujRi7GN0Tt1av6KRkvQY9oqQe/MmX6D3qKtm8kwgp7gshitmwDbRdCjF+0YvTG0B7XVx1IWe5zJH5i2akWvz5bDRVs3k2G8pmdV9IZwXEM1q3XTexjAeAl69GJsrZuLtm0mJn/gRvv712fQXJTWzeW0r7+zZ1X05jGrAuo9DGC8BD16MbbWzWWCntZNph0cjHeM3iKtm0MKepcuNaGlFBW9eWjdBNgugh692IWKnqDHtFXH6I2ldXNIr+m2RTxR0ZvHxYvJ6dPN35vWTYDxE/To3P5+8zWW9qBVgp6LJFpm3exe2zmQqOjNow3GpYzjvRmA4wl6dG66Pejs2eb+kC8mjNFjHcYc9Mbauqmit5jpYCzoAYyfoEfnpoPeqVPDD0SrVvSuXVv/MTE+Y15eYaytmyp6i5l+r9O6CTB+gh6dm57ZLRn+WLZVgl6SvPrqeo+HcRr78gpjregdrlAJLkebPl9j6LYA4HiCHp2bXqupvR3yxdeqQc+FEsl4WzdrHW/rZjvrZqKiN4/pVtdTp5qw53wBjJegR+dmBb0hX0ysGvSGHGLpRq3jXV6hrSTO27q5t9fsO4Sgp6K3mOlW12T4H8IBcDxBj85Nj9Frb7cx6Kkg0GpD2hiXV7h8ubmdt6KXDOc1raK3mMPvdUP57wjAcgQ9OjemMXq1at1kdW1I29tb7vF9VvSuXGluFwl6t902vIreqVPN76BCdbTp1s2kue98AYyXoEfnxtS6ub+fXL0q6LGaVSt6fQa9tqI3b+tmMpygN13RSwSXk8xq3fT+BTBegh6dG1PQO3ysizAmiFZb0Rvj8gpjbt3UirgY5wtguwh6dG5MY/RWCXrGBNFaNej1ubzCWFs3a1XRW5TWTYDtIujRuTGN0VtHRW+ovxvdWUfQ67uiN7bWzVlt10N+rxkCrZsA20XQo3OHw9O5c8O9mBD0WIdtCHpjq+i1lSgVqvlcvdr8t55+rxvyezMAJxP06NylS0kp1y8ch/ypsaDHOqxjjN6YWjeH8Jqe9dodwnEN1axgbB09gHET9OhcO+C/lOb78+ebT5KvXu33uGYxGQvrsA0VvbG1bh5uEW/vez3OdnjsdHtfMAYYL0GPzh0e8D/kQGQyFtZhG4LeWFs3BZf5zHqvE4wBxk3Qo3OzpvButw/NKkGvXaB5iL8X3Rrz8gqrtG7WupljmoeK3mJmna8h/HcEYHmCHp2bNbNbMsxAtErQax83xN+LbrVBb29vucf3ubzCsq2b165df2wfVPQWc9T5qrXf/44ALE/Qo3NHtW4O8QJsHUFPBYG2GrdLrZtJv+2bKnqLOap1M3HOAMZK0KNzR7VuDvFiQkWPddjF1s2k3799Fb3FHDVL6fTPABgXQY/OjXGM3nRVYBHWoSJZz2QsY2vdTIZZ0Ts46O9cDtlR52v6ZwCMi6BH58Y2Ru+WW5YfW6WCQLKeoHf1aj+TYoy1dfOoit70z7jO+QLYPoIenRvbGL1l2zYTQY/GOoJe0s9ak8u0bg4h6B1XoRJcbqZ1E2D7CHp0bmytm6sGPReVrGOM3vTzdGmZ1s0hvKaPq1AN8b2mb4IxwPYR9OjcLgU9Y/RI1lfR62NClsuXk9Onm695DaWiV8qNlUjB5WiCMcD2EfTo3NjG6GndZFVjDnpXrizWtpkMI+hdutQEu1Kubxvye03f2nNy9uz1bc4XwLgJenTq2rXk1VeN0WO3rKt1s6+K3iJtm8kwXtOHxwInKnrHuXixCXmnpq4KnC+AcRP06NSs9qBbbmkuLoYYiAQ91qENesvO3to+rq8xemOt6B1+7Q4hgA6V8wWwfQQ9OjUr6JUy3LFs65qMpY9p8RkOrZvdm/XaVaE62nHna4jvzQCcTNCjU7Om8G6/H+LFxDomY6n1+syF7KY2oI0x6C3Tunn6dBMO+55186jWzSG+1/RtVqurdfQAxk3Qo1OzpvBOtjfoaX0iGf/yCotW9JLmb39oFT3B5WizWjfPnGlCu/cvgHES9OjUrlX0BD2ScbduLhv0brut/zF6Knrzm/Ve17bVC8YA4yTo0alZY/Ta74d48bWuoOdCabft7zcTDp1a8h237zF6i7ZuJv0HPRW9xcxq3UyG+94MwMkEPTo1ptbNa9dmtzMtQgWBpAl6y1bzkv6XV1i2ddMYvfE46kOtvv87ArC8lYJeKeV1pZSPl1J+p5Ty26WUf6uU8vpSyq+UUn53cnvH1P7vL6U8W0r5XCnlO6a2f1sp5ZnJz36ylOklbtkmY2rdfPXV5lbrJqtaNeiNbXmFZJgVvVOnmt9FRe9mR32opXUTYLxWrej9T0l+qdb6ryf5N5L8dpIfTvLJWutbk3xy8n1KKW9L8kiStyd5V5IPlFJOT57ng0keS/LWyde7VjwuBuq4oDe0i4mjjnURgh7J+oKe1s35zaroJcNdyqVvKnoA22fpoFdKeW2Sfy/Jh5Kk1nql1vqHSd6d5MOT3T6c5Hsm99+d5KO11su11s8neTbJt5dS3pzktbXWT9daa5KPTD2GLTOmMXqCHusy5qC3SkWvz7/744LL0D5UGgJj9AC2zyoVvQeTvJTkfy2l/MtSyt8vpdyW5E211heTZHL7xsn+dyf5wtTjX5hsu3ty//D2m5RSHiulPF1Kefqll15a4dDpy5jG6K0z6Lmw3G3rGqM3ptbNPpdXqPXo4KKiN5vWTYDts0rQ20vyZ5N8sNb6byb5RiZtmkeYNe6uHrP95o21PllrfajW+tBdd9216PEyAGMao7eOoGfyB5ImoLVVuWVo3VzMlStN2FPRm08bjLVuAmyXVYLeC0leqLX+2uT7j6cJfl+atGNmcvvlqf3fMvX4e5J8cbL9nhnb2ULtBdauVfSG9rvRrYOD3Wzd7CvoHdU50G7zerzR5ctN2DvqfAnGAOO0dNCrtf5Bki+UUv7UZNPDST6b5BNJ3jvZ9t4kPz+5/4kkj5RSbi2lPJBm0pVfn7R3fq2U8s7JbJvvmXoMW+bixeTs2ZvXEzt/vql69NGadhRBj3XZ5eUV6sz+jM06aixwu01wudFJ58v7F8A4rdBMlCT5oSRPlVJuSfJ7Sf6zNOHxY6WU9yV5Psn3JUmt9TOllI+lCYMHSX6w1np18jw/kOSnk5xL8ouTL7bQcQP+k+aCY5UL4nVaZ+umC8vdNublFVZp3ay1WaZk1mt+k06q6H35yzdv32XHvdcJegDjtVLQq7X+30kemvGjh4/Y/4kkT8zY/nSSd6xyLIzDceNA2p+/9rXdHtNR1hH09vaai2QXSrttV2fdTJr2za6DnoreYk4Kxs4XwDituo4eLOSomd2G2OK4jqCXGBPEeIPewUFy7dryrZtJP3/7gstiTgrGly8nV6/e/DMAhk3Qo1NHVfSGODvluoKe1ifGurzClSvN7bKtm0k/E7IYc7aYk1o3E+EYYIwEPTp10hi9IV2ACXqsy1grepcvN7ertm52TUVvMSedr8Q5AxgjQY9OzTNGbyjaYzl7drXnMSaIXQ56fbymVfQWc9L5SpwzgDES9OjU2MbonT+flLLa8xijxy62brav6SFW9A4OhrWUS9+O615Q0QMYL0GPTo2tdXPVts1EBYEmVOytMMfxmCt6QxyjN70PxwfjIb43AzAfQY9Oja11U9BjHQ4Odrd1c4gVvUTQmzbPZCzewwDGR9CjUycFvSFdfK0z6A3p96J762rd7DroraN1c4hj9BLBZdpx50swBhgvQY9OjXGM3qpU9Fg16J0+ff15ujTmil4pswOq4HIzrZsA20nQozO17uYYPZOxsGrQK6UJe2Nq3exzMpb2A6VZEykN8b2mbxcvNu3Bs/5GnS+A8RL06Mzly03YmxWezpxpLjSGdDGhose6rBr0kub1MabWzdOnm6VJ+vjbP+oDpURFb5ajOi0S5wtgzAQ9OnPcOJB2+5AC0bqDXq2rPxfjtI6gd+bMuFo3k+Zvv8+K3iwqVDc77r3O+QIYL0GPzhw3s1u7fUgXE+sMeteuWbdrl421ordq0Lvttv7G6KlQzW+eCuiQ3psBmI+gR2eOG/CfbG/Qc6HEWIPeKq2bSRP0+pp187j3mcTrcZrWTYDtJOjRmV2u6LXPx+6pdY6g99RTyf33J6dONbdPPXXTLmfOjK+i11frporeYo47X6dPN//9vX8BjM9e3wfA7tjlMXrt87F7rl5tbveOerd96qnksceu/4E891zzfZI8+ugf77a3N74xen21bhqjt5jjWjeT5meCMcD4qOjRmTG1bu7vN1/rDHoulHZTW4U7sqL3+OM3/+FfvNhsnzLW1k0VveE7Lhgnw3pvBmB+gh6dOal1c0jrzZ1UfVyECsJua6twRwa955+fa/tYJ2MZ6vIKXo/XndS9IOgBjJOgR2fGNEbvpGNdhAvL3XZi0Lv33rm2W15hfsdVqE6dan4fFb3rtG4CbCdBj86MaYzeOoOeit5uOzHoPfHEzX9o588326f02bp55PjCE/TZunlccBnSe80QaN0E2E6CHp0Z0xg9QY91OTHoPfpo8uSTyX33JaU0t08+ecNELEl/rZu33toc1jL6XF7huNeuCtWNTmrdHFJbPQDzM+smndnV1k2Tsey2E4Ne0oS6Q8HusL6WV1i2bTO5/pqudfmwuKhaVfQWNc/5+tKXujseANZDRY/OzFvRq7W7YzqKih7rMlfQm0MfyytcubL8jJtJU9GrtdsPOdpxhSp68zk4OHmGYcEYYJwEPTpz6VJzsXvUBe/588m1a91fzM5iMhbWZZ1Bb2wVvdtua267HKd30gdKieAybZ4ZhgVjgHES9OjMPO1B7X59U9FjXXY56PXxty+4LGae9zrBGGCcBD06M89aTe1+fVtn0DtzprlIH8LvRffaoLfszJWtPpZXWEfrZqKiN2TOF8D2EvTozDxTeCfDuKBYZ9Brn0cFYTe1VbhdrOj1EfRU9BazyPkawvhpAOYn6NGZXa3oJaYn32W73LrZBr0u//ZVqBYzb+vm1avDGD8NwPwEPTqzq2P02ucZwu9F99YV9MbYutm+flT0hmveYDy9LwDjIOjRmbFV9Pb2Vr84bwl6u0tFzxi9IZs3GE/vC8A4CHp0Zmxj9NZVzUuM0dtlgp5ZN4ds3tbN6X0BGAdBj86MrXVz3UFvCL8X3Rtz0Btj6+a8Fb12ofBdN8/5UtFbn//5f05+6Zf6PgpgVwh6dOak8DSkhcXXHfRMxrK7xjxGb4ytm1oRFzPP+RrSh3Bj97f+VvKhD/V9FMCuEPTozNjG6KnosQ5jruita8H0IVb0pvfdZVo3u1NrcuFC8od/2PeRALtC0KMzuz5Gbwi/F90bc9BbtXXz1Knuq9kqeovRutmdr3+9eT8Q9ICuCHp05qQxetvcumkylt3VBr29vdWeZ4ytm0nzt991Re/UqeODdfva9pq8/n579uzR+wzpQ7gxu3ChuRX0gK4IenRif7+pRhwXnvb2murBEC4mjNFjXcZc0VtH0Lvttu7H6J0/n5Ry9D5D+lCpb/OcL0FvPV5+ubkV9ICuCHp0Yt4FyIfS4qh1k3Vpw9kYg96qrZtJE/S6/Nuf57WronfdSZ0WidbNdZkOerX2eyzAbhD06MQ842banw/hYmITQc907rtpnbNuXrvWfHXh2rXm2MfWunnp0vzBxYcviwVj52s1bevmwYFzCXRD0KMT8wz4T4ZT+dpE0EuGEWLp1jpbN5PuqnpXrjS3Y2vdVNFbzEmTZCUqeuvSVvQS7ZtANwQ9OqF18/rzslvGHvTG2Lqpoje/ed7rbr21GcPnfK1G0AO6JujRiUVaN/u+mKh1M5OxJP3/bnRvf7+5SD59erXn6TroXb7c3I6tojdPhUpF77p5gnEpw3hvHru2dTMR9IBuCHp0YkwVvcuXm7Cnosc67O+vXs1Lrj9HV+M81xX0+lheQUVvfvME46Q5Z4LxaqYrel/9an/HAewOQY9OjGmM3ryhdBGC3u5aV9Abc+umit5wzdu9MIT35rF7+eXkm76pua+iB3RhxSV8mfajP5r88i/Pv/8jjySPPba54xmSMVX0Nhn0XFjunrEGvbG2bs5T0WsXB+/7vWYI5jlfyTDem8fuwoXkT/yJ5Dd/U9ADuqGit0ZXrzYXYfN8PfNM8oEP9H3E3RnTGL1NBD2tYrtrf/96SFtFGxbHFvRuv715/bcVwk2bp6J36lTze/ngRetml15+OXnwwea+oAd0QUVvjR5/vPmaxw/9UPKRjzRjwUrZ7HENgdbNG5+b3bHuil5XY/TW1bp5553N7csvJ29+82rPNQ8VqsVo3ezOhQvJN39zcy4FPaALKno9efDB5I/+6MZZuLaZ1s0bn5vdcXCw262bbdDr6r1OhWox8wZj52s11641E7DceWfyutcJekA3BL2ePPBAc/v5z/d7HF2ZNzydO9fsW+vmj+koxuixTrs+Rm+6ordptTavMRW9+bTnS0Vv8/7wD5uw9/rXC3pAdwS9nrR9+r/3e/0eR1cuXWrGxZzUBtZecLz66uaP6SgqeqzTWJdX2ETr5qa17xsqevNZ5HwJeqtpK9oqekCXBL2e7GJF79y5k8cjDiEQmYyFdVLRa267CHrzjgVOBJdksfMlGK+m/fsX9IAuCXo9ec1rkje8YXcqeosM+G/378smgt4ttzQVzV2/sNxFgl5z20XQm3d230RwSRZ7rxOMV9P+/WvdBLok6PXogQd2q6LXRdB76pmncv9P3J9Tf/NU7v+J+/PUM08t/BybCHqluFDaVetu3RzTgulPPfNU3vbk/cnepfz3/8f/stTrcREqeotZJBg7X6vRugn0wfIKPXrwweTpp/s+im4sMuA/We6C4qlnnspjv/BYLu43D37ulefy2C80K9I/+q2Pzv08mwh67fPtegVhF411eYVVK3o3vB7PXcgf/eGZpV6Pi1DRW8wi73XnzjVj+q5da7oTWMys1s1dWV4J6I+36x498EDy3HPNQuvbbpG1rZLlLsAe/+Tjfxzy/vjf3b+Yxz855+KG7WMWqAoswifiu2lXWzdveD2eezm5eOdSr8dFqOgtZtHzlfQ7UdaYvfxyE+pe97rm6+rV5Bvf6PuogG0n6PXowQebi7YXXuj7SDavi9bN5195fqHtR7l4MTl7dv2fWrdLR7Bb9vevh7RVdB30Vm3dvOF1d/7l5NKdN29fMxW9xSx6vqYfw2IuXGgC3unTyR13NNu++tV+jwnYfoJej3Zp5s0uWjfvvf3ehbYfZd5QuigVhN001uUVVq3o3fC6m1T0btq+Zip6i1l0Mpbpx7CYl1++PjHR617X3BqnB2yaoNejXVpLr4uK3hMPP5HzZ278R86fOZ8nHn5ioefZZNDzafjuOTgYd+vmshW9G16Pk4reMq/HRajoLWaZ1k1BbzkXLjQzbiaCHtAdQa9Hb3lL0x64CxW9RcfoLXMx8ei3Pponv+vJ3Hf7fSkpue/2+/Lkdz258MQPKnqs01jH6F250vyby7YwT78ec+5Ccun1+bvfufjrcRGLBpeDg+4qpEOkdbM7KnpAH8y62aMzZ5J771XRm7bqp8aPfuujK19IbironTuXfPGL639ehm2eoPfUM0/l8U8+nudfeT733n5vnnj4iZv+jvto3Vx1Db329fh3riZ/9f9Kvuu+zYW8ZPngso4gPkZaN7vz8svJn/7TzX1BD+iKil7PdmUtvS7G6K2Lih7rdFLQa5cheO6V51JT/3hZkMNrzvXRurlq0Gt1tWi6VsTFOF/d0boJ9EHQ69mDD+5ORW+ei4mzZ6/v3xdBj3U6KejNuyxIH62bqyyWPq2roKcVcTHt7z7Pe7Pztbz9/eSP/uj66+D225tbQQ/YNEGvZw88kHzpS9sdAK5ebaoD81x8nTrV/zIEJmNhnU4KevMuC7INFb0LF9bzfEe5eLGZvn6eVkwVquZ3v+WW+Zb/cL6W1/7dtxW9M2eS224T9IDNE/R61s68uc3tm4t8yt7ut61Bz0XS7jkp6M27LMgYx+i1umzdnPe1q0I1f6dFcv287vL5WlYb9NrXQdK0bwp6wKYJej3bhbX0Fg1621rRO3euaYfrqiLDMJwU9OZdFkTr5snmHQucqFAli52vNhDu8vlaVvt3L+gBXRP0erYLa+ktMuA/6b/ytcmKXuIT8V1zUtCbd1mQMbdu3nFHUko3Fb1532dU9BZ7rxOMl9f+3betm4mgB3TD8go9u+uu5n+g21zRW2QK73a/bQ56Fy8mr3nN+p+fYdrfP3kM1DzLgoy5dfP06ebCVkVvWATjbhzVuvnii/0cD7A7VPR6Vsr2z7w5pqC3yMQxi1LR2z3XriW1jnfB9HW1bibNRa6K3rAs8qHW3l7zd7zLwXhZWjeBvgh6A7Dta+mNaTKWRY91Eca47J62+raOoHdq8m49xtbNpGlbU9EblkXOV9J/t8VYXbjQBOXpTo477ki++tX+jgnYDYLeALQVvVr7PpLNWGaMXl+fsi9afVyEC8vds86gV0rzPGMNeip6w7Nom/q5c7t9vpb18svNBx2lXN/2utclr7zSVP0BNkXQG4AHHki+8Y3kK1/p+0g2Y0ytm4Ie67TOoJc0VYGuxuiNsXVzmYreLgeXRYJxoqK3rJdfvrFtM2mC3rVryde/3s8xAbtB0BuAbZ95c0ytm10EvV2+sNw1mwh6KnpHWyS4nD17/TG7SutmNy5cuHHGzaQJeolxesBmCXoDsO1r6ano3ficLpS68bnPJd/5nf2eb0HvujvvTL72taZSuCmLBJdTp5qwt8sfvGjd7MZRFb1E0AM2S9AbgDbobWtFb0zr6B0V9J565qnc/xP359TfPJX7f+L+PPXMU0c+x1H7moylW5/6VPJP/knyO7/T3zGsO+idOTPu1s3k+lTzmzCronfca/fcud1+PS56vlT0liPoAX2xjt4A3HZb8sY3qui1zp9PXn21Gb9wquOPImYd61PPPJXHfuGxXNxvfvjcK8/lsV94LEluWvvsuH3//TseveHfYLPaMa9f/nJ/x6Cid117ofvyy8k3f/P6nnfa4YreSa/dPid+eTYQtQAAIABJREFUGoJFz9e5c5tvv91GWjeBvqjoDcQ2r6XXXki1Y2JO0udYtllB7/FPPv7HFz5/vN/+xTz+ycdvevxx+2rd7NZLL9142wdB77rpoLcJ1641HxBNV6hOeu3uckVvf7/5W1rkvU5Fb3GXLjVfKnpAHwS9gdjmtfQuXmxC3rzVuT4D0ayg9/wrz8/cd9b24/Y1GUu3hlTR21tT70RXyyvUOr7WzVdfbW4Xee3uckVvVkv9POdL0FtM+8GGih7QB0FvIB58MHn++e4+re/SogP+hxb07r393pn7ztp+3L633tqso+RCqRtDCnpjW17h4KAJe2Oq6M0KLie9dne5ojdrNuR5zteuBuNltR9sHK7o3X57cyvoAZsk6A3EAw8kV68mX/hC30eyfmMPek88/ETOn7nxFzh/5nyeePiJmx5/3L6l7PaFZdeG0LrZfnAzttbNy5eb2zEFvVnB5aTXrore4ufL+9di2r/3w0Fvby/5pm8S9IDNEvQGYpvX0ltmraZkOEHv0W99NE9+15O57/b7UlJy3+335cnvevKmiVjm2deFUne2taLXRdBrl0BYZ+vm+fNNcOyyonfS63GXP3hZ5nx5/1pcW9E73LqZNO2bgh6wSWbdHIhtXktvkUWMk36XIbh4sRlLePgC99FvfXRmsJvluH13uYLQpVq3M+h1tbzCJip6pWx20fRZFb3E6/EoR82GfNz5Oneu+aBhf399f9Pb7qiKXiLoAZunojcQ99zTfFrfVUWv1uRDH0q+/vXN/1tja908f765KN0En4h34xvfuD45h1k3F7eJoJdsNugtul5nu++uvh6PCsbHMaHU4o4LenfcIegBmyXoDcTeXnLvvd1V9D772eSv/JXkH/7Dzf9bYwx6m7LLF5Zdaqt5d9yxXRW9MbduJk37WtcVveOo6C3XbbGr52wZFy40s07POs+ve13y1a92f0zA7hD0BqTLtfS+9KXm9g/+YPP/1tjG6G0y6KnodaMNem97W3O+v/GNfo5D6+aNVPSG46jWzeNYC3RxL788u5qXaN0ENk/QG5Au19Jrqxxt4NukRcfoCXqsqm3XfNvbbvy+a2Ot6I0x6KnoLWaV1k3vYfMT9IA+CXoD8uCDzQVpF+Pm2qDXRVvbsq2bfVyAdRH0dvXCskvTFb2kv/bNsQa9TbVutkGv1vU+b7J8Ra+dXGTXLFPR07q5uAsXZs+4mTRB75VXkmvXuj0mYHcIegPS5cybbYWji4pe162bf/fvJj/3c8s9VkVvOwwt6O2taX7jbajoHRwkX/vaep83Wa0VcReDyzLBWEVvcSdV9GrdzOsBIBH0BqXLtfS6bt1c5OLr1lubWS+XvZj4O38n+cAHlnusyVi2w0svJadPJ299a/N930HPGL3GJhdNb8PaTcHlqaeS++9v1k25//7m+9y47y6+Jo9s3TzmfAl6izsp6CXaN4HNWTnolVJOl1L+ZSnlf598//pSyq+UUn53cnvH1L7vL6U8W0r5XCnlO6a2f1sp5ZnJz36ylE1Nbj9sfVT0Nn0BXOviY/RKWa3y9eKLyRe/uNxjjwx6x1z8LLKvil43vvKV5A1vSN74xub7vsbotdU3rZuNTQa9mRWqp55KHnssee655s3oueea7yevyV2v6N20ZugJ50vr5mJqPbl1MxH0gM1ZR0Xvv0ry21Pf/3CST9Za35rkk5PvU0p5W5JHkrw9ybuSfKCUcnrymA8meSzJWydf71rDcY3OnXcmr3lNtxW9l15Krl7d3L/TrmW2aJVs2UD09a83X1/84nLjgGYGvRMufhbZ1xi9brRB77bbmnO+LRW9bWjdTDZX0Ttz5tC5fvzxm99ILl5stme3K3rtB3A3fKx6wvlS0VvM177WvF5V9IC+rBT0Sin3JPlPkvz9qc3vTvLhyf0PJ/meqe0frbVerrV+PsmzSb69lPLmJK+ttX661lqTfGTqMTullO5m3mwvfK9d29wseMlyM7u1+y9zMfHii83txYvNIPdFzQx6J1z8LLLv+fNN+DX4frNeeim5667m/hvfuD1BT+vm0WZ2Djz//OydJ9t3uaI3c+z0CedLRW8xFy40typ6QF9Wrej9RJK/lmT6svVNtdYXk2RyO2meyt1JvjC13wuTbXdP7h/efpNSymOllKdLKU+/1Fcv1oZ1tZbeSy8lb3pTc3+TF8HLDPhPlg960+sCLtO+OTPonXDxs8i+LpS60Vb0kiboWV5hMZtu3WwvgNdpZnC5997ZO0+273pFb9HzpaK3mPYDDRU9oC9LB71Syncm+XKt9TfmfciMbfWY7TdvrPXJWutDtdaH7mo/rt8ybUVvE9OPt65caf7H8o53NN9vckKWZWbCa/dfpaKXLB702vGEi178LLKvC6VuHA5621LRG3vr5h2TEdudVfSeeOLmF/T588327HZFb5Xz5f1rPoIe0LdVKnr/TpLvLqX8qyQfTfIXSin/IMmXJu2Ymdy2l1gvJHnL1OPvSfLFyfZ7ZmzfSQ8+2Fx0bDJ8tVPPf+u3NrddVPTGEPT295vxijcd6wkXP4vs60Jp865ebSpG7WdBd93Vf9A7ffr4/eY19qC3t9dc3G5qjN5Nr91HH02efDK5776mN/6++5rv/3/2zjw8qup+4+8NCUvCviQsSgIoSwD9tW6taKuiIFZqxbqmda1pXapU69bUWm1R26q12kpFtNoarRsuuOCCCmhVsC7EEFzYIgGakEgChJCQ3N8fL8eZTGbues6duTPn8zw8Q27u3Lm5c5fznve7lJQAyGxHz8vx6tkz8l6NPXahm/368VULPY1GowrPQs80zetM09zHNM0isMjK66Zp/gjAcwDO2bvaOQCe3fv/5wCcYRhGD8MwRoFFV5bvDe/cbhjGt/ZW2zw76j0ZRxCVN8WgNwhHLxk5emJQ7VboJRSlNoMfN+tmsoMQFF99xRzI2NBNlS55Itra6ObJqiMcVI6eqtBNINI0XTYJq/uWlADr1/OkWL++03WbyddjwgrDFsfLMHSLGDfYOXrdugF9+2qhp9Fo1CGpjW8nbgXwuGEYFwCoBnAqAJimWWkYxuMAVgHYA+AS0zRFvceLADwIoBeAl/b+y0iie+l9+9tqPkMIvbFjOcMeROimlxw9Lzl2mzcDI0awEIs0oQdwsBNP2MXDYl3t6KlHONbRQq+1FWhqisygB4UQerIIu6MHqBN6cR0qGzLZ0WtuZpVnt+gWMc4R53kiRw+gw/3VV8Hsj0ajyTykCD3TNN8E8Obe/9cDmJpgvTkAusS7mab5PoBJMvYl7BQV8VWloycKUxQUqM9f8hq66XXWePNmYNgwltWXKvQkkckDy6AQQi86dBPgeZ4MoZctcTotSKGXlSV33wUDB0a+I5k0N0dyAJ2S6Y6eKMjlhl69MvN4eaGhgWLaarKnf3/t6Gk0GnXI6KOnkUivXhQqKitvCmGXn88HfbqFbg4bBgwfnppCTzt66hETGdGOHpCcPD3Zjp4I3VQdhtraqiZsE9COXqrg5XgB2tFzQ3194rBNgRZ6Go1GJVropSCqe+nV1XHA2K8fB8HpVnVTCL2aGnfvDVLo6RlxdcQL3QSS02Jhzx75oZuA+j6Mu3erCdsEkpCjZ0EmFxfxcrwA3sMy8Xh5QQs9jUaTbLTQS0FU99KrrWU4m2HQ0UuXPnqtrXywCqG3ebO7AbF29NKDWKEXHboZNCpy9AD14ZuqhV5Tk/yiMl4cqqwsir1MvB4TFmOxQRdjcU5Dg3V+HqCFnkajUYsWeinIqFHAxo2Ryneyqa2NuBzC0VMVCubH0WtrczegFc7k0KEUem1t7pwDLfTSg7o65miKyYV0EnpiW0EIPZWhm4D8punaoXKHDt1Uj3b0NBpNstFCLwUZPZpOVHW1mu3X1UUGvwUFHNQ1Nan5LDGA8uLoRb/fCaKHnnD0AHd5eroYS3oQ3SwdoDPVr19yQjdVOXqqWyy0tqp19AD54ZtehUsmOlQdHUBLi/fjlYnC2AsNDc6EXlOT+nBsjUaTmWihl4Ko7qUX7eiJqmuq3I7mZjoDbqv3eXG+wiD0tKOnnq1bIxMZAtXVZROhQze7okLotbdzn7Wj5wyvE3CAdvSc0t7OtglOQjdNU91kq0ajyWy00EtBonvpqSDa0ROCT1VBFj/hVOL9TgmD0Mvk4g9BUVfX2dEDeL5roeecIEI3ZQq9lha+akfPGV6rIYv3ZNrx8sK2bRRwThw9sb5Go9HIRgu9FGT4cA4OVTh6LS3A9u3BOnpeBxPi/U7ZvJkFZvLzKfaA1BN6WVmZObAMktjQTYDnRDqEbopt6dDNzngt+iTek2kTL37udZl4vLwgclCdOHqAFnoajUYNWuilIN26sXG6CkdPDHZjhZ4qR89Pwj/gXujl59P16NGDg30vQs/LYNENWuipRYdu+idsoZt+HapMEy5+7nXa0XOGOL+1o6fRaJKJFnopiqpeemKwKwbCwvlIl9BN4eQB7pume80ndIseKKmjpQXYsSN+6GZdXfAFD7ZvZwVQWaRD6GZeHrcts+qmX0cv065HGcJYVaXmdMGp0BswgK9a6Gk0GhVooZeiqOqlJ4SecPRycvggSpfQzVih56Zputd9dYsWeuoQPfTiOXodHfJL+ttRUwOMGCFve+kQumkY8puma0fPHX5DN02TkwGaxOjQTY1GkwpooZeijBrFB0Vjo9ztitDN6IFwQYFaRy+ZQs+toxeU0Mu0gWVQxDZLF4iJjSDz9NraeE7us4+8baZD6CYgX+hpR88dfkM3o7ehiY/b0M2vvlK7PxqNJjPRQi9FEZU3ZYdvxjp64v9hz9Frb+ffMHRoZNnw4VzmdFCsHb3wI4RcvNBNINg8vS1b6HyEVeipCt0E1Ak97eg5w48DKsRhph0zt9TX070WQi4RfftyPe3oaTQaFWihl6Ko6qVXV8eZ+t69I8sKCtSGbgYxa1xfT7EX6+h1dDj/24ISepnoIASFVegmEKzQ27iRr2EUeipDNwF1oZva0XOGX2EcvQ1NfBoamH+XZTPKysqi2NNCT6PRqEALvRRFVS890SzdMCLLVDp6XsWTGLA5HUxE99ATuO2lpx298JNKoZsiP1Sm0AsqR0916ObAgdrRSyY6dFM99fX2YZuC/v210NNoNGrQQi9FGTAA6NdPTehmdNgmQEevqSnSdFgmQYVuxhN6oghGKgq9TBtYBkVdHScxRCU7waBBXJ4MR09mMZZ0C92UVbnRr6O3Z4968ZxK+C3GAuh7mB1a6Gk0mlRAC70URkXlzbq6rmFtKpumexVPOTnsJ6gdPY0btm7l4Kpbt87Ls7PpIgUt9Hr2tK+654Z0Ct1sa2MrDBnICEXMJOHit0opoO9hdjQ0OL/2tdDTaDSq0EIvhRk9OhhHT/wsO3zTNL3n6BmGO0EUT+jl5zP/IdWEXibmBAXF1q1dwzYF+fnBhm5u3Miwzegwab+kS+im7Kbpfh09ILOuSRmhm5kkjL2gHT2NRpMKaKGXwoim6TKbPFs5erKFXlsbC6R4FU9uhV7//nRQBNnZ/NtSTehpR08ddXXWQi9oR09mfh4QjKPX3s5/qkM3AXlCTwsXdzQ3U8jbFQqJRyYKYy80NGihp9Foko8WeinM6NGcWd+yRc72du7kwzlejh4gfxDsJzxIvM/p4Cu2h57ATdP0oHP0ZOUnpSKmCRx+OPDgg8F+7tatXScyBEOGaKHnhNZWvobN0evePXJ83JCJwsVr7jSgQzed0NoKbN+uQzc1Gk3y0UIvhREtFmTl6cXroRf9s2xHz0/ejHifG0cvuoeewE3T9CCFHqCm+E2q0NAAvPMO8MYbwX6uXehmUEKvo4PnXRiF3u7dfA2T0PMaIg5krqPn9V6ni7HY09DAVzeOXlMTnXSNRqORiRZ6KYzspukiPynW8cjNZV892YNgP+FUgHuhl8jRcyL0Ojr8zXK7IRNmxMXkxPr1wX2madoLvYYG9YVMAF5rbW1yK24CweToCaEXptBNP9duJjp6MoRxJh0vtwih59TRE1WCm5rU7I9Go8lctNBLYQoLWchBtaMHMHwzrI6eaTK8NZHQ27o1MnhNhHDXgirGAqT3QElMTmzYENxnNjZSxFmFbgKRXnsqUdEsHQhh6GZ5OVBUxGSwoiL+jMgAWDt6yUGHbqpFnNduHD1Ah29qNBr5aKGXwvToQUdAtaMHqGmaLiNHz8lgoqmJn5VI6AH2eY5+RakbMmGgJCYnNm4MxkEDEjdLF4gJjiDCN8Ms9KSFbpaXA6WlVPumydfSUqC8HNnZ7BMqnA+/aEfPHX5CN0Xrm0wSxm7RQk+j0aQKWuilOKNGyXf04gm9goLwhm7Ga60gcNo0PRlCL50HSmJyor3deTEcv4iJDDuhF0SLBVVCL1Shm2VlXS/g5mYuR6Rpugy0o+cOP8cL0JWD7XAbuimE3ldfqdkfjUaTuWihl+LI7KVXWwvk5fFfLCocvaBCN62EntOm6drRk8vatZH+cUGFbwpHL1HoZpCOXk0N3bd4YdJ+CFXoZnW15XKZQk87eu7wm4+shZ412tHTaDSpghZ6Kc6oURw02uWYOSFeDz1BQQEHyjIHkFroxScThN66dcA3vsH/B1WQxS50U5z7QYVuDh/urU+ZFaEK3Rw50nK5dvSSh98Kw716Zdbxckt9Pd333r2dra+FnkajUYUWeinO6NGR9Ba/1NYmdhgKCvg5sgZeQHA5elZCb9AgPnDtwgeDFHrp7iC0t/N8/e53+XNQjp5d6OaAAcwtCip0U3bYJhCs0PMdujlnTtcLKjeXy6EdvWSiQzfV0tDAsE0R1WCHFnoajUYVWuilODJ76Vk5eip66cnK0bNrLL55Mz+jb9+uv8vKogDUjl5wiAIsxcXsbRiko9ezZ/zQZIDnQlBN01UJvSBy9KSFbpaUAPPmRcoHFxby55ISABwIp4Kj17MnXzPJodKOnlrq652HbQJAnz68RLTQ02g0stFCL8WR2UvPztED1Ag9P45ee7v9oFY0S080e+qkl54uxiIPMSkxahQr6geZozd4sPUsehBCzzTTw9GT0l6hpIRKv6ODr3tFHsCBsGiJ4Rc/jl5WFsVeuk68xEPn6KmlocGd0MvKYhVaLfQ0Go1stNBLcYYO5YDLr6NnmnT07ISezEGw39BNpyFViZqlC1JV6KXrQElMSoweTRMnKEevri5x2KYgP1+90Nu2jed+2IWeyobpQGQgLKPFgoxQxHSdeInFNHXopmrq651X3BT076+FnkajkY8WeilOVhadEb+O3vbtHMAFHbrZrVsk3MwtTgVR2IReuucErV3L733ffenoVVfT0FHN1q2Jz29Bfr76HD1VrRUA3g+yskISummDEHoywjf9OlS9eqXv9RhLayuvRx26qQ63oZuAFnoajUYNWuiFABm99ISLkcjR69+fM/gy3Q4xa+w0IT0Wp0JvyxZ7odfYCOzcmXgdLfTksW4dCytmZ9PRa2uLFMxRiQjdtCKI0E0h9EQPR9lkZ4codNMCWUKvvZ3iRTt6zvAbaSHem673Lxm4Dd0EtNDTaDRq0EIvBMjopSdcjESOh2HI76XnN+HfidDbtYsPRyuhJwbcVmIjSKHXrRsH0ek6sFy7NlJEqLCQr0Hk6TkN3WxqktOuJBEqHT0gOKEXVOimX6EnQ7hkkqPnt0gWoIWeFc3NQEuLDt3UaDSpgRZ6IWDUKD4AvvrK+zbsHD3xO5lCT0bCv9hOIrZs4audowdYh2+KQYuowKeadB4orVsXKSJUVMRX1Xl6bW10bZ2EbgJqwzc3buTEidU56YecnJA0TLdBltCTJVzSdeIlFhmTWjp0MzFum6ULBgzQQk+j0chHC70QIKPypp2jB7Agi+zQTdWOnlUPPYFToZeb6z3M1C3pKvR27uRkQdCOnhhcOXH0ALXhmxs3soiS19xUO7Kz1ebohS10U5ZwScfrMR46dFMtoriQdvQ0Gk0qoIVeCJDRS08MbK2EnorQTb+z7GI7iegi9MrLaSNlZfG1vPxroWfVNN2vKHVLug4soytuAuxpN3iwekfPrlm6QJz/KoVeTY26sE0gfUI3e/emGJYVuqkdPWfIEMa5uXR+29vl7FM64dXR69+fRdNUXtsajSbz0EIvBAih58fRq61lQ3Gr0ETh6Nk1KHdK4I5eeTlQWkr7yDT5WlqKfs+Xo1cvZ45eUKTrjHis0APo6ql29LZu5WuqhG6qKsQCqBd6InRTtdAzDA6GtaMXLDJCXcV7M0Ucu8GP0AMYgq7RaDSy0EIvBPTrxzAQP45eXZ39ILiggIM8WQ8aWTl6dkKvW7e9Tk5ZWdeVm5th/LrMtsVCMoReOg6SopulC4qK1Dt6QuilSuimSkcvJ0d96Gb37sGEMQ8a5L+Pnnb03CErdBPIHHHsBj+hm4AO39RoNHLRQi8k+K28WVtrXYgFkN9LL6jQzYICRmqiujr+StXVKSn00nGQtG4dQ/KiBZdw9GQ5xfFwGrrZp4/8NiLRbN/OiZKwh26qdvME2tELHlnHC8gccewGv46eFnoajUYmWuiFBL+99Jw6eoBcoReEo/d1ft7IkfFXGjky5YReug4sRWuFaDeoqIjlxlW6aMLRsxtciTYiqvZF5IGGWei1tqovxCKQIfS0o+cOWVVKo7elidDQwOPjtoKzFnoajUYFWuiFhNGj6Yp4TX534ugJoSdrEOxXPDlpLN6pWfqcOV0/MDcXmDMHI0ZQ6CVylbSjJ4fo1gqCICpvbt3KgZKTSpf5+epy9IIQekGEboZJ6Ol2Ae7QoZtqqa93H7YJaKGn0WjUoIVeSBg1ijPtVq5UIkyTA9ugQzf95uhlZzOEzLGjV1ICzJtHZWEYfJ03DygpwfDh3E5TU/zt6Bw9/5hm52bpAtFLT6XQc9IsXTBkiDpHTzRLD3MxliBDNwcO5MDYT1ivLEdvzx61AjpV0KGbaqmvdx+2CWihp9Fo1KCFXkjw00tv2zYOYuxCNwcPpj6S6ej5GXwB1s7Xnj3c10499EpKWPmjo4OvJSUA7HvpaUfPP3V1/JsSOXoqC7Js3Wp/fgtUhm6mg9ALOnSztZX9F72ihYs7dOimWhoatNDTaDSpgxZ6IcFPLz0xqLVz9LKzKfZkOHrt7RzAJRp8lc+9GEVXZSPrtwaKrspG+dyL465nJYhEKwirZukCu156Wuj5J17FTYBVY/v3Vx+66dTRUxm6uXEjB3l+JzisCMLRC1LoAf7CN3Uoojuam1mp2EmYcyLE8coEYewWr6GbvXuzqJgWehqNRiZa6IWEkSP5EPDi6IlBrRPHQ1bTdKvBV/nci1FaMxcberfDNIANvdtRWjM3rtizEkRdmqVbkGqOnijGorISZdDE66EnKCxU6+i5Cd3Mz+ex9+MiJUJ1awUguPYKQSBD6Om+cO4QIfV+2mc4yZ/OVLyGbmZlcVJMCz2NRiMTLfRCQvfuHECqdPSASNN0v1iFU5WtnYfmmNnk5hwuj8WqOqUQekOH2u+PEIOpIvRycynydu8O7jNVI85NkZMXTVGROkfPNN2Fbor1VIRvBiH00i10E/Dv6PXosbfFikcyzdGTEVIvtqWJYJreQzcBYMAALfQ0Go1ctNALEV576bkRerIcPatZ9uq8+KVD4y2X5ej17g307Rtf6LW18V/QQg9ILwdh3TqK7njHUTh6KhzMnTspmN04eoCa8M2aGu9C75lngI8+sl9Ph252RoZwySRHT8akViYdLzc0NTFtwUvoJsAQdy30NBqNTLTQCxFee+k5bSYN0NGTKfTiDShG7uwW9z3xljsRek4cPQAJe+nJyPFxSzrOiMeruCkoKgJ27AC++kr+57o5v4GI0JPt6LW0cF+8FmL56U+BW2+1X0+HbnbGb3VfID2vx0To46UOr83SBVroaTQa2WihFyJGj6a4cTuLWlvLkBAnyfcFBRyQ+32AW4mnOaNLkRszUM1t4/JY7ITeoEHOB6WJhJ6Mqn1uScccl3g99AQqK2+KZunJDt0U55YXR2/3bu7Pli3266ZT6KZwPrSjFxwyHb10un/JoKGBr1roaTSaVEELvRAh3BK3g+W6Onel5wH/g2Ar8VRy0T2YN+IiFO7oBsMECnd0w7wRF6Hkonu6rGvVb65TDz0HiKbpbvZVFek2I97WBlRXWzt6gJo8PSH03PTRA+SHborWCl6EnhB4wqW2Ip1CN3NyGFKtHb3gkCGMs7J4jmSCMHaDOI/9hG6qiHrQaDSZS3ayd0DjnOheehMmOH9fba2z/DyAjh7A8M14RTWcYlcJr+Sie1CCrsIuFitHb8sWd0JPOHqm2bninNh+Xp7zbfkl3QaW1dVsXZgMR89t6GZeHv/JdvT8CD3R9iMVHL0gQzcBuh/a0QuOXbsiPdv8kI4tYvyiHT2NRpNqaEcvRHjtpVdb69zRE0JPpaPnBrvQTbdCr62t66AymY5eugwsrVorAJzh7t1braMXfY6XV5Sj6M4iZN2YhaI7i1BeUd7pPUOGyBd6Qqx5EXrCaW5qsh88q87RCzJ0E/Av9LSj5w5ZFYatoi0yFRmO3s6daq9vjUaTWWihFyIKCjjz7LbyZl2dc0dPrOe3IIusAieJhJ5penP0gK5N03Xopn8SNUsXGIa6Xnpbt0ZCAAGKvNKFpdjQuAEmTGxo3IDShaWdxF5+vhpHr08f/nNLdEix3bWXTqGbAIWecEK8oB09d8g4XoB165tMRYbQA4DGRjn7o9FoNFrohQjDcF95s6ODA2EvoZt+kNHEGIgIvdiy/A0NdB68CL3YPD1djMU/69ZRbFlVnFTVS080SxfhuGWLy9Dc1vnANrc1o2xx2dc/5+erydHz2lohevLBLk9Ph252Rjt67pBxvAAduhmPhgY2Pc/2mBQjhJ4O39RoNLLQQi9kuO2l19BAsec0dLNnTzojqRS6Ga+xuNvWCkBqCb10G1ieIqIEAAAgAElEQVSuXUvHrlv8zhkA1Dp60fl51Y3VcdeLXq7K0fMq9KLPSbs8PR262RkZDlXPnnzNFEdPxr2uV6/MOF5uqK/37uYBWuhpNBr5aKEXMoSj57TxtJtm6QIZTdNlCr3o7QncNEsXiHVTSeily0DJqrWCoKiIAxjZYUmxQm9kv5Fx14teLnL0ZDZw9+vojRnD/9sJPZWOnmkGL/QGDuR54fVvkuFQGUbmhCLKCt3Ujl5X6uu9F2IBtNDTaDTy0UIvZIweDWzf7jynRYSnOXX0AIZv+nU7hIARM+VekSn0evTgQziVhF66DJSsmqULROVN2eGbse1D5kydg9yczl9mbk4u5kyd8/XP+fl0xZqa5OzDnj0UaH4cvQMPZNn6ZAq91la+Bh26CXgvKy8z5yxdJl4S0d7O71iHbqqhoUELPY1Gk1pooRcy3FbeTKaj16tX5zYGXpAp9ID4vfR0jp4/mpo4k+3E0QPkC71YR69kcgnmzZyHwn6FMGCgsF8h5s2ch5LJJV+vI6tfpGDLFoZI+3H0Ro6kYLXL0cvJUS/0gg7dBLyHb8qsIpkO16MVsopkAZkhjN2iQzc1Gk2qofvohYzoXnqHHGK/vhehV1AALF3qft+ikTn4EtuLZvNmluvv3dvd9kQvvWhkFY5xQ3Y2B+zpMLAUOaNOHT2ZeXrt7ZxFj+2hVzK5pJOwi0U4gLW1wP77+98P0UPPqhhNIrZvB3bs4Lk5dKgzR09Vjp7IhQ2L0Gtro+jVjp4zZN7rMkEYu8Vv6OaAAXzVQk+j0chCO3ohw62jV1dHV83Nw6eggA8sP66BzMpuQNcBhdvWCoJEQi8nh/+CJF0GSnY99AT5+QzllenoNTQwr8xNaLLYF0Be5U0ZzdJHjOA57UTotbfLzS8UCKGXjNBNL0JPpkOVLtejFTKjF3Qfvc60t1Og+RF6eXksaKWFnkajkYUWepKwa9Dsdd1YevfmoNZN6OagQdbVEGORMQgOwtHzKvS2bOFDWSBrX92SLgMlux56AtFLT6bQE83SYx09O2SHbvoRemLiwY2jB3Q+h2URttBNcf1oR88ZskM3010Yu0HkmPoJ3TQMhm9qoafRaGShhZ4EnDRo9rJuIkaNct5iIbZQhRNk9NKTWSBBbC8aP0Kvo6Pz35ZMoZcOA6V169g7SoQdWSG7xYJXoSfWlyn0evb0NsgTQm/EiIjQ6+hIvL5wnlWEbzoK3SwvZ8JlVhZfy53fu+LhR+jJdqjS4Xq0QkXopgpnOYyIAml+HD3Am9Dbvh34znf8p1xoNJr0Qws9CThp0Oxl3USMHu3O0XOTnwdEhJ6fQXAqO3pA5/DNZAm9dJkRFxU3nRTekd003UtVWYBCpl8/eUKvpoZunpfiQyJ0c9gwCr09e6yr6gpHT0VBFluhV14OlJbySzRNvpaW+hJ7ffrwb9KOnnpkC+OODrU9HcOEOH9lCD23FWiXLAGWLQN+9jP9fWg0ms5ooScBJw2avaybiFGjgOpqZwM9L0JPrO/H0VOZo7djB/+5aZYuSCWhly4OgpMeeoLCQoqznTvlfLZXRw/geS4zR89LIRaA52LfvgzLFpMXVuGbKoWebXuFsrKuJ21zM5d7ROQQa0dPPbJDN4H0P2ZOEeevn9BNwJujJ5y8qirg73/39/kajSa90EJPAiOz49/Z4y13s24iRo/mIE/kBVmRzNBNVULPa2sFIPWEXtgdhI4OCj27/DyBaLFQ7XxewxK/Qk9m6Kaf1gpCJIrJCyuhl9TQzURfnM8v1KvQ046eO2QL4+htZjqOQjcdhD17EXrLlgFTpgBTpwI33OC9VYlGo0k/tNCTwJzXgNzWzstyW7ncz7qJEINquzy9PXt4w3fr6PXty4Ge39BNWXkgQOcBmB+hV1DAZ2yqCL2wD5K2bAFaWtw5eoC8PL26Oob+eSkeMmSIHKHX0REJ3fTCpk2RCQgh9Kx66SU1dHPkSHfLHZIKjl66hFJbITNHT2wj3cWxU2wdPYdhz26F3s6dwPvvM0fvz38GGhuB3/7W05+g0WjSEC30JFCypAHzFgKF2wDD5Ou8hVzuZ91EiEG1XZ6eePC4dfQMw3/TdFniKV54kB+hl51NsaeFnhyc9tATyG6aHtss3Q2yQjfr6uiuBeXoJTV0c86crhdLbi6X+2DQIOu8xETIdPTSwWG3Q3Y7CiD89zBZ1NdzErFfvwQrlJWhfEwzimYDWTcARbOB8jFdw57dCr333uO94MgjgcmTgZ/+FJg7F6is9P63aDSa9EE3TJfByJEoqdiAkoqY5YVxZrndrJuAffdluwQ7R89Ls3RBQYE/t0NWjl5WFqsZyhJ6QNdeeroYi3fEZINTR2/YMIYeynL0ZAi9jg6eZ14RxVS8CL2ODp7PwtHr04fnYrKEnq2jV7K3CX1ZGcM1R46kyCtJ3JzeCYMGAcuXu3+fdvTcoSJ0M93FsVMaGujmJbqXlPfdgNKZQPPeSZQN/YHSmQAWbkD01dO/P7+n1lZn/SyXLuVnHn44f77pJuDRR4ErrgAWLfJWIEqj0aQP2tGTgZtZbgkz4tnZHF/ZOXp+hJ4MR0/GLDvQ1fnavJkPQK9J76ki9NLF0RP98ZyQlcVzV5aj5yUHVTBkCIWWFycpGpEr66UYy9atdAPFew2Drp5V6GbS2yuUlFCpd3Tw1afIAyKhm25L9ct2qNrb07tqoYrQzbDfw2RRX2/9TCqb3u1rkSdo7s7l0fTvz9fGRmefu2wZcOCBESdx8GDm6b3yCvDCCw53XqPRpC1a6MmgpASYN4+jXTHqnTcv/gDIzboWOOml57X0PEBHz6vQ6+iQ5+gBXQXRli0cDHudqRw+POLCALoYix/WruXx7NnT+Xtk9tLz6+gB/sM3ZTVLF9g1TU9q6KYiBg2iyHQrGrRwcYe437i5XhOhQzc7U19vXYilune7o+VC6DkJ32xtBd55h/l50VxyCTBuHF291tb479VoNJmBFnqycDPLLWFG3EkvPRmhm16a4ba08FWV0PPaQ08wfDgFgnAvku3ohbnhsJvWCgKZvfRkCD2/BVk2bqT48nKdpZrQc+ToKUA4IW4LsqjIOQv75IsVItLCT6iyIBOOlxsaGqyF3sh+8cMeYpe7EXr//S+Pf6zQy8lhYZbPPwfuvtt+O4no6PAX2aPRaJKPFnohZdQoDlCt+pHV1TGXb8AA99vPz+dA0m3jVkDu4EtsJ1boeemhJxCD6i1bKLKSmaMX9lAx0SzdDYWF/A7FhIBXdu3i+e8ndBOQI/SGD+e15hbhLEeHfQ4blsLtFRQhBshuhZ64L8hwqDLB0ZMZUp8Jx8sNdqGbc6bOQW5O5wdNbk4u5kztnLYhntdOhN6yZXw94oiuv5sxg/9uusnbPW7nTuCUU3g/evRR9+/XaDSpgRZ6IUW4KFbhm7W1dDu8zN766aUnM5wKUOPoAXRTdu+m2EuWoweEd6C0ezeFihdHDwC+/NLf5/vpoQfIdfT8tFYQeXmCoUM5wZJICKdr6CbgzdHr2VM7VE6ROakV9vuXbOwcvZLJJZg3cx4K+xXCgIHCfoWYN3MeSiZ3juhx4+gtXQqMH584muCOO/j9XH+9wz9iL5s20SV87jmGgJ59NvDii+62odFoUgMt9EKKk156tbXe3Q4h9LwMgmVWdhPbEdtsbeVgUJbQk72vbgj7wFK0g/Li6AH+8/QshZ6DxsSDBlFk+c3R89NDr6aGgzTh0gER0ZdokiUdQzf9OHraoXKOzNxp3UcvQmsrsGOHTbN0UOytn70eHTd0YP3s9V1EHuBc6LW3A2+9xbYKiRg/nvl68+cDH39s80fs5cMPgUMPBT77DFi4EHj3XbZt+OEP+XkajSZcaKEXUpz00qur85Y3BETe58XRUxm6KULa/Ag9ESaXKkIvrANLMcng1dHzm6eXsNiQw8bE2dkcmPlx9EyTjp6XiptA52bpArteekEIvTA5erIdqnQWLrKrIYttZjq2zdJd4FToffIJK3PG5ufFcsMNDAedPds+H3zhQgrHrCzg7beBE05gNc9Fi9jW6cQTnQtGjUaTGmihF1IGDwby8uwdPa9CT0bopgqh57eHHsBBZU6OFnp+EZMMbh29ESOYz6bM0Ssr63pQm7s2JgYoEv0IvW3buGkZzdIF4txOJPRU5ui1tvK78ZJv6AevxVhUOHrpLvRk3eu6d6cgCOv9SyaiRYudo+eE3FxO5tjlx4v8PCtHD6DIu+km4M03gaefjr+OaTLM86STgOJiNmE/4IDI7/PzgVdfZZ/P6dNZ5EWj0YQDLfRCimHYV97002Ns0CA+xP2EbsocgMkUellZfH+yhV7YQ8XWrWOIn9vvIjubwsivo5dQ6FVXx39DnOWiabpX/LRWAKwdvUS99FQ7ekGHbQIUDX36pIajF9br0Qkyj5dh8B4WJmH87rvAT38qf5JEnLcyhJ5h0NWzc/SWLmVPUic9TEtLgUmTgF/+smvub1sbcNFFwJVXArNmURDGu6ePHEmx194OHHdc5xZFGo0mddFCL8RY9dJrbeWDwquj160bB9Dp6OgBkabp2tHzztq1kTQ4t8jopVdXx8/uUlV25Mj4b4izPD/fn6PnR+i1tfGzY4Vefj4He8kK3UyG0AMiTdPdoB09d8g8XkDXQlmpzt13s22tn5YD8ZAZugnYCz3TpNCzC9sUZGez3cK6dcCdd0aWb9sGfO97wL33AtdeCzz+uPWzcPx44KWX+PdOm+b+etVoNMGjhV6IEY5evLh74XZ4dfSASC89t6jO0TMM7wJWIJqmp4LQC+vA0ksPPYGMXnpbt0ac507MmdP1C83N5fIY/IZuilltL0JPTFrEhm7m5HCSJVmhm0Hn5wm8CD3t6LlDdiuZ3Nzw3L86OoBXXuH/b7gh0sNSBjJDNwF7offFF5yEtQvbjObYYxmaOWcO7z3r1gGHH04H74EHgFtucTZpd/DBrMa5Zg1z+LZvd74PGo0meLTQCzGjRvHBHS/0zE+zdEF+fuo4ei0tfFBv3sz9Eq6GV7Sj5x8vPfQEhYUUSX7ESsJm6SUlnLYvLOSsQGEhfy7pWuEuP5+DNK/u2MaN/AgvDnO8ZukCq6bp2tGLoB09d8gUxkDnsPpU54MPeM+48UZOaFx1lbxtB+3oLV3KV6eOnuC223iNn302cNhhvMe88gpw3nnutnP00cBjj7Fh+8knR4o4aTSa1EMLvRBjVXlThtArKEidPnoAxZ7fZumC4cNZsUyIZC303PHVVxyI+HH0OjoioY9esMxBLSlhbGhHB1/jiDwgcn0IB9wtGzfyOoluj+CUeM3SBUOHZlaOHuBd6GlHzzkqHL2wHK9Fizgpc9FFwNVXA488AixZImfb9fV0wvPy5GzPTugtW8Z737hx7ra7336svvnaa6ym+e67wFFHedvHk04C7r8fWLwYOOssNfcjjUbjHy30QoxVL72Epedd4DV0U4WjJ7brt1m6QLgoa9Z0/owgCXMxFnHO+XH0AH/hmwkdPReI68Nr+KbfZulAajl6yQ7dFCFwTtm1S96EUs+ekW2mK7Jz9MJUjGXRIuCgg3jNX3cd70GXXionBFo0SzcM/9sCnDl6Rx7p7fNuuIF5eu++C4wd630fAeCcc5j7t2ABi9zYtW/QaDTBo4VeiBGDbFWOXn4+sHMn/7lBPPhV9GuSLfS++KLzZwRJmHP0xDnnx9ED/BVkkSH0xPWRDKFXUxPJx4tl2DAKvXgDJ5U5esl29LZtY1U/p8h0qEQVSTcTLxUV4QlbM035oZthcfS2baOwOf54/pybS4HyySfA3/7mf/v19fLCNgFrobdxIyfa3OTnRZOXB1x+ubx8wtmzgeuvZ57fddfJ2aZGo5GHFnohJjeXrlsiRy8nh+EZXvHaS6+5ma6ArF5cYmCyfTv3RYbQE+FyqSD0wjBQisWvo7fvvhxYe3X0Ojoo9Pw41kBE6HltsVBT47+1QrxZ+aFDI5VzY0nn0E3TtO8fFo1MRw9w51C9/jp7je2/PzB3buoLvt27eXwzUegtXswJBCH0AOAHP2BPuBtuSOyeO+Htt+kWjh/vfz8FAwbwPIx3Ton+eW7z81Ry443Aj38M/PGP7ieGNRqNWrTQCzmJeunV1nIQ7CeURAg9t26HijwQgG3Q2tvVOHoibCtIcnI4aA/DQCmWtWs5g+11IqF7d34HXh29xkaeC8l09HbsoBDz4+jFC9sErHvppWvoppem6cnMOfvDH3j+7LMPcPHFzH+6557UFXyyc6fFtsIQkbBoEe9Vhx0WWWYYwF13cf+vvtrbdj/5BDjxRE5czZ0rZ18BOnoA73OxLFvGnpMHHijv8/xiGBTOpglUVSV7bzQaTTRa6IWcRL30amv9tyAQ73fr6KkIDwIi+XQyhF6/fhykbN/OVy+94GQQpqp10fhprSAoLPTu6CVslu6S/v3pPHsRelbFVJywaVPi9wqhF89pSOfQTcC50Gtro9hPhnD56CNWK/zFL+jovPIK2zRecgkwZgzDAWMbUycb2W1vxLZS/f5lmhR6xx7btVrz2LFsIv6vfwFvveVuu+vX0xHMzQVeftl/dEE0QujFc/SXLgWmTJEXMSOLSZP4+sknyd0PjUbTGc/DW8Mw9jUM4w3DMKoMw6g0DOPyvcsHGobxqmEYn+99HRD1nusMw/jCMIxPDcOYHrX8IMMwKvb+7i7DkJXSnP6MHk2nK3bQZ1mR0CF+QjdTXegZRsRNSUbYpiAMA6V4+GmtICgq8u7oySg2BFDgDxniLXTTT7N0IBK6GQ9xjscTeukcugk4F3oqWqM4vR5vuw3o3Rv42c94LznuOAqFV1/leX3ppRR8f/1r6gg+Vccr1R29Vat4rUaHbUZTVkZH7pJLnF9TtbVsGN7cTJEnco5lkUjo1dcDlZWpFbYpGDOG947KymTviUajicaPj7EHwJWmaU4A8C0AlxiGUQzgWgCLTdPcH8DivT9j7+/OADARwPEA7jEMQ8xJzQVQCmD/vf8S3JI1sYwaxXylL7/svFyGo+e1IqHsym4qhB6QOkIv1QdKsbS304mT4eh9+aW74hsCWY4ewOvEi6PnR+jt2AE0NXlz9NI1dNOt0JNd9Elsy+563LAB+Pe/gQsvjAzIAQq+Y49laN1rr3Hg+/Of8/Xuu5Mv+FSFbqb6RNWiRXydPj3+7/PygDvuAFaudBZ+uX07G4Vv3Ag8/3zEyZJJIqEnXEevhVhU0q0bMGGCdvQ0mlTDs9AzTXOzaZof7P3/dgBVAEYAOAnAQ3tXewjAD/b+/yQA/zZNc7dpmusAfAHgUMMwhgHoa5rmO6ZpmgD+GfUejQ2JeunJcPR69mSIY6o4euJvlNFHD0gdoZfqA6VYNm2iIJDh6O3ZE2kz4AaZQm/IEH9Cz0voplVrBYDXXY8e8XP0ROjmI4+wCMLzz/srJhGNdvTsr8c776Somz07/u8NA5g6lT3aXn+duXuXXcaiLSLcNxmoOl4tLZxsTFUWLQImTqRrl4hTTqFIv/566+fd7t1sEP7RR8ATTzCEUgVC6MUWJlq6lNfnIYeo+Vy/TJyoHT2NJtWQkplkGEYRgG8AeA9AgWmamwGKQQDCVxoBINp32rh32Yi9/49drnFAvF56u3Zx1tGvowd4a5quKkdv7dpIbp0MtNDzhjjXZDh6gLc8PVmhmwCvEy+hmzU1FCdezke7/D7DSNxLLzubxSPa2yn0Zs6kyz1iBJsY/+53wIsvur9ugeQKvb59+belsqP31VfAffcBZ5zBnDwrDAM4+mgKvldfpWi/4w55++oWFTl64tgn261MxM6dFEeJwjYFhkHXtbkZuPba+Ou0twM/+hEreD7wAPC978nfX0EiR2/ZMhaUSdY1asekSYzSiFdERqPRJAffQs8wjN4AngIw2zTNJqtV4ywzLZbH+6xSwzDeNwzj/Tqv9dDTjH324eAo2tGTPQhOlaqbu3bJC9sEUkPohSH0KRZxrslw9ABveXpbt/LYyfju/IRu+inEAiR29IBIL714/OEPnDlvauJA9s9/pqj47DOWi//e9ygU990X+OEPI9Vl7Uhm6KZhsPJmKjt6c+dSPFx1lbvtHnsscOaZwLx57tpHyERF6Gaqt4hZsoTndKKwzWjGj2dxnQcfBN55p/PvTJN5l08+Cdx+O3D22Up292viCb0dO4APPkjN/DzBxIl8XbUqufuh0Wgi+BJ6hmHkgCKv3DTNBXsX/29vOCb2vooh1EYA0cET+wDYtHf5PnGWd8E0zXmmaR5smubBQ2SWuAox3brRGYl29ITQS5ajpypHD5Ar9MQgXTt67li3jkVM7BwNO8T7vTh6MpqlC4YMoWBy60r4bZYOWAu9RI5eNL17M19n9mzg4YdZ2ryxkQPc22/noPC11yj8nAiMZDp6AB3SVHX0WlpYjn/6dPbPc8tVV3Gw/ve/+9tHr6gSxtHbTjUWLeJ36jSn7frr+Vy45JLOucM33sjv7ZprgCuuULOv0fTqxRDtaKH3zjvcp1TMzxMIoafz9DKbRYsYVaZJDfxU3TQA3A+gyjTN6ICU5wCcs/f/5wB4Nmr5GYZh9DAMYxRYdGX53vDO7YZhfGvvNs+Oeo/GAbG99IQ7IUvoJdvR69Ej0g8w3Ry9MBZjWbuWAsev89OrF89RL46ejBxUgdem6X6E3qZN7IXVp0/idYYOjZ+jZ0efPhR4V1wBlJcDCxdSnJ9xhn0BlzAJvaAdvX/9i5NeXnuuHXAAReJf/pKcUEeVoZupeg9btIhOt9M+qb17M7z2ww+Be+/lsr/9jULv/POBW25Rt6/RGAZdvWiht3QpJ3a//e1g9sELRUU8v3SeXuaycSMwYwar2WpSAz+O3hQAPwZwjGEYH+39dwKAWwEcZxjG5wCO2/szTNOsBPA4gFUAFgG4xDRNMWd2EYD5YIGWNQBe8rFfGUdsLz3ZoZsNDe56dsnO0TOMyIAiHYVeqs6GJ0JGDz1BUVHyHT0vQm/3bq6volm6YOhQip7WVm+fITjySIYcvvKKfchhMkM3AQq99eud/c1BOnodHWyp8M1vUjh45eqrKRb/9S/v2/BKpoVurlkDfP65fX5eLKeeChxzDAeqf/sbK6d+//sUfkE2foon9L7xDevJoWSTlQUUF2uhl8l8/DFf//EPRspoko+fqptvmaZpmKZ5gGma/7f334umadabpjnVNM399742RL1njmmaY0zTHGea5ktRy983TXPS3t9durf6psYho0dz4CusctmOXvQ2nSDb0QMi25Mp9MS2dI6eO2T00BMUFnrP0ZMZugm4O8dFjp0fR88uv0+cn17yB2O54ALg8stZMfL+++Ovs2cPBU0yHb2SEp4P559vX8kxSEfvueeY/3j11f4G+0cfDRx0EEWjl7YiflAZupmKjt7LL/PVrdAzDPY/3LGDeXlHHMF2GrHN1lUTLfR27wbeey+18/MEkybp0M1MpqKCrzt2UOylE+XlnPwJG1KqbmqSS2zlzdpaDtZ69/a/bbdCzzTl5+gBaoRe7950TZKZ7hk2R2/XLoYTynT0qqvdl2dXEbrpRlD5aa0AWDdLF1j10vPCbbexsfdFF0X6cUWzezdfkyn0Tj0VmDOHD9Rf/cp6XVWhiO3tXSMY/vQnnqunnOJv+4ZBsfjZZxSPQaLKAQVS8x728su8T+23n/v3TpgA/P73dPaee07+88wJ0UJvxQpen2EQehMn8p7lNARbk16sXMkJ3MMPZyXboCe0VGCawG9+w6q7Tz+d2u1k4qGFXhoQ20uvro6DVxlhJmIQ7LQgS2srL4IwOHoA8PbbwHXXyd2mG8KWoyfcN5mO3u7d7kRWaytDQpIZuumnWbppOnP0hNDzkqcXj+xs4LHH+N3NmtXVSRXhkskM3QR4Pf7sZ6wsajV7GlQo4ttvA//5D3DllXJcnVmzeM/+4x95LgRFczMLfIg+jDJI1dDN1la2QZg+3ftz8JpruA1RATNoBgyICL1ly/h6xBHJ2Rc3iAbyOnwzM6moACZPZu/QNWuAl0KeiLVrFysm/+53wHnnsXVRVsiUU8h2VxOPeI6ejLBNIOLoORV6KmbZo7cnq1m6YPRo9u9KFrm5dA/c5EAmEzGZINPRA9yFb4qZYlmOXp8+dLG8OHpehJ7Iuwva0QM4eHzuOX7+SScxvEbgxtF76SXun4ouN6Kn2cyZzI96+un467m511x/vbMBR7ziIn/6E9s+nHee/fudkJ1N0fjuuxSRQeEm0uL224FHH7VfL1VDN99+m20w3IZtphLRjt7SpXTKBg1K7j45QVTe1EIv82htBVavZuGpWbM4mfmXvyR7r7zzv//R1X/sMeDWW5n2kOyJUC9ooZcGDBxIsRLt6NkOgsvLOcrOyuJreXnc1dyGtamYZQfUOXpOaWmJHC4n/3Jzgf/+1367qTpQSoSYTJDp6AHuCrIIcSHL0TMMXi9uhV6fPt4mCZy0VgAikywyhR4AjBsHPP4482h+/ONIGIoboffmm3wIPvOM3H0TZGczL+rQQ4GzzqKjFou419hVVNy9G7j55sS5idHEOlSrVwPPPsty+3l5zvffjnPP5fn7xz/K26YdTotkmSaP11//ar9uqoZuLlpE59JP4ZxkI4ReezuFayq3VYhmn314X9R5epnH6tXM9Z48mdffxRezvU8YRX9FBZ8/H38MPPUUHf4gizHJJOD0Yo0KDKNz5c3aWuYYJKS8HCgtjTydN2zgzwCrIUTRpw8HUk4dPRUJ/2J7PXsC/frJ3a5TVq3iYTrzTPucj7Y2zv4sXcrCC1ZED5SS6Sw6Ze1a7rMQIX4RQs+No7d1K19lCT2AExpuQzft3LzyinKULS5DdWM1RvYbiTlT56BkcsnXhVzsQjd79OAkjqzQzWimTWMZ+dmz2WD9d79zF7pZVcXXBQuACy+Uv38Ar/mFC5nrMXMmxd64cZHf72LF0FYAACAASURBVNrFc9Hu4fvFFxSzYp+tiHX0br+d951LL/X2NyQiN5du5Q038N5SXCx3+/FwWiSrro6VlquqKPqsjm+qhm4uWsQwx1SuUGlH//6cYHzvPRZaC0N+HsDzZeLEcA7uNf5YuZKvos9oaSmfLXffnbz+oV546SXg9NNZx2HpUuDgg5O9R/7QQi9NGD2asykAhZ6lo1dW1vXJ3NzM5TFCzzDcNU1XFbqZl0cHJFkzKuKhdf31NiJ6L/ffzwGcHak6UErEunWcVJD1PfTpQzHjxtETQk9mEZ38fHeOXk2NtdArryhH6cJSNLfxi93QuAGlCzmZsquG15idowc4a5rulcsu46zl73/PvBoRcuXE0ROiafFiug6q8piGDOGg/dvfZhjeO+9EQlqdhiKKff38c842W+XZRV+PmzcD//wnK4DKCoWP5uKLOSF0223AAw94387WrUBjI/NWrf699ZYzV1Icr6++4jVhNamTin30Nm3igPMPf0j2nvhDXFMLF/I1LI4ewHvJ00/bTxRo5FNfz+vRa0VoP1RUcKJw//358+DBHFL+85+MEhg4MPh9csvdd3MC9IADeO0l4zjKRodupgnC0duxgxe55cCkutrVcjeDYFWO3rXX8gJMFpWVDEVwWsGtuNid0Nu50/u+qcI0gS+/pGtz3XXAscdypktWfp6gqMidoyc7dBPwFrpp5ciVLS77WuQJmtuaUba47GtHz0kY8rBh6oSeYQD33EPn49xzI+GRdkKvpYXO7rRpdK+ff17N/gnGjAFeeIHfz4knRvIKnYYiCuHS1hYJb09EtHC5+26+58orve+7FYMHs+3Fww9Hwnnd0NbGWechQ3hf+uY3gaOOYs+3H/2IQvLaayl4yst5Pc+cab/d6PuW3T0sFUM3X3mFr9OnJ3c//BIt9EaNCteAc9IkCg4ZrWHa24MtWhR2vv99Cu1khM6uXMmJ8OiCT5ddxvvp/PnB748b9uxh5MZll/E5s2xZuK45K7TQSxNGj+YATPQwsRR6I0e6Wu7G0VOVo3foocAJJ8jdphsqKxk25rRiXXFxJPTJCiGahBubTBoaWJL897/nw2L4cJ4Sp5xC12HbNhakuOkmuZ9bWOjN0ZM5O5ifz3PcieDcs4duj9VDoLox/qRJdWM1amo4OHcSIqnS0QO4D089xWtchCfa7dfnnzMU8uyzKXYXLFC3f4JDDmFC/IcfsgVDW5s7R0+4Cnbhm0I41tayyfysWd7K8zvliis4kHVbsKC1lSLv8ccpRB96iA7K4sUsxf/ppzxHd+7ksWpo4Ll9yy32266qArp1i/zfiuxsni+vvcY8slQYkC9axOtGhI+FFSH0KivD5eYBcguynHoqv8vPP/e/rXRn+XJO2O3cSbHidNwmi4qKrtfdAQdwAupvf+OzMwh27+ZxWLWK4wW7dgiNjZwE+9vfeD9dsEBOe7JUQQu9NEEUx3jvPb5ahrXNmdN1Kjw3l8vjUFCQfEcv2VRWusujKS5m6JPdjXbSJIrH99/3t39+uO8+DmYHDWJ43G9+w4fqtGl0Nd59lzki77/Pwe83viH384Wj53SQuHUrRZ7MBsbTp3PAvd9+dEOsZkO3bOGDw0rojcyOr0JHZg901FpBMHQoB+wqB9D5+azEKQSenaMnBv/FxcDJJ3NgHYQjfeKJzPNYtIjtF5zmnFVVMfRT/N8KIRzvuosTG1dd5W+f7Rg1CjjtNP5djY3O3tPayvc8/TQF4m23UXT/4AesEHfwwcDYsTx3cnPdh86tWsVrvE8fZ3mNF1/Me8MRR9B9vf765E1ctbfT0Tv++PCHDEaHQ4clP08gWiz4dZVaWujmf/IJJ3tfftn/vqUzd9/N6/a11xj58v3vB+e2NzQwMmHy5K6/u/xyBow9+2ww+/LrXwNTpnDCYcgQjrHy8/nz0Ufz/nnJJcBvf0txN2UKj9m8ebyfiomudEELvTRBOENC6Fk6eiUlPKMLC/k0LCzkzzH5eQIRuumkSaSqHL1ksnMnw2LFLKUThCi0C33q0YMzXk4qdKrizjv5euutwOuvc4BbVUWX4NJLgcMOs69s6IfCQj6MnDbYrauTG7YJsJH4unWMzX/mGT6sRAGQWJy0VpjzGpDb2nlZbiuXO2mWLhg6lIOdpiZn63vlgAMY3jdoUKRATiKEQzZuHN3eXbuCG4BdeCGFxAMPUPDZOXodHXS3DjuMx9xOgIj71rJlHFwfdpic/bbiqqs4kXLvvfbrtrbS4Xj2WQ7qLrtM/v5UVfFeN368M6H35z9zQuuhh5ibc/PNDN86+GDeW1Q60rGsWMEJtjC3VRCEWegVFHAyzq+jt3w5z/m77mJ0yQknsN1JKjjHqcaWLYx6OO88OmiPPMLr4eyzg2nwLaLJ4jnpM2dyQveuu9TvR309J6RnzmSLmLvuYvmJWbN4T2tv577++9/AjTdyjFNTw+eJqsJiScc0zVD+O+igg0xNhF27TBMwzVGj+Lpunbxt33knt1lXZ71eR4dp3n8/112zRt7nJ5sVK/g3Pfmk8/ds3sz33H23/bqlpabZvz+PX9A0NZmmYZjmjTcG/9mCZ57hsVqxwtn6U6ea5uGHq9uf+nrTvOkm0xw0iPt15JGm+cILke/nySe5/KOPLDZiGObDk2EWzoZp3MDXhyfDNA3DLCgwzZ/8xNm+PPwwP2v1at9/liOcnIOnn26aRUX8f1ubaQ4ebJpnnaV2v6Lp6DDNc8/lcTnmGOt1163jevPmcd1DD7Ve/8svuT5gms8/L22XbTn2WNMcNsw0W1oSr9PSYpozZ3Lf/vpXNfuxbRu3f+utpnn22dwnt2zaZJp33GGa3/wmt5WVZZrTp5vmv/5lmtu3y9/naG64gZ+3davazwmCTZt4/AoKkvNs8Mt3vuP/Pn3TTXw+NTSY5o4dpnnqqTwmZ51lms3NcvYzXbjxRh6bTz+NLLvjDi67+mr1n3/XXfysmpr4v7/tNv7+ww/V7sf11/NzKivt121rM80tW9Tfl4IAwPtmAr2kHb00oWdPzliLFgsyKxKKqmsXXsgZ/OnTGabzf//HGdxhw9gaoFs3FhcAwl3WOhYxK+nG0SsoYHNqJwVZDj6YLppdoQgV/Pe/HNYeckjwny0QDtLrrzubedy6Ve75HcvAgXSNNmxgaNz69cD3vsfz/dFHI3l8luGXI0eipAJYfyfQcSNfSyqAtn1Ho7bWXegmEJwr4iTcbfXqSOXZ7Gw2Xn/++UgfPtUYBgMQzjzTvk+acKTGj+c+r15t7QYIR6+4GJgxQ87+OuHqqxmim6CdKXbv5r134UIW0LnkEjX7ER2WO2EC98lpSKlg2DDgF7/gvaWykgVhVq9mz8aCAoZGv/SSmnydl19miF8YGovbIRy973wnnGGoosWCH/dt6VI6RAMGsGLsY48xw+TRR5m3+OWX8vY3zLS20sWaMYNh24LZs4GLLmK/zvvuU7sPFRW87hIVGbvgAt5fVTZQb2pipMPJJztLtcnO5j0pnfLx4qGFXhohwjfz8uQ29z3oINruH33EMKjGRoYcFhZSIMycyYu4rIzJ/o89pnYgHjSVlcxfclOUwTA4UHIi9ESvvWTk6S1fztdkCr1x4zgQv+YaThzcfjvj/ROhInQzHnl5DI1bswZ48EEWtTjrLO5njx42g8kEebBbrvwTTNNd6CagppeeF9rbeQ+IbjEyaxYfsK+/Htx+5OQwNOnXv7ZeTwiXCRN4jjU1WR/Lfv04gLz1ViArwKfjscdyIuG227pOdrS08Bi/8AIHcxddpG4/xP1qwoTIQMlJ+GYiiot5Kaxdy3DYH/8YePFFhuCNGMHcnRUr5ITi1dfzfhb2apuCnj0ZdhfWcLKJEzlW8FJRFuD99j//6Ry2ahjAr37FnOLPPuMk6VtvydnfMPPkk5wMjA3lNgyGLh5/PO8br76qbh9WrmTKQ6JJif79gXPO4X1bRjXWeMydy0nzX/1KzfZDSyKrL9X/6dDNrvz4x7SsRViVRg4nnGCakye7f99PfmKaQ4bYr7d7t2l2726av/yl+8/wyymnmObo0cF/biytrab52GMMkwRMs2dP07zgAtP84IPO63V08Fhdc03w+9jebppPP22a3/62aZ54ooM3PPywaRYWMvaosNA0H37YfPddd2GBW7dy/T//2ceOS2TNGu7PffdFlrW0mGbfvs7DUYPkwgsZWmqapvnaa9z3115L7j4loryc+/fcc5Flu3aZ5owZXH7vver34corTbNHD9Pcs8c0P/uMn/vAA3I/o6WF19Epp/CzANMcO5Zhen5C/v/9b27rnXfk7avGO2++ye9j0SJv73/nHeuUiVWrTHP//U0zJ8c0//537/uZDhx2GK+h9vb4v29s5Bimb1/T/OQT+Z/f3m6aeXmmedll1uutWsXv9He/k78Pzc2mmZ9vmtOmyd92GIAO3cwMhKOnorlvJlNZ6S5sU1BcTPdJ9H1LRPfuwIEHJqcgy4oVDHVKNjk5rIS1dCnw8cecyX70UfYGmzKF/29tZQ+11tbkOMZZWaxs+J//RJoYW1JSwjjPjg6+lpR8Pbvt1NEbOJDHJsiCFlZEO2SCHj1YEfOZZ4Irn+2UqqrIvorXVGhlEo9TT2WUxB//yJ9bWhiC9NJLDFUtLVW/D1VVdNi7dWNF0O7dnUUluKFHD15HwoWYP5/Xw29+w6qdU6YwPFW0UXHKokUM8UtmdIImgt8WC0uX8jVRa4kJE+jgTp3KKrwXXcRnQ6axfDmL8F16aeIohL59GV6fm8s0BNltF9atY9G6eBU3o5kwgdW877lH/nd1//10CsvK5G43HdBCL40QLRacDILLK8pRdGcRsm7MQtGdRSivSJAckuHs2MFcLa9CD3AW+nTQQRR6QVTHEmzZwpLHqTYwOuAAViCsqWFFv7o6hkyOHMmwSSCY0E0ViGbpTnP0DEN9Lz03xBN6AEMLt25NvTCqqiqGbALMHXHaMsAJTu+hTtfLyWFfvbfeYhjsSScx52z+/ODC91atity3srOZ7yPreMWjf3+G/b/xBu+zt97KcL9LLqH4O+ccZyX6TZNCb9q09CuNHlYGD2b+k9cWC0uX8tq1mrju358C5ppr2KJk6lROxFVXM/QzExAtFc45x3q9kSM5QVlby3uLqJAuA6uKm7FcfjnD5596St7nt7ZygmzKlPD1nAwCLfTSCKeOXnlFOUoXlmJD4waYMLGhcQNKF5ZqsRcHMcjxI/ScFmRpamI+WFCsWMHXVHD04tG/P5PJV6/mIO7QQ/kwB8LrWtfUcADtRqiKXnqpQFUVj31ss/rjj2dOURDN051SV8e8LSFKRd6sDEfP6T3U7b32ggt4bGfMYD7N/PmRAleq2bmTYiu6iEFxsVqhF42YyKmoYD74z35G12/yZLoQb76ZOJdv5UpOhqRDW4V0QhRkcUt7Oyc8nLSV6NaNEwSPPMLJ0ilT6Iz36MF750EHsZ/cRRcBv/898I9/sNdiVVX42zREt1To29d+/YMP5nFavlxu24WVK3l/dTJOOv545uLLLMpSXs7CPGVl4SxcpBot9NIIp45e2eIyNLd17qLZ3NaMssXa847FS8VNwT77sJpTqhZkWb6cD0nZDdBlk5XFAgvPPUchPH8+i1eEkU2b6Cy5KfSRao7ewH3/18WhysvjA3zBgmBdaSuEoIt2HydMkCNcnN5D3d5r8/I4493Wxl6B55/vf1+d8umnHPjGHq916+TO/tthGAxlv+suOjO/+x0npY4+mn0Nn3ySQiAa0cdx2rTg9lNjz6RJfIa6vSesXEln97vfdf6eM89kgZbnn2dEyPXXM6S8oIATGE88wWXnn8/nSXExcMcd7vbLimRESd17L+8Vl17q/D0/+AGLPj35pLyiJRUVDLt2UgQwKwv4+c8Zbir6PvuhvZ1C/xvf0BM9idBCL40YPpwXsd3Drrqx2tXyTKaykjODY8a4f69h8GHiROhNnMjPCTJPb8UKPohlVmj9mvJylmrNyuJrorrxLhk1ig5HTo6UzQXOpk3OwzYFw4alhtAzTeDjT3bji6yFcR2qWbPoWAqnONnECzMdP57fgduWAbE4vYd6udf++tccmJ57rufd80R0awXBhAn83j/7LNh9EQwaFDkef/87m6GfeirzCOfOjQjQRYsYNmaV++pmIK5TG+Qcg4kT6RRXuxxa2OXnJWKffej+lpayGfb8+azy+vHHDC3ftYuThUuXssrtww+7234ikhEl1drKa+KEE+iQueEXv6Bj/oc/AP/8p/99qaiwz8+L5txz6UDKaKD+1FO8P/3qV2rdvDDfE7TQSyOysoCnn2acuhUj+410tTyTqazk4NBr3odToZeTw1nsoBw906SjpyQ/r7ycT9oNG/hBGzbwZ0liL8zU1DgvxCIYOpRhiMkudFJbCzRv74E9gyo6LRcO1cyZDEtNlfDNqioWH9h338gyIfo+/dTftp3eQ73ca7OyOu9zUKxaxftcdBsZcbxkF2RxS69ewE9/Spf2qacoAC++mOGev/kNw/ysZvPdDMR1aoO8YzBpEl/d5uktXcpJPdnXQc+eTHE58kiGLn70kZxJjGRESYliRj//ufv3GgZz+w4+mO2M/LBrF/D5587y8wR9+tBZffzxSN66F0wTuPlmjtFmzfK+HTvCfk/QQi8DmTN1DnJzOvf4ys3JxZypc5K0R6mL14qbguJi5ldt22a/7sEHAx98EEzo25o1nB1Xkp9XVgY0d37ooblZl8OCN0dv6FA+0FT1HnLK1yGPg7vGPlY3VqN/f04yLViQGrkvooJkdJisKMziN3zT6T00TPfaqio6A927R5aNHcvjF1Senh3dunFA9+67wJIlwLe+xdDOtjbr/nluBuI6tUHeMRDusJs8PdOk0HMTtumFU0/l6+OP+99WMqKk7rqL16fXcOXsbB6DlSv95YCvWsUxixtHD2C4aXs7nXmvCLf22mvV9j0N+z1BC70MpGRyCebNnIfCfoUwYKCwXyHmzZyHksklyd61lGL7doac+BV6gPPKm9u3c3ZMNUoLsVRXo3wyUDQbyLqBr+WT4T5+J83YuZMhg14cPSD54Ztfn8NDup7MwqGaNQv44gvvlfZkEt1aQTBmDN1zvwVZnN5Dw3SvXbWq6/ESDkiqCD2BYbBQx8KFPNfmzweOOirx+m4G4jq1Qd4x6N+fE1tu7gdVVQyzdFKIxQ/77MPCLY895n9bXzv0DaOB8ueBnYM6L5eMyG/7+c/9CRwhEv00Ul+5kq9uhd6YMcyhvPdetpJxi2kCc+aw8M5ZZ7l/vxvCfk/QQi9DKZlcgvWz16Pjhg6sn70+JQceyUaEK8kQek4rbwLBhG8uX85wKD9/WyLKvzsQpTOBDf0B0+Br6Uwuz2TctlYQDBvG11QQej1z29BrUEOn5dEO1UkncRAus3S2F0ReUKxwyc6mayVDuDi9h4bhXtvaSoEenZ8nkFXARhUTJzJv12rA6yaEVqc2yD0GoiCLU0R+nmqhBwCnn04R6jc0+Wvn/t3Lgc+/B1TNUurcO22pYMcBB7BYjShm5IWKCo4lvNQx+MUvmJZw8smMMHLDkiXAO+8AV1+tPmc/7PcELfQ0mgT4qbgpKCzkTdDJg6S4mDPoQRRkWb6czcizs+Vvu+xYoLl752XN3bk8k3HbLF2QSo7epOIc3Pf9xA5VQQHzX5Kdpydy8ESoZjTjx6e2cEkGn3/OMKpYYQxw2WefJT9H1A9uQmjDFG6rCpnHYOJEXm+xlVITsWQJ75GiXZRKfvhDTkz5Dd8smVyCvx53P7IqqLx6rf+hMud+yxbu73nnUez5ISsLOO44tpvwmjKyciW/Yy91DI4+mqGbixczusjNhMDNN/N5E0Rl4rDfE7TQ02gSUFlJ4SXaVnghK4sDSydCLzublcBUO3ptbcwFVNU/r3pPg6vlmYJw9NwKvYICvia7l54IhbRzqGbN4ixvECHIiUjU2F0sW7OGLpaGiPtTPEevuJj3jCB7fMrGTQhtmMJtVSHzGEyaxNC8tWvt143OzwuiH9qwYXQOH3/cf15xj8/PQMeufhQ966fh1HFqzhcvLRWsmD6dobIffeTt/RUV7gqxxPKznwFvvMG0lW99y9kk4YoVDDe98kqO0VQT9nuCFnoaTQIqKzko9FpxU+C08ibA8M0PP3Q+++mFyko+eFUJvbCHOajCa+hmr15Av37JdfSamuhIxnPIYjn5ZL4m09WrquJ1G6/s+PjxvL6++CL4/UpVqqo4sB43ruvvhFgOuwvqJoQ2DOG2qilZCay/E+i4ka8lK71tR0TEOHFr1q7lfTKIsE3B6afz3PabVzx/PieFf/97YMcOVoKVTWsrHTAvLRUScdxxfPUSvvm//7FImNv8vFimTGEkU3ExcMop7Hdo5TDefDMwYABFYlCE+Z6ghZ5GkwC/FTcFxcXMF9q+3X7dgw7iQ0Jl36rly/mqpLUCwh/moIqaGvYs9BJuk+xeevGajydi5EhOWCRT6K1ezZyR7t27/k78DX4LsqQTq1ax3WVubtffyapUmnQU9fZMSyS2yBEusRMhFWR+nmDWLJ4SfoqyfPEFXakLLgCOPZb3nRdflLePgieeoLi67DJ52ywoYCTRK6+4f2/F3k47fhw9wYgRDNs9/3yK5e9/P3618k8+AZ55Brj8cv+hq5mCFnoaTRwaG4GNG+UJPcDZwFIUZFGZp7d8OTBwoLociLCHOahCtFbwEpI0dGhyhZ5VKGQ8Zs3iefbll+r2yYqqqsTuo3CtfAuXNBIOVVXxwzYBNjYeMSL5vfR8oXt7ukNii5zevXl5OHH0liwBBg92fp+RQUEBc8X8hG8+8ABvA+eey7/3yCOBl16SupsA2FJh3LiICyeL6dOBt9/mJLMbvFbcTETPnnRG//pXOoyHHtr1Pn3LLTzGXvoHZipa6Gk0cZBRcVPgpvLm+PEM1VOZp7d8OW+gKnMgwhzmoAovzdIFQ4cmN0dv9WrmkDqtrHbKKXx95hl1+5SIPXuYH5hosNi7Nxsx+xJ6boRDigvC9nYWr7EaXBcXh9zR07093ZGoFY7HFjkTJzp39L7znWDy86I57TTeM7zkqe3ZA/zjHwynFGH5J5zA5/2GDfL28b33+Oy+9FL5PeOmTWPe3xtvuHtfRQWfTUOGyNsXwwAuuYQFWhobgcMOA559lr9bswb497+Biy7iZLXGGVroaTRxELOPiWa53TB6NEM5nBZk+cY31Am9nTv5t6nKz9MkxkuzdEGyQzdFM22nZazHjuXgLhltFtas4aDFSrhMmOAzdNOpcAiBk7RuHbB7t/W9Thwvr5X5ko5k4ZL2jEyQT51ouQ2TJnEyoa0t8TpffslzMciwTcGsWczp9VJ988UXeW/+yU8iy2bM4KtMV09WS4V4TJnCsG234ZsrV8pz82L5znc4Dho3DvjBD4Df/pZuXk4O2zJonKOFniZjKa8oR9GdRci6MQtFdxahvCIy+KqspLPmp+KmIDubNyunoU8HHaSuIMsHH3Cwpio/TxMf06TQ8+Po7dzJ0Bqr8zYWN+taEa/5uB2zZgHLljFZ3wmy9tVJPqFv4eJUOITASRL3JbvjtXMnw9lDiWThkvbMmdM1YTM3l8s9MHEiRZ5VAaRk5OcJBg8Gpk5lnp7b8M3583l/PuGEyLLx42neyxJ6mzdThJ5/vpq8tB49GL7qpiDLnj28d8jIz0vEvvvyGXLOOcCNNwL33888SNFbVuMMLfQ0GUl5RTlKF5ZiQ+MGmDCxoXEDSheWfj24FBU3ZYVIuK282dyspliE6kIsmvg0NNA18SP0AODeJc9anrfR2J3jTmltpUvmReh1dADPPWe/rqx9BSIhhvEqSArGj6dwEb0NXeNUOITASXIq9KLXDR2ShUvaU1ICzJvHRrCGwdd587jcA5Mm8dUqfHPpUlYXVikcrDj9dDqKbvLja2qAF15gbl50tINh0NVbvJj3fb/Mm0dhJaulQjymTWP46rp1ztb/4gtW71bl6Al69mRo7F138Ty6+mq1n5eOaKGnyUjKFpehua3zTHtzWzPKFnOmXVbFTUFxMW+gu3bZr6uyIMuKFXxmi95smmAQgsJr6KYQen9a9LDleRuN3TnuFKtm2lYceCAdcSfVN2XtK0ChN3w4B42J8N0ywKlwCIGTFMjxSjZuhYvCvMpVq5LfE9MRJSXA+vWcrVm/3rPIAzixkpVlXZBl6VLgiCP8tzPyysknU6y5qb750EM8PPGads+YwcmkZcv87Vd7O4u9HHsssN9+/rZlxfTpfHUavimz4qYdhsHiKxUVvHSTQornWluhhZ5E1q6lY+L0X2xEjyY4qhvjz6hXN1Zj2zaG2ckWeqbJPAU7xo1jGX4VeXrLl2s3Lxl4bZYuEKEq//tf/CoF8c5nq3PcDW4rbgoMg0VZXnstfplsJ/vkdl8BZ2GmvlsGOBUOIXCSVq2yz0UeMgQYNCjEQg9wLlwU5lW2tbEZeJD9v4LALuy6Vy8Wckrk6NXWMoIlGWGbggEDWM3SafXNjg4KsKOOit/T7phj5LRZeO01BgBE5wCqYOxY3sachm+uXElRHmSFVNk4ThcIQa61FVroSeTGG1khyOm/I45Q2xhbkxirpt5i1lG20AOchT5168aCLLIdvbo6uoq6EEvweG2WLhCO3oA98Z+q8c5nWY3rnYRCJmLWLA5uX3jBej1Z+2qaHDDaDT7y8zmw8xUe7UQ4SA6Bk41pOhPGhsF1Qi30nKIwr3LxYmDrVg7eW1p8by4lcBp2PXFiYkdP5Od997uKd9aG00+nqHrvPft1lyxhSHsiAZaXx7/Hb57e/fdzkuWkk/xtxw7DYPjm4sUME7WjooICt2dPtfulClfpAiHItbZCCz2JXHEFBzRO/t1yCwtuPPBAsvc6M7Fq6q1C6O23H4uyuC3I4uSG65QVK/iqhV7wxc4uawAAIABJREFUiNBNr0nkgwZxAuCIgaci1+jcBTzX6B63Gb2sxvVVVYw0zMtzv9+HHca/2S58U9a+btoEbN+euIeeIFDhIjEETjZffsnwMifVhSdM4P3La6+x0KAwr1JUdWxu9h/Slyo4DbueOJFh4PFy1pYupdH9zW+q3FN7TjqJLpyT6pvz5wP9+3MyKxEnnMDJJKd5b7Fs3coWNT/6EQumqGb6dKCpyZnQXbkyefmUMnCVLhCCXGsrtNCTyIEH8sJ28u+aa1jS9te/5sAkE2lp8VEMwSdWTb0rK/nQkRkL3r07xZ6bgiy7dsktyLJiBcPLDzpI3jY1zti0iZXdvD6ss7KYVzn4816Y95yJwm2AYQKF24B5z5koWdn1PSWTSzBvwDko3NGN6+7ohnkDznHd09BLxc3o/T75ZM5qWzXjlbmvgLP9HT8+QxwqC8Tf71ToNTQwMiCtUZRX2doKPP00w5m7d1fTUDsZVDfGbxYXu3zSpEjPxliWLgUOP9x5+xZV9OsHHH88hZ5VRd6GBraO+dGPGJaaCL9tFh5+mBERF1zg7f1uOeYY3rPtwje3b6d4VV2IRSWu0gVCkGtthRZ6ScIwgDvuYGz6rbcme2+Sw69+xVm+nTuT8/klK4H1dwIdN/JVDJZFzorspqRuKm8KMSYzT2/5cg7WeveWt02NM/w0SxcMGwZsWV6Nkv+2dT5v/9sWP4SkvBwlv3wI629r57q3taPklw+5yivo6LBvpm3H6adz0mLhQouVJOwr4E7oTZjA+29Dg6uPSCucVNwUCDGY9uJYUV7lq68yV/W88+SE9KUKI3fEr54Su1xEyMTm6X31Fd2hZObnRXPaabxf/+c/idcpL6czaZc3N3YsC1J5+a5Nk2GbhxwSnKAaMIBRGHYFWcR3mIqOntO8O1fpAiHItbZCC70kcuihjOK5/XbmdmYSHR2sbtXY6D9Z2RMWybWyK24KiotZkthJueWxYynIZAk906TQ02GbycFPDz3B0KHAlpYEpRHjhZBIyCuorqZI8yP0jjiCf/u//22xkqQciKoqzsqLnEYrRHinijYmYaGqik7zkCH264a+8qZTFOVVPv44Q/2OO45Oz+rVjOQNFAWVA+e83I7c1s7Lclu5PJpx45i+EJun99ZbfD4lOz9P8P3vM/IiUfimaQL33ceomwMPtN6WYTCC6/XX3edkrlhBQRWUmyeYPp2fbTUBJipuppqj5ybvzlW6QIrnWtuhhV6SueUWnjfXXZfsPQmW5csjBSqcxMNLJ8HA8qtr/4DNm9UJvfZ25inYIbsgy/r1jPfXQi851NR4L8QiGDoU2Nxtn/i/jBdCIiGvwGvFzWiysujqvfQSZ+9d7ZPLHIjVqyngjPjFSTsh/qZMFnqrVjn/bvfdl3maaS/0AHd5lQ7EU0sLc61OPplhm35D+jyhqHJgSVMh5i1E53DyhVweTffuLN4RK/SWLuXvUuXZ1KcPxdkTT8Qvlvf++xQ6TqtgzpjBoYYoOOOU++9nWOgZZ7h7n1+mTeNpv3hx4nUqKnicktbqIAFu8u6s0nfiksK51nZooZdk9t0X+OUvgUcfBd59N9l7ExwLFjAev6SExWms8neUkGAAWbmRjokqoQe4y9P76CM5BVl0IZbksWcP8L//yXH0ajsGo71XTOxtohASCXkFMoQewMFKWxsHu672yWUOhJt8wqIiztxnhHCJg2k6a60gMAyK6NA2TVeBQ/H0yisscnHaafx53Dief4EKPVWVA+fMQcma3M7h5Gvi35MmTeoaurlkCcMFU6l64+mnA1u20G2MZf583nLPPDPmFwkE/9FH8z7j5rveuZNjwtNOs+5vqYJDDqHzbJWnt3Ilv0vZ6S1+cdump2RyCdbPXo+OGzqwfvZ61znhYSHFvqbM5JprOIj7xS8yoKIZ+DcuWABMncpn4q5d9uXXpZNgAFk58EgAaoTe2LG8MboRei0tcgZWy5fzYZNqoRaZwJYtPOf9OnrDhgEdZha23vagsxASCXkFIrRv8GB/+37IIcxVSRi+KWFft23jsXYq9Lp14zWZqUKvtpYOqxsRX1ycuccrLmVlKB/TjKLZQNYNQNFsoHxMV/H0+OPAwIF85gG8dGfMYEifk1B+KVRXo3wyOu/rZPivHOgirG3iRPYbFnpz+3bggw9SJ2xTcOKJdNNim6fv2AE88ggFWN++Ub+wEPy5uey15yZF5YkneGyCDtsEGF47dSqFXrzxqGnS0UvF/DzXbXpC3ATdDVropQC9e3M88+67XW8s6UhFBfvPzJrFyqPDhiUhfDPBwLLy4LPRu7eaYkq9egGjRyenIMvy5QwFTXZVs0zEb7N0gcg72zLlFGchJBLyCvxU3IzGMOjqLV5MgaFqXwF3+zthQuaGbor7kFNHD+DxqqmhO6UByvtuQOlMYEN/wDT4WjqTywW7dgHPPsvnXfT9d8YMOjdBtVko/+7A+Pv63YH+N+4wrG3SpEjvRgB45x2GR6ZKIRZBXh7F3lNPdY6oeeIJir0uYZs2bumMGcBnn1HkOuH++zkJdcQR3v8GP0yfDmzcGP/eWFPDCaJUnDSeM3WO4/ZDYW+C7gYt9FKEc84B/u//6O7t2qX+87ZtA773va5hFEGwYAHHciedxFn1H/6Qs12BtplIMLCs3DMexcXOcny84Kby5v77Mw7eb57enj3chg7bTA5+m6ULhNDbvNnFm3zkFYgBmV1POqeceSYHdU89lWAFnzkQYlDiZn8nTGCZ8HRpXu0Gr8IYyFxxHEvZ9G5o7jyuRHN3LhcsWkRxIMI2BcccE2ybhbJjEX9fjw3m84FIpIzI01u6lGOAb387uH1wyumnc1JqyZLIsvnzeX85/PCYlW1yjE84gT86+a4//ZQho+efr24cYse0aXyNF74pCrGkoqNXshKO2w+FvQm6G7TQSxG6dWP1zepq4C9/Uf95jzxCcfWnP6n/rFgWLACOPBLIz+fPp53Ggdbzz8vZ/gkn8CZpS5yBpaqKm4LiYs7stbXZr5uVxQayfh29qirevw45xN92NN4QvSKlOXpb/G3HKXV1rLwmw9EDOJtfXGxTfdMh8UpoV1Vx4DxqlPPtjB/Py99JgaQwE+94rVrFiSQ3ExDiXNB5eqS6d5xqHTHLH3uMoc9HH915nbw8OllBCb3qPfHLKCZaroL99uM1KiaYlyxh5EoqtvyZMYPfkYg2WrWKLRd+8pM4Aswmx3j//YExY5yFbz7wAMeD55zjfd/9UljIPNJ4bRZW7hVNkyYFu0+OKCtz3n4o5E3Q3aCFXgpxzDEs7XvzzSzeoJIHH+TrE0/Q3QuKzz/njNCsWZFlhx/OQbCM8M0VK/jgfPBBtjJwQ309j7tqodfWxtBVJxx8MPDxx86EYSKWL+erdvSSw6ZNfHA7KWFvRdBCT1YhFoEI31y2jGFBXklUQvu19zZi7FjmmDglE1oGlFeUo/Tp8zsfr6fPxxsrtmDCBHeuwZgxDD9M5+PlhpH94pcdFMubm9k/8pRT4p+XM2bwWAbRXsl1/pICsrM5uVJZycil5ctTLz9PkJvL8dhTT/H5e//9PPd//OM4KzvIMT7hBOCNN6wjttragIceYrSVkxYxKpk+HXjzza7RDhUVLCI4YEBSdssaN+It5E3Q3aCFXorxpz/xRvCb36j7jFWrKIh+9CN+VpAhyQsW8PXkkyPLsrKAU0+lQPOb+3HXXZwdzMlhQ3o3iHAS1UIPcD5QOuggJuvHlqR2w/LlrKK1337et6HxTk0N81C7xe8r7Ji8PDowQQk9EZ4nS+gBDIcyTU4weSVRCe2KVXtc7+vYsRQ66Sxcyp67HM1m50ZnzWYrPl1luMrPAzhQD6KAjdOmx8nGrhfXiy9S7MWGbQqktVlwUFTCVd8whUycyOfZ8uVAa6t1fl753ItRdFU2sn5roOiqbJTPvTi4HQW/t/p6ht/+859MNxGRSJ1wkGM8YwbHW9GhoLG8+CInm5NRhCWWadO4v7GVR1eutMnPS2aBEzfiLeRN0N2ghV6KMXYscMkljAUXsdCyeeghDjpvu42hgffdF1y1zwULGEIYe92ddhoFzcKF3re9eTPDZM4/nyL2H/9g+JlTghB6In/ITeVNwF/45ooVPOa+SyFnSIUq2choli4YOtRljp4Pqqr43Nt3X3nbHDuW95xHH/W+jerGOPZHWw/sqd/XdT5hr148lUOZc+bweqxuq++6cFd/tO8ocC30AAp/lULPTdPjZGPXi+uxxygMErlW48dTE/gSeg6LSrjuG6aISZO4i88/T000ZUr89crnXozSmrnY0LudxWN6t6O0Zm6gYu/44zm5duml7ENr2TvPJsf4qKPYQsLqu54/n/d4kdOXTI46imG20eGbra28VyYUeskucOJGvIW8CbobtNBLQX7zG/ZOueIK+QKsvR14+GHeSAoKgAsvZGigjMqOdnz5JWfxosM2Bd/6FrDPPv7CN++9l4VHLr2UvQlbWoB77nH+/spKlkzeJ0FPahnk5fF+4lTojRnDffJakGXXLs7A+c7PS/YNPMTIaJYuGDo02NDN8ePl90o64wxOPjgNX45l5I441mjD/oDZzZP7qFq4KMHF9TiyMc7763igvB6vtWvVFbBx0/Q4FUjUi2vHDrYN+uEPE7v5os3C4sVd2yw4drNcFJVIhb5hYiL1H/9gQY9EIYBla+ehOaZKdHMOlwdFz5508aqrOTl9rI/CNb16MU8zkdDbtImO3rnnugs/74TEydi8PFb9jC7I8umnDC9NWIgl2QVO3Iq3EDdBd4MWeinIwIHADTcAr73mrveKE159lTcUkeh71lmc8LjvPrmfEw/RLDme0BPhm4sWeQvf3L0bmDuXAnb//TkYOfFE4K9/7XrfSURlJZRW3BS4qbyZlcXwTa9C/MMPKe4T5uc5fTA47Bel6YpMR2/YsGCFnsywTYEIY/PaSmbOy+3I7RyJiO6bvQuX8eM5gOno8LY/ScHF9Tjno0Fdj9cWWnleHb2ODhaVUuHyu216nKq88AIn2hKFbQpEm4XoEDlXblbIikqIIh719db5edV5CQrdJFiuitNP5+v55/sPv58xg3UK4tUPeOghXleOCsnFQ8Fk7LRpnCgWUSQiyiyho5cK52KGiDc3aKGXolx8McOcfvlLf4U4YnnoIQrJE0/kz3378kb26KOcgVTJggWczRs7Nv7vTzuNoQHPPed+2489xlLIl10WWXbVVQy3eOghZ9sQQk81xcUMf2h3+Lz6//bOPM7Gsv/jn8vMWCaSLVOYkZJGhBDSYouoVKTU1KOFEel5UvIU9YsypUdZEpUsyUwiW4lCtKekssRIlrGU7MYyzJg5398fnzlmO+fMuc9+Zr7v1+u8zsx97nPf17mve7k+13dr0YI326ys4tctzM8/892h0LPwYHCnXpRSlIwMJjsKN4veyZO0wPtD6MXFMQGTp9k3E47HYcpiFEihfesP8TCwOb23uCI+ntapQCTE8BVWrseEvhMwZVlUgePVce2VKB+VjTjHuURcci7O+N3v/GLlj410XNfN2fJQZc4cTswUVwvNUZkFS9asMEsqcckltG4BruPzYk85VlXOlvuLrl3pGTR4sG+2BRS16okw2+YNN3Ci2iP8YE3r0oXvK1bwfcMG5j9o0MDJF8LsXCwtqNALUaKimJhlyxZann3BsWPAwoWsZ1WuXN7yvn05sPNF2nNnHDzImjmOrHl2WrXi/cCq+6YIk7DExwM33ZS3/PrrKXBef714UXXwIF/+jM+z07AhB5Zpae6t37w5RZ4nNQ/XrKHIuOgiBx9asAq4Uy8q1PjtNyb9sRKn6Wt8VSzdTkwMLd7uWqk9xZOadFa4917ODnuUZCgpCQnbowuk0I463AhxNTKKhGe4g/03hpP7pqXrMSEBCYNnIG1hHGwvGqQtjIPE3IMrroz0yEJxLoHNrLV+sfInfYEiFsjoLC4PF06coDeOK7dNOxUr8lmVf/BvyZoVZkklypTJm0C6/nrn6yXVS0R0oUnu6LNc7ggriVusrBsRAQwYwElxb7nsMgq5wkLvm29o5fMqCcvu3UhpjILXY2N4ZU276irGmNrdNzdu5P2ybFknX0hKQkrzqIJtaB4VsudiaUGFXghz222c7XvhBeDoUe+3N3cuXRwffLDg8jZtKD6mTvV+H8745BNa0l0JPWPovrlsWaGSD8W4B/3wA2PY/v3vgm6XxtAiun17ntuoMwKRiMWOfUZ88+Sv3HJ7sidk8SROb80a526bVqwC7tSLCiVEeD4sWhRc71JfCz27YPe3Vc/XpRUKc9ddPO09ct90EIexpVZHxLfwrBhXOBYBt3w9FnJnSk2/2OO+rVCBVpnPcZFfrPwJXx8pYrGdspjLw4XFi/msLc5t007XrnTnt4/JLVmzwjCpRPv2tOo7zGCZS8KAyZhSawDiTkbwPDgZgSm1BiBhQNHAeyuursFO8tK1a9EyC9OmUUjedZfn2025sarj6/FGzy3hZcrQfXPFCt46NmxwXSg95Sogsbsp2IbuBikhWFy9NKFCL4QxhtaoI0d8MyHy3nsUGc2bF91Pv37ATz95menThSCbP5+DgyZNXG+iiPumG+6Fb7zB8gGO6tv06AHUq0frqKvENoEUeueKDk9c6ZbbU716/H1W4/SOHOEsoTOhZ8UqUFy9qFBj2TLGvDRowAkMlyLZj9lE7cXSfem6CfhB6BU6Bqkf/Y6ICP+V5IiJ4WDvww89TDiVT7jkbE/DHweqeixcqlVjjcNwsuh5cz2ePMnbjTdu6vHxwK9RDf1j5Y+NRcJGFCx6vBFh5f41Zw6v+WuvdW/9wi59Vq1Z4RaXNGYM62kWR8KAyUgbkw3bCEHamGyHIg+w5uoa7CQv3brRo+err/j/sWPAvHn0cvDEI8HO8E5wfD16kUAGoPvmwYMUp3v3ui6tMHzlcIelXEI1kVJpQYVeiNO0KYNz33jDegHw/GzdCqxeTWueo2QjDzxAc3yRpCzuDoJdCLL0dCaW6dGj+EQnLVtyQvKc+2Yxfud79lBE9u3LLFGFiYhg9tKffgK+/975fu0ZN301IHdF5cpArYh92Hy20CjaiT+9MZ4lZLGv70zoWbEK+LUGk4+Fls0GDBvGiYXvvuMg/vHHnQgKqwHsFtvqD9dNwMdCz8ExSF26A5ddmO7cRccH9O7NxAS//ebddnbt4sDJG+vjFVfkWvTCpISIN9fjH3/w3Zvj1bAhcPbo5UBOUVHntZU/zFwRC5OezqRivXq5n7E2Pp461i70rFizwhFjfJvN14qra7CTvNx4I63i9kR7s2fTuudt7bzd2Y4t3s6Wu4s9HOb11/nuyqJnNZFSsOsklhZU6IUBL71EEfbf/3q+jZkzeWO9/37Hn1erBvTsCcyalc+lwMog2IUgW7KECWV69iy+nXb3zeXLc91Vi8niNHkym/bYY863+dBD/H1jxjhfZ9MmWvP8nXHTTsOcjdgMB1PqTn5v8+a0thZOwe2KNWvyvusIK1YBv9Vg8oPQWrCA4mHkSKB6dWD0aE5yONyklQB2D7Ka/fUXx6iVKztdxRJ2oefTWnoOjsGWnPqIP/ajD3dSlB49mEbc29hguyXOm3jC+Hggdd2ZsCkh4s31aM/4661FDznlgGOXFPnMayt/GLoi5ueTT+iVYs/W6A75yyzYk265a81SrLm6BjvJS/nyDMmxi/pp0yie7CEanhJb2bHF29lyd6lZkwYHe3tdWfSstCHYLrSlCRV6YcBFFwHPPMMB7NdfW/9+Tg7w/vs0wTtMypFLv350I5g/P3eBlUGwC0G2YAH326qVe+29+24Kw48/hsssTqdP8/l/++0c8zsjOppC8JNPHMfhiOQJvUDRsNJepCIeNhRSlk5+b4sWPCZWXGvXrOHg15nIsGoV8EsNJh8LrZwc4PnnOYi97z4u69OHluKhQ5kkoQBW0kF7kNXMXlrBVxMINWpQ4/rUolfot55FJLbhMsSf/tWHOylK1aq8J82Z411pA/s17Y2FKj4eOHyyPA5mFLIkBbIGlEU8vR43b6bA9sYt136syx5tWmC5z6z8YeaKmJ85c3gbd/d5Z6drV7rV5i+zoLiHFVdXy26xfqBrV+YOmDePYQWPPOL8GZGyMQV1x9dFmZFlUHd8XaRsdDzxZNnKb8F7oXNnvlep4trryUobrLjQunsMFMeo0AsTnnoKqFOHbohWB0V232p77TxntGvHh/85900rg2AnAiWj9uX47DNmP3TXVaNFC9535s6FSzeelBTGof3nP8Vvc9AgzqTZ3Q/yc+AAa/oEVOj1uhKnUBF7UCdvoQv3JLtVzt2ELCKuE7EAfrTSWcHHQis5mQP/UaPyst2VKQNMnEgrWJHDayUdtMUaQQcOMA7j0ksdf80TIiKYwMCnQq/Qb92Gy5CNKMRXO+DDnTimd28evtWrPd9GaioFcLVqnm/Dbg3cAgdmwRCtR+YpqanM/BcVVfy6zrALvTuqDQvu/SPEOHqU3ii9elmf3OnQgX3irKC24hwrrq6h4BZrj8kcMIDeWs7mMVI2piBxcSJ2pe+CQLArfRcSFyc6FDqWnucWvVPsZRYaN3Z9Xltpg7sutFaOgeIEEQnLV/PmzaW0kZwsAojMnGnte/ffL1K5ssjp08WvO3o097Fli4jExUlyY0jcExDzAt+TG0MkLs5h45KbRxVct3mULHziKwFEvvjCWpuHDhWJjBQ5fFgkefIAiRsSwe0OiZDkyQPEZhNp3FjkqqtEbDb3tvnooyJly4rs21dw+apV/M3Ll1trozd8+y33eWHPbgV+lzNsNpEqVUT69nVv+7t3c/tvvumjBnvAli0iJ08Ws5KVc8wYx+saIyIiZ87way1aOD4nHnxQJCpKZOvWfAudnLeSnOxVW7OzRTp1EilXTmTdumKOgUWaNhW59VYfbrDQMahx650CiPz84lIf7sQxx4+LlC8vMmiQ59u49lqRG27wrh1pabxeHq7Vz71z0V8kJ3N/xvDd0XnoJfXri/Ts6f12Lr5Y5F//8n47JYkZM3gerVnj2fc7dBC58kqfNkkJURo04Llyzz3O14kbFycYgSKvuHFx3u08Lo47L/xycq87c0bkggtEnnrKu90WaMKQCMe/bUhEwfX8dQxKGADWihO9pBa9MOLee2mhefZZ4NQpBys4MMUfP05XzN69adEqjj596NYzdSqQ8t9uSOxeKF1vdy4vsmsnaXVf31YHVau6LozqiLvvBrKzgWfeXI3EozML+nEfnYnnpn+BjRtpzXN35vTJJ+n+OHFiweWBzLhp53f5CABw4ESDAr/L2SyVMbR0umvRs8fntWzpi9ZaZ/ZsHs8OHZycq7lYOseKSR89dSonJpOSHJ8Tr7zCayB/4Vsr6aCttHXUKCYgevPN4jPNWiUmxrcxeoWPwcEMmmvWdz7uu504oVIl4NZbab3Pzrb+fRFaqLyt91enDlCubCber3WFW/3rFzyIAbVKZiZdxnxRNiM+PgQylYZY8pw5c9gMT+Otunbl82jPHmvfS9mYgjpJTWFGqGtbuGC36rlKwmI1uYnbWKy5V64c8OuvwIgR3u02P+660PrtGJQiVOiFEWXKAGPHMu7ntdcKfehkkDBvyGqcPl20dp4zYmJYv2/mTGDYqRWOfagzlxb5nsO0umcFP6yshu7drbsJXX01ywrMmp2FjLMF3fUyzmZg3HgbqlfPi8Nyh/r1gTvuAN56i7EQdjZtYvkCV/GLvmb0L08D5+0HDuZlRMg4m+EyDXHz5kwyUqlS8a/77uMx97XIcIdZs5j058ormfmzVy8KbEcMz1zq/jnmIn30qVNMWnTjjXlZwgoTE8OalEuW8AVYSwftblu/+IKJYB54wPtMao646CLfum4WOQYH44Hz9+Cl1V5kf7JA7950c/Uk/vjgQbrLeStcypQBpMafyD5ScEPOzkW/4EEMqFW2bqXrvzeJWOzYhZ5H5TF8QQCEsRUOH+a1f/fdnsfkFi6z4A4pG1Pw8KgV2Pv8WuDNVOz6rAf6zXlGxV6IM3gwE4V17Oh8ndhIxzXwnC13F09q7l1yCVDRs1KlDnHXhdZfx6A0ERnsBijWaNuWD5L//Y8lBc4FxjoZJLz3fhlcfrm1wPB+/YCFCwGsaQZcWbSmg6OZFIezK2ntYTtd2WWRdGcYw985+tW2QEZVIDpfiuCjdXH6904YPMw9K2V+nn6av236dBbUBgKfcRPIPV41NhcQeueWO2HAAA7S3LV8NG3KmbhAMn06z8v27Zn85oMPOPbq1w+YMaPoMbYyW+cqffSbbwL799N67aofH3+cMaiDBwOdOlncvxvr/v03RXZ8PCcU/HFOxcTwt9psvklRXuR3HYoHqqcGbMa0WzcOID780PWgxxG+LOyeVXU9sLttkeUBmzm2GAPqCb48XvHxnDDbu5cW0YDjShgHIXnLokW8N1vJtlmYhg15LD/7jPdNd3hizI/I+mgaUGsNYGzA8rE4vfJl9F+2BJdMANq0CeyzTXGP2NjiM6knfQEkXltwgjM6C0j6CoAX8z/DOwEZhcYR9knTQF45CQMmIwGuYyP9dQxKEyr0wpDRo/lQGT6cRdABOBwMbEc9fJvZCi8/aO1G37kzb0IH1g/CmSs/KvK5o1S5sZVjsSt9V8GFqT1gyp3ETTd5Ng10993A6NGRQOqdQPNpeR+sGQSUsWHgQOuj3DZtKJbHjQMGDmRyi02bgLvu8qiJHhNbORa7amwGNiQAAtiTb7pKhRwbC7z6amDa5wlTpgD9+/P8WbSItYL69aOb4QsvUKCMHl3wOw7PG1g4xwDULtsIr74K3HIL+9YVZcsC48dz5nzCBN/s375udjatU6dOMQmLo7qOviAmhvs6coTlI7ylwO+yGeDQFcDVU71Oy+0uFSrQ0j5/PjBpEizV7vOlcKlc+x+kb6wLZFWZ+eIPAAAgAElEQVQAyp4+tzxQxwGxsbRKOVruIzZv5rOgQQPvt2W3CqamBknoBUAYW2HOHCZeatbM823Yyyx88AHLLBR3LcyZAxxKHg/EfgckdAPKZgD/NAZ+6Y9T6x9A27ZMoPHoo/SyOP9819uz2Vjb8pdf6I3xyy8U8q1acfKuQwd62qhwDAwJXx8BDgPDOwK7KwOx6UDSSiDhd+9q4/mr5p4/8NcxKE2o62YYcskltEjMnJkvZsvBYOB9/AsGNjzwgLXtR0SwSPuZrTeg/ImCIyhnqXKLpNW1lQG23IlW7Q9ZtrrZadoUqBl7HGVS781bmHke8NsjaN15r8fFzZ9+mhm7582jZeTIkcDG5wE8XlEx24DMC4AT9Bn1WWryIDBpEkVet24si1GhQt5nzz/Pgcarr1Jc5cdKOmZn616dloKjRxkX5w433wx0705XzyGNx3q9f/u6//d/wLffAu+84xvXOGf4upZegd91vDZw9jxE1dwe0HOxd2+6YK5YYe17qalMVlu7tvdt6NMxN6D1cJ4KCug1GYBC4ampfH7kvz49xS6ugxanFxvrOM7IhTBevRoYMoTxRr7k4EFg1Spa87wVQfYyC99/73q9uXNpuCx3yVrgvlso8gAgZiNwyyDUebE1pkxhzP1jj7HMS2Ji3m+32ejKO3s2s3q3a8cQhiuu4Hbfeotis2lTZu5OTGRW7rg4xvLPnFniEtKGHrGxSNgIpI0HbCP5nrARXk/++Kvmnl/w0zEoVTjL0hLqr9KYdTM/6ekiNWow25zNJkUy58X+x0j1sjvlpkZ/e7T93btFypQRuT1xg8SNixMzwkjcuDhJ3uA8C1zyhuRz69YcdJcAInPmePgDcxk2TKRMRI7UfqmZmBFGqvYcLoDIDz94vs2cHGa8uvpqZgP1JCuoLxg2dQWTXT3QqdhjG8qMG8djePvtzM7liOxskR49uN7s2QU/y3/eWDnH4sbFyaQvP5LzznOducwR27YxA+sDD3i3f/u6S5bwt/XrZ60dnvDNN77PEmv/Xbi/iwAiw6cHMAWtiGRmMqvs/fdb+95NN/E69gUbN/K4Vr//cbfOBRGeR/v2uZ/5tzgcZRj2JY0a+S5jqz0TcP/+vtmeVZInD5Do4QUz8UUPh9NjNn8+s+DaEwy2bCkybZobmYHd4J13uE1fZNg9fpzZgZ9+2vk6c+eKRESIXH+9yNQfP5TopOiCxyEp+ty5a7MxC+jDD4tUqMB2Xn65yPnn5x2LcuVEWrUSGThQZPp0kfXrRc6ezdufzSayebPIpEkid90lUq1a3nfr1WM26JQUkQMHvP/9Sj6Sk0WiowtmxoyO9jobb/KGZJfnTEjhp2NQ0oCLrJtBF2yevkq70BMReftt9uCCBbkX7siyeRdunxsFEBnwynceb79bN6bQzn/Dd5cnnuDD4/hxj3cvInxwAnyQ2gVay5beD6ymTOF277iD74VLLgSCf/7hvidMCPy+fcX//sff0KMHB+uuOH2aExNRUSIrVvhm/4MHc8Dzxx/WvztsGNv+/ffetWHXLpGqVUWaNBHJyPBuW+6wdSvb/f77vt+2XbTv3+/7bRdH374iFStaO4Z16ogkJPhm/2fOcHLr+eeLX3f9epEuXfLGHeXLc/B80038HaNGicyaRVG+ezcnOorD34Ovs2eLFxBWufZaio1gYCXt+htvsGJFmzYU5xMnsowBQMEzaBCFvrucPSuyerXIiy/m3dOuuMJ3gr99e4pyR9hF3nXXiZw4wWXuTlYdPcrf3qWLyIABFLrr1olkZVlrX04Or4Hx40W6d2f5JoDCf88ea9tSisFPJVesTHAGnQCUnQl3XAk9w8/DjxYtWsjatWuD3Yygkp1Nt4ozZ4Czj16O3af+zPtw4QwgtQfqvHQNdg/d4tH2Fy1iofNPPmEmTncRYYrpJk34XW8QoStJnTp0ubz55rysjt5w5gxdUA4cAKpWBQ4dCnzcgQhjrHr1At5+O7D79gUvv8w40bvvZqFydzKrHjvGUhs7dzLL4tVXe77/PXuYSTUhAZg2rfj1C3PyJM+tmBjgp5/yCqxbISuLmT43baIbdf361rdhlRMnGGtTqxa9VypVYjITR9lXK1akO1abNkDNmsVvu39/ujQH43pYtYrJWObNA3r2LH79kyf5G0eN8l1iyssv533ro6KhyQAYr/T883Rbq1yZ96Tzz2doXVoa33ft4n0lP5GRjN8aMgR46CHH51rd8XUdxoDGVY5D2hNpXv+2rVsZmzdjhvtZmIujb1+6ah886JvtWaHMyDIQFB2/GBjYXrABoHviM88AY8YwDjQlJc87VoTuke+8QzfIrCzguut4Ddx1V8FEXyI8fl98QffiL78Ejh/nNdKsGTP9PvKI767/MWOAoUPpGpk//nHePLo5t27NhC2VKvlmf96SkwP88AOfzx078pzQOD5FCRzGmF9ExHFhF2cKMNRfatEjy5blzip3fjJvVvPZ8wRRJwTN3hUzwni87awskZo1RW67zdr31q5lm2bM8HjXBXjuOc60t24tEhNTvOXIXV56ie0M1oy0CGdlg7l/Txk5kscuIcG6xfevv0RiY0UuvJCz657Srx/dL9PSPN/GBx/wd7z7rmffHzyY35871/M2eMKIESJ33smi7K1a0ToRG8sZ9cjIPEuT/WW3ZoweTRcsZ5aH668Xads2sL/FTnY27zetW4usWlX8df7zz/xt8+b5rg233ebYknLsmMizz9JyV7asyJAhIocPO9/OqVMiqakin39Ob4Rnn2U/Adz+0qVF+8CMMA4tVN7cw/OzcCH3/+OPPtmciIi8/jq3efCg77bpLsVZ9M6cEbn3XrZv4EDXVtWDB0Vee43F5AFa6J98klbzhx6i5dh+LdWty3vP3Ln++912N+IpU/KWzZtHS17btt57yvgL+/ngKmQjrCxJEn7tDTb//CPSrJnI44+LHDkS7Nb4DpuN45W1a4PdEsdAXTdLNl27ipjy6YKnq/Fhd8e/+FB6qK1DNxYrPPMMRdbeve5/Z9gwPpAOHfJq1+fYsCHvITtihG+2KcKBWsWKIv/+t++2aZXERA4qfOXy429sNgpvQOTBB91zSXNEairjPC69lA8Gq2zdynPM276z2Si2a9SgW5MVFizgcXj8ce/a4GtsNrrJHjggsmMHB/YvvijSvHnedVS/vshTT4l8/XVBoV69usgjjwSv7RMm0A0OEKlUiWL23Xcd33/ef5/rbdrku/0PHUohZz8mmZl0+6tePW9iY+dOz7Zts3Gwfuml3FanTiK//Zb3uRVXRE94+WXuNz3dJ5sTEQpWgC6qgcaVq+uxY3R/BEReecX9+2tODuO1e/XKmzCpUkWkZ0+GSngzMWUFm02kdm2e/yKML4yMpKtsqIo8EV43LVpwEs/R879IiMkISPTIsiErnsKtvcHGZmOsflQUx43Vq3OywtNxQjA5dIj3txEjOMa2x6W2bBnsljnGldBT180SwObNQOOrbCjT8h1k3zwQeG8lkB6LCk81wbvdpyChseeVUbZvZ6atl14Cnnuu+PVFmI2tTh3rGfRcbbNhQ2DHDrqyuOOC5i7bt9N9snJl323TChMmAE88wfprgSzYnh8RlgRIT6c7kqv3HTvojtu3L12evKnj9uOPTNcdH89yBFbckO67j+5BO3Z4fz6sW8di9A8/TDevOnWKT2u+fTu/06ABM21aKQkQTPbuBRYvZh+uWkV3tWrVWJqiQwe69L32GrPwBYsTJ4CVK+matnQp2wwAV13FrK7dutEV9YUXmMk1I8N3x/+99+ha+ccfwPr1wLPPsq87dGDt0ubNvd9HVhZdtUeOZKbRf/2L7qdfH01B4uJEZJzNqw0XHRWNKbcVvYdnZdHdeNUqtjU7O++Vk1P075wcXitly9Ll2VekpTGL5/PP87jVrFk0cag/SdmYguErh2N3+m7EVo5FUscktKuagK5dmQ10+nRYzjptZ/9+ZrZt3Ngzt25vSUxkbcmpU+me3rIl8PnnxZdICDbr1wMtWjC8YsaMgp/VTaqOXdmHi3wnLrIa0oYfcri9tLS8++0FF/ihwS7wpL3BRIT3sN9/B7p0YVhBIGvpzprF+9nrr9OF9/HH+Xxs3hyYOJH37UAgwvOmTBmGlERF8d5nf4+IKOhafOYM8NtvwJo1fP30E885gOs1bMgSI9dcQ7fpJk0C8zus4Mp1U4VeCeGxx4C337Gh+iMP4cCUmajc9XVMejXGK5Fnp2NHnvQzZjDmISfH+evQIV7ckyezwLevWL6cD90+fXy3zVBg/XrGqd1zD+NHAh3XkJPDeJRFi4pft1IlDjLuu4/18HxRrHvJEuD221mjadIkxp0VV39u/XrGpg4b5rvM8489xnMWYB9cdBFjOOvW5Xv+V0wMY3J27mSq8rp1fdOGQHP8OK+rjz9mPxw9yuVLlzLFeyggwvjHpUv5+v57CpfKlTmAqVIF2OJZCLJDfvyRg5HatSkwGzWiwLv5Zt9fm8eOMc51wgQOPJ58Eqh76xyM+vG/BYRLQuMEZGdzILJqFV/ffUeBawzrmpUrx21ERrp+v+UWnuu+wmYDatRgiRo7lSpR8MXEuH6vWRMel95xxqZN7Kv0dNZlvOkm324/kCxcCPTowb/btAkPkWdn+HCe28uXF+yDMiMMxMF1ZASwjSg6Fl23jjHdJ07w//r1KXjtr2bN/DuxYLW9wSQjg2WMZs3itZ6TwxjtLl2AW2/lBNmFF3q2bUcTKoXHln/9xTJVjRox/j4igvfvDz9kbPLff1MEvvpqXokgf/DXX5zc+fJL1+vlF38nT/K5AvDef801fLVqRZEaKrGwrlChVwo4eJA3waws4PRpDkJ9NQCdO5dCxF3KleP+g2WhCjfsSU3eeos36kAybBjwyivAv//NWavKlTmYyP9euTIfGP6a1bZbUexUqcKbrbPX0KEc8O/YwXV9QU4OZx537sxLqGFPrrFnT95DID+LF/MBWhLIzqZ42LCBEzTuJNYJBunpTIjx2WfAsmVM2jJ+vG+3HxPDBE0vvcSJJX9bc9LSeP1/8AEHYiNGMLHHli15wu6bb9g2gAOpDh34uuEG310DnrJnDy0I+/cD//yT957/b/skQmEqV3YtCOvU4eSKOx4XX3/NSaPoaE4KNG3q298ZaI4f58RXo0Y818NF5AG0kDRpApw9C2zcmDd5V3ewwS4HVrm4Y0DauIJj0bQ0CtzISODNN+m5tGYN8PPPHMwDvDavvLKg+KtVi2MQ+8ubCUkr7Q0mO3dyUmD9euDFF1ln+csvgU8/5euvvzgp1Lo1n1m33cbzyp3Jq5SNKUhc+DAyJOvcsmhTFlPunH5O7IlwEunrr9mGyy4ruI2TJzkpO3Ys++T//o9jDl97wnz6Kb1STp/mPi68kOdgVlbB98J/V6zIc+eaa1hvMhxRoVdKeO01ZoFr356DA18hwkKzWVm8sRb3qlKF7pCKe9hsvPmuXMnj7E0mSissWMCBcr9+wJQpgdmnM9at42Bx796ir/37i67/8st0qwsEOTm0JtsF4K5dzNZ5552B2b8SWHbuDLwLIsAB7JAhFHVly/J+C3DQZBd27dr51nU9UGRmMhNpYTHo6N0uaPNz/vkUfLGxeZZ1+9+xsZyk6dOH1s3PP+fyksCePRysBtL9zld88w1dB598kq58AJDSvjoSrz2MjHwD/OgsYMoP1ZDwZZ4r5OHDQNu2PB+++45iLj/79vF6yf/Kb1XOT2RkQeFnf513Hr2PXGWgdbe9wWTZMuDeezlOS0mh5S4/IvQG+PRTTk7ah81xcRx3PPooRZ8z3HFfnTaN4RwTJwKDBjnf1p9/UoQuWcKwhzfeADp3tvqLi5KZyQngN97gBMOHH/IZXZpQoVdKyMykL3///uHtslIaOXyYbiiRkXQH9HcswubNdEu48krOwoXyQCIri24fduGXns5Bna/dvhQl2IgwfnL5cs4ut29PIVOaOHMmLz5u926+du3Ke9+1i26vhbnuOrohV60a+DYrjnn0UeDdd+kS3bIlgJQUpIx7CMOvP4vdlYHYdCDp2ygkDJ7BwQvogtixI8XJF1+wX4tDhB4ea9cyfCQz0/ErKyvv7z//5D6efJLu2Q4t9260N1iI0Bvnueco1BYuZAmX4ti3j0Jr8WLmUYiMpAXc2XEuzn111y7GsbZowf5yx4K6ZAlzE2zbRiv8s8/yfueJe/wff7DkyLp1tBK++mrpHBuo0FOUMGD1arpi3XorrW3+itdLT+dN9dgx1n6rXds/+1EURfEHx48XFIHZ2fRMKI0DvFAmPZ0hAdWrU4RFRYFmp+HD2XmxsfTpyxVN2dl0Qfz0U9YMtMco+oPsbIq8iRMZkzx7thMXYRftDRbHj9MSuXAhrXnvvlt8bLsj/vqLonrvXh7zdu2KruPKfXXnWMFNNzF5ycaN1sKFMjOBceN4OE+epFhMTGQSH3cmukVYz3TQIF73M2ZYq/dc0tA6eooSJowdyxS+Y8cW+iA5WSQujgXR4uL4vzNcrJuTI9K9O0sTfPWV79uvKIqiKHYWLeIzLSnJ9Xo2m0jfvlx30qTAtE2EZTMiI0Xi470vn3H4sMhHH7EGsb/YvFmkQQM+w8eN87400759Ig0bilSoILJ8edHPk9tVk+hhBUu+RA+DJLerJpMns7/eecfz/aensw/s5X8qVBDp00fku++c/7b09Lwame3aWSv/VVKB1tFTlPDAZmPtpMhIke+/z12YnCzJzaMk7gmIeQES9wQkuXmUY7FXzLr2IvHjxwfuNymKoiill169RMqVE9myxfk6L7zAZ9OwYQFr1jlWrWI926pVRb780vr309NFRo4UOf98/oY77hA5c8b5+smTB0jckAg+o4dESPLkAW7tZ8EC1v6tUcOzdjrjwAGRJk3YR0uWFG6s4zHF9rGL5LzzRDp39l0d4LVrRfr3Zw1VgAJ0/HgKaDs//SRSrx6F7ksvhWeNPn/gSuip66aihBjHjjGlb1YWYwiW9XI/INxV8HjVoYdwyy0sjzBrVuBLOSiKoiilj3/+oQtno0asmVo4jmvKFOYWePBB1j4MxrNp2zage3fG7k2aRDfC4sjI4Lqvvso4+zvvZL3PkSNZ1mDBgqJJnVLeGojEv95CRr7MxtFngSm1BiBhwGSH+8nJYa3KV15h2MX8+b4PuThyhIlRNmwAPvqIsXN5jS7ovmp7KQntpyacS6JWp45v23LyJDBnDs+LNWuYQ+Cuu1izc/RoZsb84AMm7FGIxugpSpjx669MLd2+PZB6RRnsrlL0OnWU4tmZP/3FaZciY9E21K3L0gSBziioKIqilF5mzAAefrhoGaFPPqFA6tKFyXSCWdolPZ0xb599xoycY8cyWUlhMjMZF5eURBHbpQswahQTkgAUq337MuZ+8eKCddjqPh2JXRVzimwz7mQE0sYUreOzbx/j1latYhzqxIn+S5527BjrUP7yC4VUr16O15swgclUZsxwnbXUF6xfz2M9axZjE3v25P/BLisTaqjQU5Qw5O23c4vOtx8O3Phykc8dFW11mCErKxqYuhpVzl6FX37hrJiiKIqiBAoRZgNfs4ZZn2vXZgKyjh1p6fvyS88SivianByWqRo3jhauOXPykoNkZwPvv0+L3e7dFHKjRgHXX190Ox9+SIHWogWFo12YWCnCvmIFt3HiBGsJPvywj3+sA44fZ028H36guLrvvoKfb93KGpUdO1KkB8r6mpEBbN/ufv2/0oYroedFKUlFUfxJ//65N9mvXgR2tivyeWxUteKXCYCPpwEHG2H2bBV5iqIoSuAxBnjnHYqlgQOBLVuYYbpWLabbDwWRB7DMwtixwNSpFJ+tWzOF/+zZdD995BHWsly2jG6ojkQewJT/8+cz/KJ9e9aRBIDYU47qOBRcnp1NT8kuXYAaNVgnMBAiD2DNys8+o4i9/35mtrSTk0MLXvnydKsMpOCKjmZmThV51lGhpyghiv3BeHHdk8D82cCJmHOfRZuySOo+och3krpPQLTJF6C3ejCwqTfufnwDunQJRKsVRVEUpSiXXgq8+CLdGdu0oVvksmUUM6HGI4+wLtzhw0B8PCddy5cHFi1iOYHOnYsXHbffTqvX1q0sHv/330BSvUREny24XvRZLgeAPXtY5uDll9mGNWuKFoz3NxUrUnx36gQ89BBdJQEWvl+9mtbFiy4KbJsUz1HXTUUJcX7/HWjeMhum1lpk3nc94qrWQlLHJCQ0dlzLJ2VjCoavHI5d6+oBs5ajZfu/8dOKWJ0JUxRFUYJKdjZF3pYttIg1bx7sFrlm504mQrnlFuCee9wrCF6Yb77h9y+8EFi5Evj+s4EYvmMKdp+Xg9hTEUiql4iEAZOxeDEtZllZnOQt7DYZaM6cYUzc0qXA0KHA+PH8HfPnq2Ut1NAYPUUJc2bO5ANg4ECgW7fi18/MpOtnjRqcfcwfDK4oiqIoweL4cSY+8XW2xlBmzRq6YlaqREvh5ZfnfZaVBTzzDOMCmzVjXGD9+sFra34yMylwP/6Yhe83baJgVUILFXqKUgLo149xA+5y/vl8uDRo4L82KYqiKIpSPOvXMyFNmTIUe40aATt2MJ7v55+Z6XPMGP9l1fSUs2eZgKZTJ7qVKqGHCj1FKQGI8EFx9mzx6wJA3bqhGfugKIqiKKWR1FQKpjNnaMUbNYrCb9o0oEePYLdOCVdcCT0HFUIURQlFjGFaY0VRFEVRwo/4eMbsdezIuLdWrViKoW7dYLdMKamo0FMURVEURVGUAHDppaxT9/nnwAMPBLdIvFLyCZnyCsaYm40xfxhjthljngl2exRFURRFURTF11x8MWvjqchT/E1ICD1jTASASQC6AmgI4F5jTMPgtkpRFEVRFEVRFCU8CQmhB+AaANtEZIeIZAH4EMDtQW6ToiiKoiiKoihKWBIqQq8WgD35/t+bu6wAxphEY8xaY8zagwcPBqxxiqIoiqIoiqIo4USoCD3jYFmRug8iMkVEWohIixqaN15RFEVRFEVRFMUhoSL09gKok+//2gD+DlJbFEVRFEVRFEVRwppQEXo/A6hvjLnEGFMWQG8AnwS5TYqiKIqiKIqiKGFJSNTRE5FsY8wgAMsARACYLiKbgtwsRVEURVEURVGUsCQkhB4AiMhSAEuD3Q5FURRFURRFUZRwJ1RcNxVFURRFURRFURQfoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYShQk9RFEVRFEVRFKWEoUJPURRFURRFURSlhKFCT1EURVEURVEUpYRhRCTYbfAIY8xBALu82ER1AId81BwlMGifhR/aZ+GF9lf4oX0WfmifhR/aZ+FFaeuvOBGp4eiDsBV63mKMWSsiLYLdDsV9tM/CD+2z8EL7K/zQPgs/tM/CD+2z8EL7Kw913VQURVEURVEURSlhqNBTFEVRFEVRFEUpYZRmoTcl2A1QLKN9Fn5on4UX2l/hh/ZZ+KF9Fn5on4UX2l+5lNoYPUVRFEVRFEVRlJJKabboKYqiKIqiKIqilEhU6CmKoiiKoiiKopQwSqXQM8bcbIz5wxizzRjzTLDboxTFGDPdGHPAGPN7vmVVjTErjDF/5r5XCWYblTyMMXWMMV8aY1KNMZuMMf/JXa59FqIYY8obY9YYY9bn9tnI3OXaZyGMMSbCGPObMebT3P+1v0IYY0yaMWajMWadMWZt7jLtsxDGGHOBMWaeMWZL7jOtjfZZ6GKMaZB7fdlfx40xT2ifkVIn9IwxEQAmAegKoCGAe40xDYPbKsUB7wG4udCyZwCsFJH6AFbm/q+EBtkAnhKReACtATyWe11pn4UumQA6iEgTAE0B3GyMaQ3ts1DnPwBS8/2v/RX6tBeRpvnqemmfhTYTAHwuIlcAaAJeb9pnIYqI/JF7fTUF0BxABoCF0D4DUAqFHoBrAGwTkR0ikgXgQwC3B7lNSiFE5BsARwotvh3AzNy/ZwK4I6CNUpwiIvtE5Nfcv0+AD8Za0D4LWYSczP03Kvcl0D4LWYwxtQHcAmBqvsXaX+GH9lmIYow5H8ANAKYBgIhkicgxaJ+FCx0BbBeRXdA+A1A6hV4tAHvy/b83d5kS+tQUkX0AhQWAC4PcHsUBxpi6AJoB+AnaZyFNrhvgOgAHAKwQEe2z0GY8gKEAbPmWaX+FNgJguTHmF2NMYu4y7bPQpR6AgwBm5LpITzXGnAfts3ChN4DZuX9rn6F0Cj3jYJnWmFAUH2CMqQhgPoAnROR4sNujuEZEcnLdXWoDuMYY0yjYbVIcY4y5FcABEfkl2G1RLNFWRK4Gw0UeM8bcEOwGKS6JBHA1gLdEpBmAUyilLn/hhjGmLIDuAD4KdltCidIo9PYCqJPv/9oA/g5SWxRr7DfGXAQAue8HgtweJR/GmChQ5KWIyILcxdpnYUCua9JXYFys9llo0hZAd2NMGhhy0MEYkwztr5BGRP7OfT8Axg1dA+2zUGYvgL253g0AMA8UftpnoU9XAL+KyP7c/7XPUDqF3s8A6htjLslV/70BfBLkNinu8QmAPrl/9wHwcRDbouTDGGPAmIZUERmb7yPtsxDFGOCVveQAAAFYSURBVFPDGHNB7t8VAHQCsAXaZyGJiDwrIrVFpC743FolIvdD+ytkMcacZ4ypZP8bQGcAv0P7LGQRkX8A7DHGNMhd1BHAZmifhQP3Is9tE9A+AwAYkdLntWiM6QbGOkQAmC4iSUFuklIIY8xsAO0AVAewH8ALABYBmAsgFsBuAL1EpHDCFiUIGGOuA/AtgI3Iix8aBsbpaZ+FIMaYq8AA9Qhw0m+uiLxojKkG7bOQxhjTDsAQEblV+yt0McbUA614AF0CPxCRJO2z0MYY0xRMeFQWwA4ADyH3Hgnts5DEGBMN5t+oJyLpucv0OkMpFXqKoiiKoiiKoiglmdLouqkoiqIoiqIoilKiUaGnKIqiKIqiKIpSwlChpyiKoiiKoiiKUsJQoacoiqIoiqIoilLCUKGnKIqiKIqiKIpSwlChpyiKoiiKoiiKUsJQoacoiqIoiqIoilLC+H95OzKvRHe2JwAAAABJRU5ErkJggg==\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#sorting by true priece to beter visialize \n",
    "true = np.exp(Y)\n",
    "\n",
    "plt.figure(figsize=(15,15))\n",
    "plt.plot( Y.index , mean, \"ro\")\n",
    "plt.plot( Y.index ,true , \"go\")\n",
    "plt.plot( Y.index , upper, \"b\")\n",
    "plt.plot( Y.index , lower, \"b\")\n",
    "\n",
    "plt.savefig(f\"figs/wykres_predykcji.jpg\")\n",
    "plt.show"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "outputs": [
    {
     "data": {
      "text/plain": "<function matplotlib.pyplot.show(*args, **kw)>"
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": "<Figure size 1080x1080 with 1 Axes>",
      "image/png": "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\n"
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plot przekatnej\n",
    "plt.figure(figsize=(15,15))\n",
    "plt.plot( Dane[\"Przekątna ekranu\"],true ,  \"go\")\n",
    "plt.plot(Dane[\"Przekątna ekranu\"] , mean, \"ro\")\n",
    "plt.savefig(f\"figs/wykres_predykcji_doprzekntej.jpg\")\n",
    "plt.show"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  },
  "pycharm": {
   "stem_cell": {
    "cell_type": "raw",
    "source": [
     "\n",
     "\n",
     "#TODO: fix \n",
     "            \n"
    ],
    "metadata": {
     "collapsed": false
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}