random_forest_transfer-GenBank.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Code to reproduce the results described in \" Virus detection in human metagenomic samples using machine learning\"\n",
    "\n",
    "#### This notebook deals with learning a model on clean GenBank dataset and applying it to noisy metagenomic data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.18.1\n"
     ]
    }
   ],
   "source": [
    "# Load data\n",
    "import sklearn\n",
    "print sklearn.__version__\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import numpy as np\n",
    "from numpy import loadtxt\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Helper functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from numpy import loadtxt\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "\n",
    "\n",
    "#the normal random shuffle split\n",
    "def TT_split(features, labels, prop=0.9):\n",
    "    nr_samples = features.shape[0]\n",
    "    nr_train = int(nr_samples*prop)\n",
    "    indx = range(nr_samples)\n",
    "    np.random.shuffle(indx)\n",
    "    features_temp = features[indx,:]\n",
    "    labels_temp = labels[indx]\n",
    "\n",
    "    train_X = features_temp[:nr_train,:]\n",
    "    train_Y = labels_temp[:nr_train]\n",
    "    test_X = features_temp[nr_train:,:]\n",
    "    test_Y = labels_temp[nr_train:]\n",
    "    return train_X,train_Y,test_X,test_Y\n",
    "\n",
    "#split by giving an array of experiment numbers and the nr to be left to the test set\n",
    "def project_split(features, labels, projects, experiment_nr):\n",
    "    projects =np.array(projects)\n",
    "    print \"test set size\", np.shape(np.where(projects == experiment_nr)[0])\n",
    "    assert(len(np.where(projects != experiment_nr)[0])>0)\n",
    "    \n",
    "    train_X = features[np.where(projects != experiment_nr)[0]]\n",
    "    train_Y = labels[np.where(projects != experiment_nr)[0]]\n",
    "    test_X = features[np.where(projects == experiment_nr)[0]]\n",
    "    test_Y = labels[np.where(projects == experiment_nr)[0]]   \n",
    "    return train_X,train_Y,test_X,test_Y\n",
    "\n",
    "#helper to get the project ID from the sequenceID (which is a string)\n",
    "def seqID_to_projects(seqID):\n",
    "    projects = np.array([x[x.find(\"rr\"):] for x in seqID])\n",
    "    #print np.unique(projects), np.shape(np.unique(projects))\n",
    "    exp_names = np.unique(projects)\n",
    "    pr_nr = np.zeros(np.shape(projects))\n",
    "    for i,name in enumerate(exp_names):\n",
    "        print i, name, np.shape(np.where(projects==name))\n",
    "        pr_nr[np.where(projects==name)[0]] = i\n",
    "    return pr_nr"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## GenBank data - 10-fold cross validation\n",
    "#### We fist show that GenBank data is clean and valid dataset - model trained on GenBank data performs very well on GenBank data\n",
    "\n",
    "** !! please unzip the data first !!**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54259  8939]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54259\n",
      "         virus       1.00      0.88      0.93      8939\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54207  8991]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54207\n",
      "         virus       1.00      0.87      0.93      8991\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54056  9142]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54056\n",
      "         virus       1.00      0.87      0.93      9142\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54224  8974]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54224\n",
      "         virus       1.00      0.89      0.94      8974\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54118  9080]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54118\n",
      "         virus       1.00      0.87      0.93      9080\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54108  9090]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54108\n",
      "         virus       1.00      0.87      0.93      9090\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54096  9102]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54096\n",
      "         virus       1.00      0.87      0.93      9102\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54247  8951]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54247\n",
      "         virus       1.00      0.88      0.94      8951\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54335  8863]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54335\n",
      "         virus       1.00      0.88      0.93      8863\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n",
      "shapes after split  (568784, 59) (568784,) (63198, 59) (63198,)\n",
      "class counts in test set:  [54347  8851]\n",
      "report at 0.5:                  precision    recall  f1-score   support\n",
      "\n",
      "     not virus       0.98      1.00      0.99     54347\n",
      "         virus       1.00      0.87      0.93      8851\n",
      "\n",
      "   avg / total       0.98      0.98      0.98     63198\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from numpy import loadtxt\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "\n",
    "features = loadtxt(\"data/new_data/genbank/matrix.txt\",delimiter=\",\")\n",
    "labels = loadtxt(\"data/new_data/genbank/values.txt\",dtype=int)\n",
    "\n",
    "total_size = len(labels)\n",
    "fold_size = total_size/10\n",
    "\n",
    "indx = range(total_size)\n",
    "np.random.shuffle(indx)\n",
    "features = features[indx,:]\n",
    "labels = labels[indx]\n",
    "\n",
    "\n",
    "#FOR AREA UNDER CURVE\n",
    "roc_probas=[]\n",
    "roc_labels=[]\n",
    "for fold in range(10):\n",
    "    test_X = features[(fold)*fold_size:(fold+1)*fold_size]\n",
    "    test_Y = labels[(fold)*fold_size:(fold+1)*fold_size]\n",
    "    train_X = np.concatenate([features[:fold*fold_size,:], features[(fold+1)*fold_size:,:]])\n",
    "    train_Y = np.concatenate([labels[:fold*fold_size], labels[(fold+1)*fold_size:]])\n",
    "    \n",
    "    \n",
    "    print \"shapes after split \", train_X.shape, train_Y.shape, test_X.shape, test_Y.shape\n",
    "    print \"class counts in test set: \", np.bincount(test_Y)\n",
    "\n",
    "    #rf = RandomForestClassifier(class_weight=\"balanced\")\n",
    "    rf = RandomForestClassifier(n_estimators=100, n_jobs=3)\n",
    "    rf.fit(train_X, train_Y)\n",
    "    \n",
    "    # print \"feature imp:\",rf.feature_importances_\n",
    "    pred = rf.predict(test_X)\n",
    "    prob = rf.predict_proba(test_X)\n",
    "    roc_probas.append(prob)\n",
    "    roc_labels.append(test_Y)     \n",
    "\n",
    "    report = classification_report(test_Y, pred, target_names=[\"not virus\", \"virus\",\"not_classified\"])\n",
    "    print \"report at 0.5: \", report\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10\n",
      "Micro-average report at threshold 0.5: \n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "  not virus       0.98      1.00      0.99    541997\n",
      "      virus       1.00      0.87      0.93     89983\n",
      "\n",
      "avg / total       0.98      0.98      0.98    631980\n",
      "\n",
      "Macro-average precisions: [non virus, virus]:  [ 0.97965018  0.99836387]\n",
      "Macro-average recalls: [non virus, virus]:  [ 0.999762    0.87493177]\n"
     ]
    }
   ],
   "source": [
    "print len(roc_labels)\n",
    "#merge the results from experiments for micro-averaging\n",
    "all_labels = np.concatenate(roc_labels)\n",
    "all_labels = np.array(all_labels,dtype=int)\n",
    "\n",
    "all_probas = np.concatenate(roc_probas)\n",
    "all_preds = np.array(all_probas[:,1]>0.5,dtype=int)\n",
    "\n",
    "report = classification_report(all_labels, all_preds, target_names=[\"not virus\", \"virus\"])\n",
    "print \"Micro-average report at threshold 0.5: \\n\", report\n",
    "\n",
    "#for macro average, we need to go experiment per experiment:\n",
    "\n",
    "precisions_per_experiment=[]\n",
    "recalls_per_experiment=[]\n",
    "for i in range(len(roc_labels)):\n",
    "    precisions_per_experiment.append(precision_score(roc_labels[i],np.array(roc_probas[i][:,1]>0.5,dtype=int),average=None))\n",
    "    recalls_per_experiment.append(recall_score(roc_labels[i],np.array(roc_probas[i][:,1]>0.5,dtype=int),average=None))\n",
    "\n",
    "precisions_per_experiment=np.array(precisions_per_experiment)\n",
    "recalls_per_experiment=np.array(recalls_per_experiment)\n",
    "\n",
    "# If you are in doubt, you can verify that the list above looks like the print_out in the training phase (cell above)\n",
    "#print \"virus precisions per experiment look correct: \", precisions_per_experiment[:,1]\n",
    "#print \"virus recalls per experiment look correct: \", recalls_per_experiment[:,1]\n",
    "\n",
    "print \"Macro-average precisions: [non virus, virus]: \", np.mean(precisions_per_experiment,axis=0)\n",
    "print \"Macro-average recalls: [non virus, virus]: \",np.mean(recalls_per_experiment,axis=0)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under the microaveraged ROC curve is:  0.995435879143\n",
      "Area under the macroaveraged ROC curve is:  0.995438902219\n",
      "\n",
      "\n",
      "Area under the micro-averaged PRECISION-REC curve is: 0.98484962791\n",
      "Area under the macro-averaged PRECISION-REC curve is:  0.98485501676\n",
      "at precision 0.74   0.76550965918  we have recall of:  0.974984163675\n",
      "at precision 0.9   0.903553885554  we have recall of:  0.956702932776\n",
      "at precision 0.95   0.950648886107  we have recall of:  0.945134080882\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoQAAAIECAYAAACABz+XAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nThqEQgghhBB13P8DmcmuCKjYTOsAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12e0bd34d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA8YAAAH7CAYAAAAO6oabAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3Xd8VFX+//HXmUlPCAQSSgJKQglNiiLFSlEEFQVFV0UB\nUVcE1rUgaye4spavBXUti0pbFn9WioKLqIBrQ0BABKR3AQOEBEhPzu+PSWImmUwyIWGCeT8fjzG5\n55x77ufeCXE+OfeeY6y1iIiIiIiIiNRWDn8HICIiIiIiIuJPSoxFRERERESkVlNiLCIiIiIiIrWa\nEmMRERERERGp1ZQYi4iIiIiISK0W4O8AagJjjKbmFhERkdOGtdb4OwYRkT8SJcYFtGzV6SkpKYmk\npCR/hyGVpPfv9Kb37/Sl9+7kGaCsTw7GQHV+rDBGObGISFXTrdQiIiIiIiJSqykxFhEREfHRBG91\n3ipFRKRGUmIsp7VevXr5OwQ5CXr/Tm96/05feu9OXpK3Om+VIiJSIxk9W+uafEvXQURERE4HxhhN\nviUiUsU0YiwiIiIiIiK1mhJjERERERERqdWUGIuIiIiIiEitpsRYRERExEdJ3uq8VYqISI2kybfQ\n5FsiIiLiGwOU9cnBGKjOjxWafEtEpOppxFhERERERERqNSXGIiIiIiIiUqspMRYREREREZFarcYk\nxsaYh4wx7xljthlj8o0x2yvZzzBjzI/GmHRjzAFjzJvGmOiqjldERERERET+GGpMYgxMAnoDW4GU\nynRgjLkXmF6w/93AG8ANwBJjTGjVhCkiIiK13QRvdd4qRUSkRqoxs1IbY5pba3cWfL8OCLfWJviw\nfwNgF7AOOK9wmmljzJXAfOBha+3TZeyrWalFRETktKBZqUVEql6NGTEuTIpPwmAgFHileJZrrf0E\n2A7cfJL9i4iIiIiIyB9QjUmMq0DXgq/fe6j7HmhjjAk7hfGIiIiIiIjIaeCPlBjHFnzd56FuH2CK\ntREREREREREBIMDfAVShMABrbZaHuszibTxZeu4oIBAIArKAXO/bJrfg+0AgG8gpd9sSXKI+GNdb\n4HnbmhCv9Z63i/cfAjjL3bYmx0t9QMH5F98uo96U1b7w+nmrd9+2xrf2ru3Qgu3MEscre9sSWLBd\ncv8yto2P7QnAeozPy7bxsT2ZBT+PPrQnB0uYb+2Nb+1d275dL1t0fbPBLb6S28X393C9THnHyy2x\nf2H/To/brn8jZdd73g4v2C48fpiXbVc8ttT5OL1vm4rFX3zbkl0sPlPwfcltV3zGGCw5/B6ve32Z\n2ya3Au3DcSn8/RhRsJ1TLJ7C7cL9i22bcuo9bFtMif7LOv+CbeMo+D63oN6Wu/379YooqjfGtW2M\nBZONIQ/XDUwWY7KxODGOICAfm5+NcThwFm7bbIxx4Ci+7XDicARhi9oH4HQEFNTnFbR3AvkFxzAY\nhxNjKCgD43BgoKCs4KuhqMw6HOBwgMMJTifW4XRtO51Yp9NV7nC4yp0Fr4J6E+AkINiJM9BBQLCT\ngBAnAUFOAsKDcYYFExARQmCdEALCgwmqE0xQqJPAwN9jERER8Yc/UmKcDmCMCfaQHIcUb+PJ0pX/\nKvq+V8FLREREqleew0GWM5D0vDBMiCUvyEG2CSY1MwZHWC6OsDyyTB2OmXhsnXzyIix5YW1wRjWE\nmAbkxgQR17UjTc6JJaJJBBiDOQVZdlLBy2NdkutVVZYuXcrSpUurrkMRESmlxsxKXVwlZ6V+A7gD\naGWt3V6ibhauZZsirbWlkmNjjF1y+bMVj49KXDOfr7PvxzCVei992MdawGIwBXu69jWFV6Tw+IUf\nSGxhfeFmwXZBfZnbRQfML9gy5W4bLNa6tn/vz/s2NhfX0wSF/eV53y5o/3t/eeVuG7d4S/afi2tE\nr9i2Lafew/bv17dgu+j4OUCA521bOLIVUOz99LZtoWD/39/f8razfej/9/Zl91dw/iXev9/7K3F9\nPLYvdr2Krp8pZxuv/XnaNtb60H+Jn8ei/vLK2S55/nnlbhvr7efRfbuofYXjKTxe4fUt+PdW1J+X\nbWvx+d+7L/0XxVex/ou3r+jvE9c2pY7vqiv8z++/3VztHQX/Fm1BvaPg12YuFgdYR8Hv1hxco/2u\n9q5/K86CMjBkkm+DCn5/gDHHyc8PK2ifh9NxjLy88IL3NweHI528vAisdWIcmTgcGeTmRmBtIA6T\ngdNxgvy8EMgLIMBxnAAyyM8LhlwnQQGpOMkkPycEmxdIsPMoAY4M8nJCsLlBBAekEmAyyM8OJjc7\niNDgozjJIT8nAJvlICLgKEF52ZgcS0BuLkE2G0dl/p9ahqyAII5H1uFAxJkciOlNXvsEMlvlMvBv\nozGBAeTl5ZGRkUFERET5nZWj4Dej57rCH+tqolmpRUSq3h8pMb4NeBO4xVr7nxJ1W4Esa237MvbV\nck0iIiKnmrWQk4PNzCL3eCa5J7LIOZZJ9rEs0o9kcvzQcY4dSiX/uIOcI2lk/bYNcywZx/EInCf2\nE5mxlXoZvxKakklUxiGC87M9HuZEaBjfN7yOo7178mu7HMaOG4sxkJ+f77rVvBIjzEqMRUT+WE7L\nxNgY0wzXA2RbrWuYDmNMNK51jH/CfR3jgcA84BFr7VNl9KfEWERE5DSWk53Db1uPsuf7zeStP0De\n6v9Sd89azkjeQlTq0aJ2mYHBrI8/i20DJtPgylwOHznI9ddf7+ojJ4eAgIAKJcpKjEVE/lhqTGJs\njLkZOBPX/2vG4pq154WC6l3W2lnF2i4FLgKaW2t3Fyu/D/g/YBnwDtAUuA9XwtzN023UBfspMRYR\nEfkDysvNY+9nG9nxwt9o9fNPxB3cW1S3LTqB1b3+ygUv3k7jpmGsWLGCtLQ0+vbtW26/SoxFRP5Y\nalJivARXsuvJMmttnxJtLwQSiifGBXXDgHuBRCAN+Bh4yFp7yMuxlRiLiIj8geXm5mKMIeXHtRx6\n9kOaffwS4VknAPgtrCGL+z1DwJ8iuXbIVQQEuOYmPXLkCFFRUR5HkJUYi4j8sdSYxNiflBiLiIjU\nHgsWLKBp/QYcePYFOn6xnCbHXH9jX9+kPWvHJDH4visICgpi2rRpDB06lNDQ0FJ9JHHqZqUuSYmx\niEjVU2KMEmMREZHaJD09HafTSXBwMOTlsXLUw7Sd9U/CM9NJDwhl9vn/JOaeaNq0TSQxMdHf4Zai\nxFhEpOo5/B2AiIiIyKkUFhbmSoqB9KwsVnVNYN4zT7Ox3dmE5WZw+7LbcP7tQ9as/hlwzV79xRdf\nkJmZ6c+wRUSkGmnEGI0Yi4iI1Gb5+flkZ2cTFBjI5gdfpuULDxCQn8e37S4m4bM55OQfZ/369Vxy\nySVFzx/7k0aMRUSqnkaMRUREpFZzOByEhISAMWw8rzmLx95HdkAQ521Yxnc9Hyc319KiRYsakRSL\niEj1UGIsIiIigmskNjExkdWN6rHqiZfIJpDBe/7JT/f8Hw6H6yPTr7/+SnZ2tp8jFRGRqqbEWERE\nRARXYtyuXTvuu+8+ej40iuXD/wJA/0+mkPJ9Onl5eXzxxRccOnSozBmpoXpnpBYRkeqhxFhERESk\nmJCQEHJzc1neoQmrz7qI4Pxsgu+8lez0PGJiYoiOjmail/0neqsUEZEaSYmxiIiISAkBAQG0adOG\nb/90GcnBjTjrxCq+GfwwPXv2JCgoyN/hiYhIFdOs1GhWahERESktMzOTzMxMtr3+Fec8fDU5JoBt\n735Gm+t6Y4AjKSlERUWV2s8YqM6PFZqVWkSk6mnEWERERMSDkJAQ6tWrx7omR1h+dncCbS7mL49j\n8/MByM3N9XOEIiJSVZQYi4iIiHhx/vnn878r+nM8OIzEg1+z41+LAYiJifFzZCIiUlWUGIuIiIh4\n0apVKy677lq+7PoYAM7HxzLBWqy1nDhxolT7CRNOdYQiInKylBiLiIiIlOOss84iZNw5pIXX4cxD\nWxk0cR579uxh2rRpJCcnu7XVck0iIqcfJcYiIiIiFRDXKpZvLroQgOBnnyAutilXXnklwcHBfo5M\nREROlhLjAsbDK6mMtklqr/Zqr/Zqr/ZqX+vat2/fnoh7HiAlOIq2GasZ9b8txDdvTt3IyFMaj4iI\nVD0t1wRarklEREQqbMkVE+m9MInNUZ2JP7Cc35KTiY2NxZhTk7ZquSYRkaqnEWMRERERH+y/uhkn\nQkNpnbKGTW98xqJFi/jkk0/8HZaIiJwEjRijEWMRERGpuJycHO6YNZvpI0fwVcIIemz8F9baomeN\nk5KqdwIujRiLiFQ9JcYoMRYRERHfGMAaQ5qJJCjlN0Lq/j4BlzFQnR8rlBiLiFQ93UotIiIiUgn7\nYuKItGmseWYOe/bs4fjx4/4OSUREKkmJsYiIiEgl/NKxLQCOD95jx44dzJgxg/z8fD9HJSIilaFb\nqdGt1CIiIuIbA3z0zBQGP3gnKSaKyIzfcAQ5McboVmoRkdOQEmOUGIuIiIhvDJBvYWdIa+Kzt/DT\ny0vp+JeLXXVKjEVETju6lVpERETERxNwJcD7usQDkDLzXdauXcuuXbt46KFs/wYnIiI+U2IsIiIi\n4qOkgq97OrcD4MyfFuNwONmzZw9jxx7yW1wiIlI5upUa3UotIiIilfPzmp+I79GT8Kx0Ns3fROLA\n1tV+TN1KLSJS9TRiLCIiIlJJHTp35OemVwKw65/z/ByNiIhUlhJjERERkZNw+LxGADRe+T7z588n\nOTmZb775xs9RiYiIL5QYi4iIiJyM/heQ53DQ7siPNA2PIywsjOjoaH9HJSIiPlBiLCIiIuKjpGLf\nd+51AdvjWhNAHplzNvF//xdOYmKiv0ITEZFKUGIsIiIi4qOJxb6PjY1lT6ebAQhY9B4TJ3reR0RE\nai4lxiIiIiIn6dj5rlunE3cvBeDrr7/mwIEDfoxIRER8ocRYRERE5CRF9WzL0ci61M1OBSAuLo7w\n8HA/RyUiIhWlxFhERETkJCW2SWR90/OKtuPj46lTp44fIxIREV8oMS5gTOlXUpLntklJaq/2aq/2\naq/2aq/2v7dp3LgRF2xYiMF6blyF8YiISNUz1pb9C7y2MMZYXQcRERGpqCTcZ6YGWPTuIi658XIm\n2gl0md4cZ1Q9rrrqqio/tjEGa61SZBGRKqQRYxEREREfJXkoM1GGPXHNeIIJJO4Nol27dqc6LBER\nqSQlxiIiIiJVoHPnzqxvOhiAzHlLadmypZ8jEhGRilJiLCIiIlIFGjZsSH6f7gBEb/ySvLw8P0ck\nIiIVpWeM0TPGIiIiUjV2bUkmtl0sgbm5zH97BgGNo7n88sur9Bh6xlhEpOppxFhERESkinyzYjG7\nm5wJQOPtoXTr1s3PEYmISEUoMRYRERHxUVIZ5ZdffjkT7RsAZH3xA9HR0acsJhERqTzdSo1upRYR\nERHfGChzxWJjwGL4qf4FtNn/BUFBQVV7bN1KLSJS5TRiLCIiIlINWh1fyWf/XcSSJUv8HYqIiJQj\nwN8BiIiIiPzRpIbVpW56KgkmgRbnadkmEZGaTiPGIiIiIlVsW+R5ABz8dAPBwcF+jkZERMqjxFhE\nRESkip1IdCXDuT8swlqL5jIREanZlBiLiIiI+GiCt7oJkNvxHACaH/mKjz/+mDVr1pyawEREpFKU\nGIuIiIj4KMlbXRJsjcwGoP6uw1x4wWV07tz5VIQlIiKVpOWa0HJNIiIiUsWsJTWoAXVzU1g1dw/n\nXN20yrrWck0iIlVPI8YiIiIiVc0YDjVtAMCvn8wjNzdXzxmLiNRgSoxFREREqsGB2DYAhG/6jDff\nfJOUlBQ/RyQiImVRYiwiIiJSDQ6fGQ1AyIZ8Ro0aRf369f0ckYiIlEWJsYiIiIiPkrzVFVR2/fND\nAMQfXQXokWARkZpMk2+hybdERETENwYo65ODMWAtpBw+QlhsE4Kzs9m4ZDVndmtNWFjYyR9bk2+J\niFQ5jRiLiIiIVIN69aP4tWEcANvnLuTDDz/0c0QiIlIWJcYiIiIi1cAYQ3JMLAARu8K45ZZb/ByR\niIiURYmxiIiISDXJSLgOAPPLT36OREREvFFiLCIiIlJNjjU7DkCjQyvYvXs36enpfo5IREQ8UWIs\nIiIi4qMJ3uqKVSYOvgKAuNQdrFq5ik2bNlVvYCIiUimalRrNSi0iIiLVIz8vj8zwOoRlZbDru185\ns0eTk+5Ts1KLiFQ9jRiLiIiIVBOH08mOsI4A7PtsvZ+jERGRsigxFhEREalGGfF5AKR99wm7du0i\nPz/fzxGJiEhJSoxFREREqtHR2EQAIvZsY+nSpZqAS0SkBlJiLCIiIlKNgjqdC0DIjlSGDx9ORESE\nnyMSEZGSlBiLiIiI+CjJW12Jyi5DBwHQMusnsrM02aeISE2kxFhERETERxO91ZWo3GtPkBkUTL28\nVP777/ns37+/WmMTERHfKTEWERERqUatWrfmt+hGAOSs3cLq1av9HJGIiJSkxFhERESkGgUEBHCo\nYbzr+y3hXH755X6OSERESlJiLCIiIlLN8lpfCkDozpV+jkRERDypMYmxcbnXGLPRGJNhjNltjHnO\nGBNWwf3DjTEPG2N+MsakGWOSjTHfGGOGV3fsIiIiIt7sqeN6rjgudSWrV68mJSXFzxGJiEhxNSYx\nBiYDzwM/A2OB94C7gfnl7WiMMcB/cc2FsRy4D/g7rvObZox5qppiFhERkVpogrc6D5WNevcCIC51\nF4cPHyY5Obla4hIRkcox1vp/2QBjTDtgHfChtfb6YuVjgZeBm6y1/8/L/j2Ab4EXrLXjipUHAJuA\nKGttfS/725pwHUREROSPacPPP5PQuSsheVns//kwTdqX+bGkXMYYrLWmCsMTEan1asqI8U0FXyeX\nKH8TSAduLmf/yIKvbusfWGtzgUPAiZMNUERERKSy2nXowN6IlgDs/3Kjn6MREZGSakpi3BXIB1YU\nL7TWZgFrgHPL2f8H4Cgw3hgzxBjTzBiTWHAL9dl4v+NJREREpFplZGRw4gwnAMlffci6devIz8/3\nc1QiIlKopiTGscAha22Oh7p9QHTBbdEeWWuPAgOBFFzPJu8CNgJ3Addaa6dWfcgiIiIiFRMSEsKh\nhgkAhO/ax+7du8nOzvZzVCIiUqjMZPMUCwOyyqjLLNYmzUsfJ3BN3DUP1/PG9YExwDvGmKustV9U\nUawiIiIiPjHGkH5GGwBCdh/jgiuu8HNEIiJSXE0ZMU4HgsuoCynWxiNjzFm4kuFF1tq/WWvnWWun\nARcCB4A3C2auFhERETlpSd7qyqg8s/fVADQ88UtVhyMiIieppowY/wq0NcYEeridOg7Xbda5Xva/\nF1di/UHxQmtthjFmAa6R4+bAjrI6SCr2f7FevXrRq1cvH8IXERGR2mQiZSfHEyd6To53OXbREWic\nsZf1P28gINBJYmJiucdaunQpS5curXSsIiJSvpqSGK8ALgW6Ad8UFhpjgoHOwNJy9o8t+Or0UBdQ\n4qtHSWX9eVdERESkCjSIb8bxkHAiMk9wfGsKtrGnjy2llfyD/cSJE6spQhGR2qum3Er9bsHXe0qU\n/xkIBf5TWGCMSTDGlPzz6gbAACOKFxpj6gGDcE3KtbUK4xURERHxSYMGDTgYdgYAAbuhR48efo5I\nREQK1YjE2Fr7M/AqcI0x5kNjzG3GmOeB54Gl1tp3ijX/ElciXNxk4AjwtDFmpjHmTmPMw8CPQCPg\nEWutrf4zEREREfEsMTGRI/VcM1Onrdnk52hERKS4mnIrNcBfcT0D/GfgcuAQ8BKl1yC2Ba/fC6zd\nbYw5F3gc6Av8CcjAtQbyvdbaedUbuoiIiEj5Mpscg+1gty9l48aeNGrUiPr16/s7LBGRWq/GJMYF\nI7ovFry8tYsvo3wHcGs1hCYiIiLipuRf7d3qvFQea9gc+IqwA4fJzMwkN9fb3KIiInKq1IhbqUVE\nREROJ0ne6rxUBrQ6G4CQ/Wl06dKFhg0bVmVYIiJSSUqMRURERE6RFpf0ASAue7OfIxERkeKUGIuI\niIicIrZpEPnG0CAzmU1rN7BixQp/hyQiIigxFhERETllIhtEkRpZDweWrM1H9YyxiEgNocRYRERE\n5BSpV68eB0LPBCBr83F69uzp54hERASUGIuIiIj4LMlbnZfKoKAgjtZrDsCJDduqMCIRETkZSoxF\nREREfDTRW523SiCzcSYAZudXbNy4kb1791ZdYCIiUilKjEVEREROoYxGrQAI2p9OYGAgTqfTzxGJ\niIgSYxEREZFTKKrLRQCE/5ZMy5YtadKkiZ8jEhERJcYiIiIip1DDbm0BiMvRWsYiIjWFEmMRERGR\nUyiyTTR5DgcNsg+zfcM2Pv/8c3+HJCJS6ykxFhEREfHRBG913iqBmCaNSI2sC0DW1uPUqVMHa23V\nBSciIj5TYiwiIiLioyRvdd4qC+wPagHA8XUH6d69O8aYqghLREQqSYmxiIiIyCl2rEEjANI37PBz\nJCIiAkqMRURERE65jEYFI8Q7v2TTpk1s2rTJvwGJiNRyAf4OQERERKS2SW94BgDB+3OoU6eOn6MR\nERGNGIuIiIicYvXP7gdA+KEDxMbGEhsb6+eIRERqNyXGIiIiIj5K8lbnrbJAnbOaARCbrbWMRURq\nAqPlAcAYY3UdREREpKIMUNYnB2OgvI8Ve3fuoUmL5jjz89m5cStrflnHwIEDcTqd5R/bGKy1msZa\nRKQKacRYRERE5BRr2rwZaZGRAOTtyKBVq1Zay1hExI+UGIuIiIj4wZ6ANgCk/fQr7du3JyBAc6KK\niPiLEmMRERERP8ho6BoxPrFeaxmLiPibEmMRERERP0iPCQEgf/tn/PLLL6xZs8bPEYmI1F66Z0dE\nRETERxO81XmrLCY9uhEAQfsdREdHk5eXd/KBiYhIpWjEWERERMRHSd7qvFUWU7/LVQCEHdpPdHQ0\njRo1OtmwRESkkpQYi4iIiPhBUKu6ADTJ3uLnSERERImxiIiIiB/U79SUfAwNspI5tP833nvvPX+H\nJCJSaykxFhEREfGD+MR4jteJwGEt6VvS6NSpk9YyFhHxEyXGIiIiIn6y27jWMj6yeh+JiYkYY/wc\nkYhI7aTEWERERMRHSd7qvFUWk5+fT04T1wIhx9fvOtmQRETkJCgxFhEREfHRRG913iqLcTgcHG9Q\nH4Ccrd/y/fffs2WLJuISEfEHrWMsIiIi4icnopsAELAvk4SEBEJCQqqk39DQ0AOZmZla/0lEpISQ\nkJCDGRkZjUuWKzEWERER8ZP6XQbC/LcIO7SPhg0bVlm/mZmZjTSRl4hIacYYj3801K3UIiIiIn6S\nG+f62jBDt1CLiPiTEmMRERERP4k+pwUAjbP3cTj5MPPmzfNzRCIitZMSYxEREREfTfBW562yhNZn\ntyc9NJTAvFzSd57gnHPOOenYRETEd0bPn4Axxuo6iIiIiD9sDOtC24w1bJi2nHYjupXb3hiDtdbr\ngsf6bCMi4llZv0M1YiwiIiLiJ+np6dgm+QAc2/Srn6MREam9lBiLiIiI+ElgYCDHIusBkLltCx98\n8AFpaWl+jkpEpPZRYiwiIiLiJ4GBgaTXc61lzK8p9OzZk9DQUP8GVYv06tWL+Ph4f4ch4ldr166l\nb9++1K9fH4fDwRNPPOHT/rt27fJpvxEjRuBw1Lw0tOZFJCIiIlKLmLg+AAT8tp+4uDgCAwP9HNHp\nY9myZTgcDhwOB3fffbfHNsnJyQQFBeFwOOjTp49bnTGmRn5Al9prxYoV3H333VxwwQXUqVMHh8PB\nzJkzy2yfnZ3N448/TkJCAiEhIbRs2ZJJkyaRm5tboePl5eVxzTXXsG3bNp588klmzZrFNddcU1Wn\n45ExBmO8TpPgF/pNICIiIuKjJG913io9yIo+AUBoyu7KhlPrhYaGMnv2bHJyckrVFSYVnv7gsHjx\nYn755Zdqj0+kohYuXMjrr79OamoqnTt3LjeBvP7665k0aRKXXHIJr732Gr179+axxx7jz3/+c4WO\nt337dnbs2MG9997L6NGjuemmm+jQoUNVnMppR4mxiIiIiI8meqvzVulBncREABrmbuaHH35g3bp1\nlQ+slho8eDApKSke14GePn06V1xxBUFBQaXqAgICqnSE/vjx41XWV01VG86xOpV3/UaPHk1aWhrr\n1q3jnnvuwdvs8gsXLmT+/Pncf//9TJkyhZEjR/Lmm29y3333MWPGDL7//vty49m/fz8AUVFRvp3I\nH5ASYxERERE/anHhuQDUO36UVq1a0bx5c/8GdBo6++yzOeuss5g2bZpb+Q8//MCGDRu49dZbPe7X\nq1cvEhISSpVv27aNW2+9lWbNmhEcHExcXByDBg1i9erVRW2aN29Onz59WLNmDf3796devXp06tSp\nqP7w4cOMGTOGM844g+DgYM444wzGjh3LkSNHKnROmzZtYvTo0XTo0IHIyEjCw8Pp2rUrb7/9tlu7\nN954A4fDwSeffFKqD2stTZs25eyzz3YrX7lyJYMHDyYmJoaQkBDatGnDP/7xD/Ly8jxenx07djBk\nyBAaNGhA3bp1i/qeNGkSF198MU2aNCE4OJgzzzyT0aNHezzHjIwM7rvvPmJjYwkLC+O8887jyy+/\nLPN5061bt3LLLbcQGxtLcHAw8fHxjB8/nvT09ApdP4C5c+dy/vnnExERQZ06dbjggguYP3++W5vu\n3bvTuHFj8vPzS+2/aNEiHA4HL7/8slv566+/TteuXQkPD6dOnTr06dOHpUuXurUp/tzte++9R9eu\nXQkLCyvzlv9CMTExFZ5nYPbs2Rhj+Otf/+pWXphQz5o1y+v+vXv3plevXhhjit4Hp9PJ7t2uu1fS\n09N56KGHaNmyJSEhITRp0oThw4cX1ZcnKyuLBx54gLi4OMLCwujRoweLFy/22HbDhg1cd911NG3a\ntOhYffr04dNPP63QsapCwCk7koiIiIiUEt0mmmwCicg9TrYJJqhOmL9DOi2NHDmS+++/n/3799Ok\niWtCs6lTp9KwYUOuvPJKj/t4uk115cqV9O3bl7y8PG6//Xbat2/PkSNHWLZsGd9++y1dunQp2nfX\nrl306dMlPIl6AAAgAElEQVSH66+/niFDhhSNBqalpdGzZ0+2b9/ObbfdRpcuXVi9ejWvv/46S5Ys\n4YcffiA8PNzr+SxdupSvv/6agQMHEh8fz4kTJ3j//fe54447OHToEH/7298AuOGGG7j33nuZOXNm\nqfP8/PPP+fXXXxk/fnxR2YIFC7j22mtp1aoV48aNo379+nz33Xc8/vjjrF27lnfffdft+hw/fpyL\nL76YCy64gH/84x/89ttvgOvZ1ueee44hQ4YwaNAgwsPDWbFiBW+//TbffPMNq1atIiDg91RjyJAh\n/Pe//2Xw4MH07duXHTt2MHjwYOLj40u9D6tWraJv375ERUUxatQo4uLiWLt2LS+//DLffvsty5Yt\nw+l0er1+r732GmPHjqVt27ZMmDABcN09MGjQIKZMmcLtt98OuCaCGjt2LP/973+5/PLL3fqYOXMm\ngYGB3HTTTUVlN998M++++y5Dhgxh5MiRZGVl8Z///IdLL72UOXPmlHoP5syZw549e7jrrru46667\niIyM9Bq3L1auXElcXBxxcXFu5U2bNiU2NpYVK1Z43f/RRx/lggsuYNKkSdx5551ceOGFgCs5z83N\npV+/fnz33Xdcd911jBs3ji1btvDaa6+xePFiVq5cSWxsrNf+b7jhBubNm8fVV19Nv3792LZtG9dc\nc02pCe+OHDlC7969cTgcjBo1ijPPPJNDhw6xcuVKli9fzoABAypxdSrBWlvrX67LICIiIlIx3j45\n+Pqx4tChQ/ZwZJS1YPd9tbX8Y7sOoM821tqlS5daY4x9/vnn7eHDh21wcLB96qmnrLXWZmRk2Hr1\n6tnx48dba62NiIiwvXv3dtu/V69eNj4+3q2sffv2NjQ01P78889ej928eXPrcDjs1KlTS9U9/PDD\n1uFw2DfeeMOt/NVXX7XGGPv444+Xe27p6ekey3v16mXr1atnc3Nzi8quu+46Gxoaao8ePerW9uab\nb7ZBQUE2OTnZWmttZmambdy4se3Vq5fNz893azt58mTrcDjssmXL3I7lcDjKjDczM7NU2dSpU63D\n4bDvv/9+UdmCBQusMcbeeeedbm0//fRTa4yxDofDrbxjx462Xbt29sSJE27lc+fOtcYYO2PGDI/x\nFEpJSbERERG2devW9vjx40Xlx44dsy1atLCRkZE2NTXVWmvtkSNHbHBwsP3Tn/7k1sexY8dseHi4\nHTRoUFHZRx99ZI0x9q233nJrm5eXZ7t27WoTEhKKynbu3GmNMTYoKMhu2rTJa7xl+eCDD7yeb506\ndWyPHj081nXr1s3GxcWVe4zCf0MljzFlyhRrjLEPPvigW3nhezls2LCissJznThxYlHZokWLrDHG\njhw50m3/efPmlXrP58+fb40xbj8z1ams36G6lVpERETEj0JDQ0mLcI0i7fvhJz7++GM/R3R6ql+/\nPldddRXTp08H4MMPPyQtLY2RI0dWuI81a9awYcMGRo4cSfv27St0zBEjRpQqnzt3LjExMdxxxx1u\n5XfeeScxMTHMmTOn3L6L306blZXFkSNHOHz4MP369SMtLc1t0rDhw4eTmZnpNtp74sQJ5s6dy4AB\nA4iOjgZck40dPHiQESNGFPVX+Orfvz/WWj777LNSsdx///0eYwwODgYgPz+f1NRUDh8+TK9evbDW\nsnz58qJ2n3zyCcYY7r33Xrf9+/fvT9u2bd3Kfv75Z9atW8eNN95IRkaGW4znnXce4eHhHmMsbvHi\nxZw4cYK7777bbWQ+IiKCu+++m+PHj/P5558DrmdrBw4cyMcff+y2hvj7779PRkYGw4cPLyqbNWsW\nkZGRXHXVVW5xpaSkMHDgQHbu3MnWrVvdYrnyyitp3bq113grKz09veg9KCkkJMSn285Lmjt3Lk6n\nkwcffNCt/PLLL6dz584en+cvbt68eRhjGDdunFv5VVddRWLBvAqFCm/P//TTTzl27FilYz5ZSoxF\nREREfDTBW523Sg/CwsI4Gt4QgLw9x+nevXvlA6vlbr31VrZs2cI333zDtGnT6NatW6kP4d5s2bIF\ngM6dO1eofYsWLTzejr1jxw4SExNLPTvrdDpp3bo127dvL7fvEydOMG7cOM4880xCQ0OJjo4mJiaG\nRx55BICUlJSitv3796dhw4Zuy/p88MEHpKenM2zYsKKyjRs3Aq7rFBMT4/Zq27YtxhgOHjzoFkdM\nTEyZt/++99579OjRg9DQUKKiooiJiSm6JsXj27FjBw6Hg5YtW5bqo+T7UxjjhAkTSsXYqFEj0tPT\nS8VY0o4dOzDG0K5du1J17du3x1rr9h4MHz6cjIwM3nvvvaKymTNnEhUV5XZr9C+//MKxY8do1KhR\nqdiSCqajLxlbq1atvMZ6MsLCwsjKyvJYl5mZSVhY5R/L2LFjB7GxsUVJa3Ht27fn2LFjHDp0qMz9\nt2/fjsPh8PhHgZJ/DLnooosYPnw406dPJzo6mgsuuICkpKSin4VTRc8Yi4iIiPgoyVudt8oypIT0\nAFaQuzuZhg0bVi4o4bLLLiM2NpaJEyeyZMkS/vWvf1Xr8U4m8SjPjTfeyMKFC4ue/WzQoAFOp5MF\nCxYwefJkt8minE4nN910Ey+99BLbt28nISGhKLEbOHBgUTtrLcYYnnvuObeJwoor+dxoWef40Ucf\nccMNN9C9e3defvllmjVrRkhICHl5eVx22WUeJ7OqyNq1tmAW5vvvv5/+/ft7bFPVMygPGDCAmJgY\nZs6cye23387u3bv56quvGD16tNtz0tZaYmJieOedd8qcLbrkUkfV+TMSGxvLvn37PNbt27ev1LPH\nNdm0adN44IEH+PTTT/nf//7HCy+8wKRJk3jppZcYPXr0KYlBibGIiIiIn+U1SoF1YPdpLeOT4XA4\nGDZsGE899RTh4eHccMMNPu1fOLq1Zs2ak4ojISGBTZs2kZ+f7zZqnJeXx+bNmz3OhF1camoqCxYs\nYPjw4bz66qtudWXdRjx8+HAmT55clNwtW7aMUaNGuS1H1apVK6y1hIWF0adPn5M4Q9dtxaGhoSxd\nutTtdt5NmzaVatu8eXPy8/PZsmVLqRHikutIF46wOp3OSseYkJCAtZb169fTu3dvt7r169djjHF7\nDwr/sPDyyy+zc+dO/vOf/wC4jbYXxrZw4UK6d+9erQlvRZ177rnMnj27VBK8d+9efv31VwYNGlTp\nvhMSEli0aBFpaWml7hjYsGEDkZGRRbfol7V/fn4+mzdvLjVCvGHDBo/7tGvXjnbt2nH//feTlpZG\nt27dePDBB09ZYqxbqUVERET8LL+J63nWkMNbmDlzJpmZmX6O6PQ1atQokpKSeP3114mIiPBp306d\nOtG+fXumTp1a5of3ihg0aBDJycm89dZbbuVTpkwhOTmZa665xuv+hTMulxx13b9/f6nlmorH3rFj\nR2bNmsW///1vrLVuz8eCa0S9YcOGPP300263OhfKzMys8DrFTqcTY0ypJZ7+/ve/lxoZHjhwINZa\nXnzxRbfyhQsXlkqMu3TpQocOHXjjjTfYsWNHqePm5eV5jL24Sy+9lPDwcF555RW38zl27BivvPIK\nERERXHrppW77FF6rGTNmMGvWLBITEzn33HPd2gwbNoy8vLxSz90WKpyx+1S58cYbsdYyefJkt/IX\nX3wRYwxDhw6tdN+DBg0iLy+Pp59+2q38008/ZfXq1Vx99dVe97/66qux1vJ///d/buVz585l8+bN\nbmUpKSmlRuAjIyOJj48nPT29zNvFq5pGjEVERET8rF67bgCEp/1Gv3793Eb5xDfNmjXj8ccfr/T+\n06ZN45JLLqFbt27cdtttdOjQgaNHj7Js2TIGDBjAmDFjyu1j/PjxvP/++4wZM4ZVq1bRpUsXfvzx\nR6ZOnUrbtm154IEHvO4fERFBv379mDVrFiEhIZx77rns3LmTKVOmkJCQUOZayMOHD+f+++/nmWee\noXXr1qUSu7CwMGbOnMngwYNJTExk5MiRtGzZkqNHj7Jx40bmzJnD3Llzueiii8o9xyFDhvDRRx/R\nu3dvhg0bRk5ODnPnziUjI6NUknP55Zdz2WWX8eabb5KcnMwll1zC9u3befPNN+nUqRM//fSTW/t/\n//vf9O3bl44dOxZNhJaens7WrVv56KOPePrpp0uN5hZXt25dnn32WcaOHUv37t0ZMWIE1lpmzJjB\n9u3bmTJlCnXq1HHbp3PnznTo0IEXX3yRY8eO8dRTT5Xq99prr+XWW2/l1Vdf5ccff+TKK68kOjqa\nvXv38t1337Ft27ZSk2/5avfu3fz73/8GXKPbAPPnz2fPnj2AKzlv1qwZ4LquV155JS+88AJHjx6l\nZ8+efPvtt0ydOpVhw4bRs2fPSscxYsQIZsyYwTPPPMOOHTu46KKL2LJlC6+//jpNmjRh0qRJXvfv\n168fAwcOZMaMGUWTu23dupUpU6bQoUOHonMD1/PcL774IoMHD6Zly5YEBgaydOlSPvvsM/70pz+V\nOcFYlfM0VXVte1FLljQQERGRmumXuRutBbsrqEW5bdFyTUWWLl1qHQ6HfeGFF8ptW6dOHdunTx+3\nsl69erktsVNo8+bN9pZbbrFNmjSxwcHBNi4uzg4ePNiuXr26qE3z5s1L9VfcoUOH7JgxY2yzZs1s\nUFCQbdasmf3LX/5iDx8+XKFzO3z4sL3jjjtsXFycDQ0NtR07drRvv/22nT59eqlllQodPHjQBgUF\nWafTWbRslSfr16+3t9xyi23atKkNDg62jRs3tueff76dNGmSTUlJKff6FHrrrbeKlreKjY21o0aN\nsikpKdbhcJRapic9Pd3ee++9tnHjxjYsLMz26NHDLlmyxA4ZMsSGh4eX6nv37t32rrvusvHx8TY4\nONhGR0fbrl272kceecTu3bu3IpfQzp07155//vk2IiLCRkRE2PPPP9/Onz+/zPbPP/+8dTgcNjAw\n0OsxZs2aZS+66CJbt25dGxoaauPj4+21117rttzQzp07rcPhsE888USFYi1UuHySw+Hw+Cr5vmdl\nZdnHHnvMxsfH25CQENuiRQs7adIkt+W8yjuew+HwuCRUenq6ffjhh22LFi1scHCwbdSokR0+fLjd\nvXu3W7uyzjUzM9OOGzfONmnSxIaFhdnu3bvbxYsX2xEjRlin01nUbs2aNXbEiBG2VatWNiIiwtat\nW9d27tzZvvjiizY7O7uil67CyvodamwZD47XJsYYq+sgIiIiFZVE2RNwJSX5PgHXmv+tofNFXch0\nBBOSmwFeJikyxmCt9TqLkT7byOmiY8eO5ObmntSt6yK+KOt3qJ4xFhEREfHRRG913irLcGaH5mQF\nBRGSn8X3//2CVatWVTo2kZrI03PzCxYs4Oeff6Zfv35+iEjEnZ4xFhEREfGzqKh6JIdFEpN9iAY5\ndYlt08bfIYlUqSeeeILVq1fTu3dv6taty+rVq5k2bRoxMTGMHz/e3+GJKDEWERERqQn257cmhkNk\n7jpBeHi4v8MRqVIXXngh3377Lc899xypqanUr1+f6667jieeeKLU2ski/qDEWERERMTPsrKycDQ+\nAWmQtnXPKT22l8eZq4Uefa6dBgwYwIABA/wdhkiZ9IyxiIiIiJ8FBgZyPDwGgIydm4uWaxERkVND\nI8YiIiIiPprgrc5bZRkcDgdZ9RJc3/+WcUpH1jSCKyKiEWMRERERnyV5q/NW6UVQs44AOJKPEh0d\nXblOpMiSJUvo2bMnkZGROJ1OZs6c6e+Q3EyfPh2Hw8FXX33l71AqbcaMGaf9OYgU0oixiIiISA2Q\nErQLgHrZWs/1ZB09epRrr72WZs2a8cILLxAWFsZ5553n77DcGGMwp/oB72rwRzgHEVBiLCIiIlIj\nNOncA4B6WfuZO3cuZ599NmeccYafozo9rVixgtTUVKZNm8bVV1/t73A8GjZsGDfeeCNBQUH+DkVE\nUGIsIiIiUiMk9OwEQOixDHr37k1YWJifIzp97d+/H4CoqKgq7Tc/P5+srCxCQ0NPui9jTIWS4qo8\npoiUTc8Yi4iIiNQAdVo2BCAyIxWnI5LAwEA/R3R6io+PZ8SIEQD06tULh8OB0+ksqj98+DBjxozh\njDPOIDg4mDPOOIOxY8dy5MgRt34Kn5/94osv+Pvf/07Lli0JDQ3l/fff93jc1NRUQkNDGTJkiMf6\nhx56CIfDwU8//QR4fsbY0zFDQkKKjulwOBg5cmSpvj0965uSksK9995bFHd0dDRdu3blueeeq8BV\nhJycHJ599lm6dOlCeHg49erV49xzz+XVV1/1ut/x48d59NFH6dGjBzExMYSEhNCqVSseeughMjIy\n3Npaa5k8eTKdOnUiMjKSunXr0qZNG26//Xby8vKK2n377bcMGDCAJk2aEBoaStOmTbniiiv44Ycf\nKnQuIhWhEWMRERERHyVR9gRcSUmVm4Br28EDnBkYSGhOJls2p9LqnHqVDa9We+mll1i4cCFvvvkm\njz76KG3atCmqS0tLo2fPnmzfvp3bbruNLl26sHr1al5//XWWLFnCDz/8QHh4uFt/48aNIzc3lz//\n+c9ERkaSmJjo8bh169blqquuYv78+Rw9epR69X5//6y1zJ49m86dO9Oxo2uSNW/PGFf0mMWV7GvI\nkCF8/fXX3HXXXZx11llkZGSwceNGli1bxrhx47z2lZOTQ79+/fjqq6/o168ft9xyCyEhIaxbt445\nc+YwZsyYMvfdt28fU6dO5dprr2Xo0KEEBASwbNkynn32WdasWcOnn35a1PbJJ59kwoQJXH311dx1\n1104nU527NjB/PnzycrKIiwsjM2bN9OvXz+aNGnCPffcQ6NGjTh48CBff/01a9eupVu3buVeG5EK\nsdbW+pfrMoiIiIhUjLdPDpX9WJGWlmYP142yFuwnL861CxYsKKN/CkPQZ5syTJ8+3TocDrts2TK3\n8ocfftg6HA77xhtvuJW/+uqr1hhjH3/8cbc+jDG2TZs2NjMzs0LHXbBggTXG2Ndff92t/PPPP7fG\nGDt58mSvMZZ3TGOMvfXWW8s939TUVGuMsWPGjKlQ3CU988wz1hhjH330Ua/tPJ1DTk6Ozc3NLdX2\nsccesw6Hw65YsaKo7Oyzz7bt27f3eoyXX37ZOhwOu3LlSh/PQsSzsn6H6lZqERERkRqgTp06pIXV\nByDsWAQXX3yxnyP645k7dy4xMTHccccdbuV33nknMTExzJkzp9Q+o0ePJjg4uEL9X3bZZTRq1KjU\n0lAzZ84kMDCQm266qUL9+HJMT0JDQwkODmb58uXs2rXL5/1nz55N/fr1eeyxx3zeNyAgoOjW9by8\nPI4ePcrhw4fp27cv1lqWL19e1LZu3brs27ePb775psz+6tati7WWOXPmkJWV5XM8IhVV6cTYGNPT\nGPOoMeY1Y8zUMl5vV2WwIiIiIn9kR4PaApD369FSt/TKyduxYweJiYk4HO4fgZ1OJ61bt2b79u1u\n5cYYWrVqVeH+nU4nQ4cOZfny5WzduhWA9PR05syZw2WXXUZMTEy5ffh6TE8CAwOZPHkyP//8M/Hx\n8XTo0IG7776bL7/8skL7b9myhTZt2lR6xuzXXnuNTp06ERwcTP369YmJiaF3794YY0hJSSlq949/\n/IOQkBAuuugimjZtys0338w777xDTk5OUZsbbriBSy+9lKeeeor69evTt29fnn32WXbv3l2p2ETK\n4nNibIwJNsbMBb4GJgKjgBFeXiIiIiJSjoyMDHIbHAYgd98vfo5GCvk6O/iwYcOw1haNGn/44Yec\nOHGC4cOHV9sxc3NzS5Xdeeed7Ny5k7feeotzzjmHDz/8kEsuuaTCo9aV9cILLzB27Fji4uKYMmUK\nCxcu5PPPP2fGjBlYa8nPzy9q26NHD7Zt28YHH3zANddcw9q1axk6dChdunTh6NGjAAQFBbFo0SKW\nL1/Oww8/TEBAABMmTKBNmzbMmzevWs9FapfKjBgnAVcBJ4A3gL8Ct5bxKj1tnoiIiIiUEhwcTHqd\nWACCUtYxbdo0P0f0x5OQkMCmTZvckjNw3fK7efNmEhISTvoYHTt2pFOnTsyaNQtw3UZdr149Bg4c\neNJ9169fv9Ts2QDbtm3z2L5Ro0aMHDmSGTNmsHfvXm688UbeffddVq1a5fU4rVu35pdffnEbua2o\nWbNmER8fz8KFCxk5ciT9+/enT58+NGzY0GP7sLAwBg8ezMsvv8y6det49dVX2bBhA2+/7X7jadeu\nXXnkkUdYtGgRW7duJTw8nEcffdTn+ETKUpnE+E+4kuJzrbVjrLWvWGtnlPWqaKfG5V5jzEZjTIYx\nZrcx5jljTIX/ZGaMiSrYZ0tBH78ZY740xpxfifMUERER8WiCtzpvlV44HA5M486u75NDuPbaayvX\nkZRp0KBBJCcn89Zbb7mVT5kyheTkZK655poqOc7w4cPZtWsXs2fPZsmSJdxwww2Vvi25uNatW/Pd\nd9+RmZlZVJaSksL06dPd2mVkZJRaGskYw1lnnQXgMbkubujQoRw5coQnn3zS5xidTifGmMJJ4ADX\niPZTTz1Vaubsw4cPl9q/S5cubjF6ahMXF0dMTEy55yHii8os1xQLLLHWbqriWCYDfwE+BJ4D2gJ3\nA52BS8rb2RhzBrAMCAPeBjYDdYGOQFwVxyoiIiK1WJK3Om+V5Qg+s6Xr69FUIiMjK9+RuCVmhcaP\nH8/777/PmDFjWLVqFV26dOHHH39k6tSptG3blgceeKDcPipi6NChjB8/ntGjR2OtZdiwYRWO0dsx\nx44dy80330zv3r255ZZbSElJ4a233qJ58+YcPHiwqN3mzZu5+OKLGTx4MB06dCAqKooNGzbwxhtv\nkJCQwIUXXug1/r/+9a98/PHHPPnkk/zwww/069ePkJAQ1q9fz+bNm/nss8/KjHfIkCE8/PDD9O/f\nn2uuuYbU1FTeeecdgoKCSrVt27YtPXr0oHv37sTGxrJ//36mTJlCcHAwN954I+Ba0umzzz7jyiuv\nJD4+Hmst8+fPZ9OmTfztb3/zeh4ivqhMYpwMpFVlEMaYdsBY4ANr7fXFyncCLxtjbrDW/r9yuvkP\nrhHws6y1v1VlfCIiIiKnwp6s9XQDGuRt9ncopz1PawRHRkby7bffMmHCBObPn8/06dNp1KgRo0eP\nJikpqdSEZ2WtM1yemJgYBgwYwCeffELr1q3p3r17hWP0dsybbrqJ/fv3889//pP777+fhIQEkgr+\nElN8tudmzZpx2223sWTJEubNm0dWVhZxcXHceeedjB8/npCQEK/xBwYGsnjxYp5//nlmz57NI488\nQkhICK1atWLkSPcnJUvGO378eADefvtt7rnnHho3bswNN9zAiBEjaNeunVv7cePGsXDhQl555RVS\nU1Np2LAhPXv25MEHH6RDhw4ADB48mAMHDvD+++9z8OBBQkNDadWqFW+99Ra33nqr1/MQ8YXx9S9h\nxph/AZcD8dba0k/6VyYIY54EHgIutNZ+W6w8GDgMLLXWXull/4uApcBYa+1rxpgAINBam1HWPiX2\nt5X9i6CIiIhIVVn7/v/odP1FHKrbgPf+8QQjR44slcQU3KbqNWPTZxsREc/K+h1amWeMCxc0+2dB\n4loVugL5wIrihdbaLGANcG45+w8ALLDXGPMxkAGcMMZsMsYMraIYRURERKpV826uZ0DDjqczdOjQ\nKnkuVUREyleZW6lHAYuAO4D+xpgvgd24EtuSrLX27xXoMxY4ZK31NPXdPqCnMSbAywh1ImCAN3E9\nW3wLEAzcD/y7YN8KTwQmIiIi4g91mkaSi5OwvAyyA8JKrbcrIiLVozK3UufjGp31dgtPYb211jor\n0OdWIMBa29xD3QzgZiDKWuvx2WZjzGKgL7ANaFuYQBtj6gHbgQxrbZkTcOl2IxEREfFFEmVPwJWU\nVPkJuFatWkXihRcTkXGCA2sP0Lhjo1JtdCu1iEjllfU7tDIjxhOrIJ6S0oGYMupCirUpSwauZPyd\n4qPK1tqjxpj5wC3GmERvM2knFfs/WK9evejVq1fFIhcREZFaZyJlJ8YTJ1Y+MW7RogXpIeFEZJzg\nf3Nnc27kYHbu3MnSpUsr16GIiFSIz4mxtbY6EuNfgbbGmEAPt1PH4brN2ttEX3sLvh7wULe/4GuU\ntwCSTmZtBREREZEqUK9ePbYFRdGQ32gW3oGmTZvSvHlztz/YT5xYHR/FRERqt5ry4MoKXLF0K15Y\nMLlXZ0pMyuXBD7hu3W7qoa5ZwVct4SQiIiI1XqqjOQC5v2USEFCZm/tERMRXJ50YG2OaGGO6Frya\nVLKbdwu+3lOi/M9AKK41iguPl2CMSSzRbi5wDLjZGBNWPDbgamCTtXZ7JWMTEREROSUOHDiAiT4C\nQPq+LX6ORkSk9qh0YmyMucMYswnXbczLC157jTG/GGNu96Uva+3PwKvANcaYD40xtxljngeex7WG\n8TvFmn8JbCix/1FgHK7brpcbY+41xjwIfAcEAn+p3FmKiIiInDpRUVFkhLQEwB5azffff+/niERE\naodK3Z9jjJmOa0kkg2vSq18LqmKB1sC/jDHnW2tv9aHbvwI7cI0SXw4cAl4C/j97dx5f09E/cPxz\nbsgukUgICdmQ2mmtRRpBSGuLpUIRa2t7lFraqpIqRRe0T1WrliDVqj4VWlJLK1SpotRWa2KtNQkh\nmyzf3x+R+3Pdm5ubiNKa9+t1XtyZOXO+584VdzJzZqbcU07uHIaJIp9rmnYVmABMJW/7qB1AuIio\n/1UURVEURSkx9345Mcgzl1kIGxsbxMUHAF1yORo0aFD8yhRFURSLFbljrGlaL6Afec/sTgGiRCTz\nTp4N0J+8hRr7aZq2QUS+sqTeO3sKzLlzmCvnayYvhrxp1YqiKIqiKA9MpLk8c5kW0NzyNurQXb+N\njY3N/VWmKIqiWKQ4U6mHALeBYBH5LL9TDCAimSLyGXl7CmeTN/qrKIqiKIqiWOha7gEAnLL/fMiR\nKJE1OMoAACAASURBVIqiPD6K0zGuT95zv0cKKnAnb8udsoqiKIqiKIqFPGq2BcA59wJr1qx5yNEo\niqI8HorTMbYHkiwol0TeitKKoiiKoiiKhdyqVwOg1HUhJCTkIUejlDQfHx+Cg4Mfdhj/SpGRkeh0\nOs6ePftA6ldt9+9WnI7xBaCxpmlaQQXu5DXi/xflUhRFURRFUSxg5+UMQKmUbOzs1BiDOVu3bkWn\n0xkcZcqU4amnnmLu3Lnk5OQ87BCNmPoK/eGHH7J06dKHEM2/i6ZpJt/fkqz/fvzT2nnr1q289dZb\npKSkPOxQ/hbF6RhvAHyB9zRNs7o3U9M0HTAL8AN+uL/wFEVRFEVRHj2R5vLMZVrg99O/A+BmdZmc\nnNz7q+wx0bt3b6Kjo1m+fDmRkZFkZ2fzyiuvMGLEiIcdmkXmzp37j+owKcXzT2vnuLg4pk6dyvXr\n1x92KH+L4nSMZwI3gDHAcU3T3tY0baCmaQM0TXsbOA6MBa7fKasoiqIoivKv8pa5PHOZFmjWrg25\nmoZ9VjoL5s2/v8oeE08++SS9e/fmhRdeYOzYsezcuRMvLy8WLlxIYmLiww5PUf6R8jYNenwUuWMs\nImfJ22f4AnkjxxOBz4GFd/7uB5wHnhWRcyUXqqIoiqIoyr+fq5sbqdYOALRt+OxDjuafyd7enqZN\nmyIinDp1yih/z549hIWF4e7ujq2tLU888QTvvPOO0dTrI0eO0KNHD7y8vLC1taVixYoEBwcTGxur\nL2PuuVZLnknNPzcuLk4/HdzKykpf344dOwgNDaVixYrY2dnh5eXFc889x2+//WbRe7Ft2zbatm1L\n2bJlsbe356mnnmLx4sVG5YKCgvDz8+PixYv06tULV1dXHBwcaN++PSdOnLDoWgCXLl1i2LBheHt7\nY2Njg6enJy+99BJXr141KHfx4kXGjh1LgwYNcHV1xc7Ojlq1avHuu++Sm2s8UyIrK4t3332XBg0a\n4ODgQNmyZWnUqBHz5s0zKpuRkcHEiROpXLkytra21K9f36DNCnP+/Hmef/55ypYti7OzM507dyY+\nPt5k2ZUrV9K5c2e8vb2xtbXF3d2dsLAwDh48aFCusHbetGkT4eHh+Pv7Y29vj4uLC+3atWPbtm0W\nx71u3TqCgoJwd3fH3t4eb29vunXrxsmTJw3KWdJGAwYMYOrUqUDe5zg/5vy0f6Mi72MMICK/appW\nDegBPAN43sm6AGwFVt29jZOiKIqiKIpiueu68pThFlmX1dep4srvDLi6uhqkr1u3jm7dulGtWjXG\njRuHq6srO3fuZPLkyfzxxx+sXLkSgKSkJFq1aoVOp2Po0KF4e3tz7do19uzZw65duwgNDQXMP9dq\nyTOp0dHRjB49Gnd3dyZNmqQfpXN3d+f48eOEhIRQsWJFRo8eTYUKFbh8+TLbt2/njz/+oHHjxmbr\n/u677+jatSsVK1Zk/PjxODo68tVXXzF48GASEhJ4++23DWJNTU0lMDCQZs2aMWPGDBISEpg7dy5d\nunTh0KFDhd7PuXPnaNq0KdnZ2QwaNAh/f39OnjzJJ598QlxcHHv27KFMmTIAHDhwgJiYGMLCwvD3\n9ycrK4sffviB1157jYSEBObP///ZEllZWYSEhLBt2zZCQkLo27cvtra2HDx4kNWrVxtMmRcRIiIi\nsLa2Zvz48dy+fZu5c+cSFhbG8ePHqVKlitl7uHHjBi1btuTChQsMGzaMGjVqsHXrVlq1akV6erpR\n+Xnz5uHm5sZLL72Eh4cHp06dYsGCBbRo0YLff/8df3//QtsZICoqiuTkZCIiIvDy8uLChQssXLiQ\nNm3asGXLFpo3b2427m3bttG5c2fq1KnDxIkTKVu2LH/99RebN2/m5MmTVK1atUhtNHToUFJSUoiJ\nieHDDz+kXLlyANStW9dsHP9oIvLYH3lvg6IoiqIoimXMfXO4368Vx44dkzOeVURAfpr2tYn6yQ/h\nsf9uExcXJ5qmydtvvy3Xrl2Tq1evysGDB2X48OGiaZo0a9bMoHxGRoZ4eHhIUFCQ5ObmGuTNnTtX\ndDqdbN26VURE1q5dK5qmyapVq8zGEBkZKTqdTs6cOWOU5+PjI61atSpWmojIRx99JDqdTvbs2WM2\nBlNycnKkSpUq4uLiIpcuXdKnZ2VlSfPmzaVUqVJy8uRJfXpQUJDodDp5//33Dep57733RKfTycaN\nGwu9ZqdOnaRChQry119/GaTv3btXSpUqJW+99ZY+LSMjw2Qdffv2lVKlShnEPGvWLNE0TSZNmmT2\n+pGRkaJpmnTq1Mkgfffu3aJpmkycOLHQe3j99ddFp9PJ0qVLDdJHjx4tmqYZtVNaWppRHUePHhUb\nGxsZMWKEQXpB7VxQPVeuXBE3Nzd57rnnCo37lVdeEZ1OJ9euXTNbrihtZO6z/U9W0M/Q4jxjrCiK\noiiKojwgXl5e3NBVAuDqsU3cunXrgV8zLi6OuLi4v+11SZsyZQru7u6UL1+eunXrMn/+fLp160ZM\nTIxBuU2bNnH58mX69+9PUlISiYmJ+qN9+/aICBs3bgTA2TlvdfDY2Fhu3rz5wGI3x9nZGRFh9erV\nZGYWbfbA3r17OXfuHIMGDaJChQr69FKlSjFhwgRycnKM9snW6XT85z//MUgLDg5GRAqdTp2SksK6\ndevo1KkT1tbWBu9tlSpVqFq1qv69BbCxsdH/PSsri+TkZBITEwkJCSE3N5c9e/bo81esWIGrqytv\nvvlmofetaRovv/yyQVrDhg1xdHS0aEr4mjVrqFChAn379jVIf/XVV02Wv3vl+Js3b5KYmEi5cuUI\nCAhg165dhV7PVD2pqakkJSWhaRpNmjSxqJ78z+uqVasKXI29qG30uCnWVGpFURRFUZTH2RRzeeYy\nLWBvb0+6Xd70ynJaQ+zt7e+vQgsEBQX9ra9L2osvvkiPHj3Iysri4MGDzJo1i/Pnz2Nra2tQ7s8/\n/wTynp80RdM0Ll++DEBgYCARERFERUURHR1No0aNaNOmDT179qRGjRoP9H7yhYeH88UXXzBjxgzm\nzJlD06ZNadeuHeHh4YVOCU5ISACgZs2aRnm1atUCMHputlKlSlhbWxuk5U+hLWwRs2PHjpGbm8ui\nRYtYuHChUb6mafj5+elf5+TkMGPGDJYvX87JkycNFnrSNI3k5GT96xMnTtCgQQOj2Ari6+trlFau\nXDmLFmKLj4+ncePGRtPGPTw8KFu2rFH5ffv28eabb7J161ZSU1MN8u6+X0uuO3HiRDZu3Gi0CrRO\nV/hY5siRI1m7di0jRozg1VdfpUWLFrRv355evXrh5uYGFL2NHjeFdow1TYsHBGgjIgl3XltKRMS/\n2NEpiqIoiqI8giLN5ZnLtNDtMhXz/pKYatGX4sddtWrV9ItctWvXjubNm9OiRQuGDh3KihUr9OVE\nBE3TeP/996lXr57JuipVqqT/+5IlSxg/fjyxsbH8/PPPzJ49m+nTp/Phhx8yfPhwwPxzxNnZ2fd1\nX9bW1mzYsIE9e/awYcMGtm3bxpQpU4iMjOTLL7+kc+fO91X/vaysjHZi1bu742ouv0+fPkRERJgs\nc/eo6JgxY/j444/p1asXkyZNonz58pQuXZq9e/fy2muvmVyAy1IF3Udh91BU586d45lnnsHZ2Zkp\nU6ZQvXp1HBwc9KPW93aUC5KamkrLli1JT09nzJgx1K5dmzJlyqDT6XjnnXfYsmVLoXW4urqye/du\nfv75ZzZt2sS2bdsYM2YMU6ZMITY2liZNmhS5jR43lowY+5DXMS5912tLPV5rfCuKoiiKotynjIwM\nssqcBkB3fe/DDeYfqlmzZvTt25fly5czatQomjZtCuR1oEUEe3v7QleLzlezZk1q1qzJ2LFjSUlJ\noXHjxrz22mv6jnH+4l5JSUkGo7iZmZlcvHiRatWqFXqNwha1atiwIQ0bNuSNN97gwoUL1K9fn0mT\nJpntGOeP/B0+fNgoLz+tJEcHq1atiqZp3L5926L3Njo6mmeeeYYvvvjCIP348eNGZatXr87Ro0fJ\nysqidOnSRvklyc/PjxMnTuh/iZLv0qVLRiO5q1evJjU1le+//57AwECDvMTERKMZCwW1848//sjF\nixeJioqiX79+BnkTJ060OHZN0wgMDNTHcujQIZ588kmmTZvGd999V+Q2smTxuH8TS34F6UveFkzx\nd7229Hh8x+IVRVEURVGKwdraGp1b3hdbu4yEIm2Vo/y/N998E51Ox+TJk/Vp7dq1o3z58sycOdNg\nqm6+jIwM/TPdycnJRiOMTk5O+Pr6kpaWpn/mt3r16ogImzdvNig7e/Zsi0c9HR0dSUpKMko3NfXX\n09MTd3d3k+Xv9uSTT1KlShWWLFnClStX9OnZ2dm899576HS6Eh1xdnV15dlnn+Xbb78t8JnYa9eu\n6f9uZWVl9P6mpqYyd+5co/NeeOEFkpKSmDZtWonFW5DOnTtz+fJlli1bZpA+c+ZMo7L5I9P3tvPn\nn3/OpUuXjMoX1M4F1bNx40aLt+Uy9VmpXr06dnZ2+msWtY0cHR0BCv2s/VsUOmIsImfMvVYURVEU\nRVFKjk6nw9qjct6LJOfH+pm/++Hv7094eDgrVqzgl19+oXnz5tjb27Ns2TLCwsIICAhg4MCBVK1a\nlevXr/Pnn3+yevVqYmJiCAwMZNmyZcyZM4ewsDCqVq1K6dKliYuLY+PGjfTs2VO/eFSbNm0ICAhg\n8uTJXLt2DV9fX7Zv386uXbv0z3YWpmnTpixevJjJkydTo0YNdDodHTt2ZNq0aWzcuJEOHTrg6+uL\niLB27VqOHTtW4GJQ+XQ6HR9//DFdu3alYcOGvPjii5QpU4avvvqK3377jTfeeEO/lVBJmT9/Pi1b\ntiQwMJB+/frRoEEDcnNziY+PZ82aNUREROh/UdG9e3cWLFhAeHg4bdq04dKlSyxZssTke/byyy/z\n3XffMW3aNH777TdCQkKwtbXl8OHDHD9+vEQXjJowYQIrVqxgyJAh7Nmzh1q1ahEXF8evv/5qFFto\naCh2dnb06dOHkSNH4uLiwi+//EJsbCz+/v5Gi2CZaudOnTrRokULPDw8GDt2LAkJCXh5ebF//36W\nL19OnTp1OHToUKFxDxkyhPPnzxMSEoK3tzfp6el89dVX3Lp1y2DadFHaKH8v8AkTJvDCCy9ga2tL\n7dq19c+o/+uYWqr6cTt4TLY0UBRFURTln2HXW/8TAdnt0tYoD7Vdk15cXJzodDqZPXu2yfw///xT\nSpUqJcHBwQbphw8flr59+4qXl5fY2NiIh4eHNG/eXKZPny7JyckiIrJ//37p37+/VKtWTRwdHcXZ\n2Vnq168vc+bMkdu3bxvUd+LECQkNDRUHBwdxcXGR8PBw+euvv8THx8fo2qbSrly5It27d5dy5cqJ\nlZWVfoucuLg4CQ8PF19fX7G3t5dy5cpJ06ZNZfHixRa/R9u2bZOQkBBxdnYWOzs7efLJJ2XJkiVG\n5YKCgsTPz88o/fTp06LT6WTq1KkWXS8xMVEmTJggAQEBYmdnJy4uLlK3bl0ZM2aM/Pnnn/py6enp\nMmHCBPHx8RE7OzupXr26vPvuu/Ljjz+a3C4pMzNT3nnnHaldu7a+3saNG8unn36qL1PY1ln3vu8F\nOXfunPTo0UOcnZ3F2dlZOnfuLPHx8eLr62tUx88//ywtW7YUJycncXFxkY4dO8rhw4dNvp8FtbOI\nyMGDByU0NFRcXV3FyclJWrVqJdu3b5f+/fuLlZVVoTGvXr1aOnfuLJUrVxZbW1spX768BAUFyerV\nq43KWtpGInnbdfn7+4u1tbXodDqD7Zz+qQr6GapJCT2ErmmaDhgI1APOAJ+JyMNZ276INE2Tknof\nFEVRFEX594uk4AW4IiPvfwGuFaNfp/eHMzlb3psql08b5GmahoiYffhPfbdRFEUxraCfoUXuGGua\n9howGXhWROLuSo8FQgCNvEW3/gSaiIhly7E9ROo/D0VRFEVRiiL/y47JPA3u92vFoTW/U7vLUySV\ndeHw2jW0bNnyrvpVx1hRFKW4CvoZWpz1/9sBKcDWuyoPuZN+AZgG/AbUIG8EWVEURVEURSmCCjXy\n9mG1Sb2tX1FZURRFeXCK0zGuChy559eQ3cj7xWm4iEwGgoFkoPf9h6goiqIoivJ4Ke1mTS4aDlmp\nlOLx2jJFURTlYShOx9gNuHhPWgvgkojsABCRdGAHRdvzWFEURVEURQF27tpGuo0dAMnxfz3kaBRF\nUf79itMxzgUc8l9omuYMPAH8ck+5G0DZ4oemKIqiKIryeAoMDOSWdRkA4r795iFHoyiK8u9XnI5x\nAtDkzirUAB3IW4Ni+z3l3IFrKIqiKIqi/MtMMZdnLtNCDg4OpFnnjUPU9VXPGCuKojxoxekYrwUq\nAKs1TRsFvAfkAGvyC2iapgENyOtEK4qiKIqi/KtEmsszl2mh7OxsbuR4AJCVmHH/FSqKoihmFadj\nPIu8rZg6AnMBD+B9ETlzV5kW5I0Y3zuKrCiKoiiKohTi7Nmz2LhfByDx9MmHHI2iKMq/X5E7xiJy\nA2gIRAATgFYi8vo9xcoBHwJf3XeEiqIoiqIoj5kqVaqQnOsDQGL8locbjKIoymOgVHFOurPq9HIz\n+TFATHGDUhRFURRFeZyVKlWKbPu8Z4zLlW70kKNRFEX59yvOVGpFURRFURTlAcu2Kw+A3FDPGCuK\nojxohXaMNU2rcuewuue1RceDvwVFURRFUZS/V6S5PHOZRSBlTuT9JeVgyVSoKIqiFMiSEePTQDzg\nf9frBAuP+BKNVlEURVEU5RHwlrk8c5lFII7BAFjfTiA5OblkKlUURbHQH3/8QevWrXF1dUWn0zF1\n6tQinX/mzJkinde/f390uoc3odmSK58FzgFZd7229DhXwvEqiqIoiqI8FqzdKgCQk+RD2bJlH3I0\nj6atW7ei0+nQ6XSMGjXKZJmrV69ibW2NTqcjODj4b45QUUrG7t27GTVqFC1atKBMmTLodDqWLVtW\nYPnbt28zefJk/Pz8sLW1pWrVqkyfPp3s7GyLrpeTk0PXrl05deoU06ZNIzo6mq5du5bU7ZikaRp5\nu/4+HIUuviUiPuZeK4qiKIqiKA+AU94fVmnpD/XL4j+BnZ0dK1as4IMPPqB06dIGefmdh3vTFeWf\nZP369cyfP58nnniC+vXrs2PHDrPln3/+eb777jsGDRpE06ZN2blzJ2+++SanTp1i8eLFhV4vPj6e\nhIQE5syZw/Dhw0vqNh5pavEtRVEURVGUR9DFtP0AuFodf8iRPPrCwsJITk5mzZo1RnlRUVE899xz\nWFtbP4TI8ty6deuhXfvv8jjc44NU2Ps3fPhwUlJSOHjwIKNHj0ZECiy7fv161q5dy9ixY1mwYAED\nBw7k888/55VXXmHp0qX8+uuvhcZz8eJFAFxcXIp2I/9gqmOsKIqiKIryCHqiUTcA7HNucOTIkYcc\nzaPtySefpE6dOixZssQg/bfffuPIkSMMGDDA5HmbNm0iPDwcf39/7O3tcXFxoV27dmzbts1k+VOn\nTjFgwAAqV66MjY0Nnp6edOnShX379unL+Pj4EBwczP79+2nfvj1ly5alXr16+vzExERGjBhBlSpV\nsLGxoUqVKowcOZKkpCSL7vXYsWMMHz6c2rVr4+TkhIODAw0bNmTRokUG5T799FN0Oh3ff/+9UR0i\ngpeXF08++aRB+p49ewgLC8Pd3R1bW1ueeOIJ3nnnHXJycgzKBQUF4efnR0JCAt27d6dcuXI4Ozvr\n654+fTrPPPMMFStWxMbGBm9vb4YPH27yHtPT03nllVeoVKkS9vb2PP300/z0008FPm968uRJ+vbt\nS6VKlbCxscHX15cJEyaQlpZm0fsHEBMTQ/PmzXF0dKRMmTK0aNGCtWvXGpRp0qQJHh4e5ObmGp2/\nYcMGdDodH330kUH6/PnzadiwIQ4ODpQpU4bg4GDi4uIMytz93O3XX39Nw4YNsbe3L/BRgHzu7u7Y\n2dlZdH8rVqxA0zRefvllg/T8DnV0dLTZ81u1akVQUBCapunbwcrKirNnzwKQlpbG66+/TtWqVbG1\ntaVixYpERETo8wuTmZnJ+PHj8fT0xN7enqZNm7Jp0yaTZY8cOUKPHj3w8vLSXys4OJjY2FiLrlUU\nRd7HWNO0ZsAQYKGImBzD1zStOTAI+FREfru/EBVFURRFUR4tU8zlmcssAifPvO2aSLEiICCgZCr9\nFxs4cCBjx47l4sWLVKxYEYDFixdTvnx5OnToYPKcqKgokpOTiYiIwMvLiwsXLrBw4ULatGnDli1b\naN68ub7snj17aN26NTk5OQwePJhatWqRlJTE1q1b2bFjBw0aNADynpM8c+YMwcHBPP/883Tv3l0/\nGpiSkkKzZs2Ij49n0KBBNGjQgH379jF//ny2bNnCb7/9hoODg9n7jIuLY/v27XTs2BFfX19SU1NZ\ntWoVQ4YM4dq1a7z66qsAhIeHM2bMGJYtW2Z0/5s3b+avv/5iwoQJ+rR169bRrVs3qlWrxrhx43B1\ndWXnzp1MnjyZP/74g5UrV+rLaprGrVu3eOaZZ2jRogXvvPMOV65cAfKebX3//ffp3r07Xbp0wcHB\ngd27d7No0SJ++eUX9u7dS6lS/98F6d69Oz/88ANhYWG0bt2ahIQEwsLC8PX1NXqEYO/evbRu3RoX\nFxeGDh2Kp6cnf/zxBx999BE7duxg69atWFlZmX3/PvnkE0aOHEmNGjWYcucfa1RUFF26dGHBggUM\nHjwYyFsIauTIkfzwww88++yzBnUsW7aM0qVL07t3b31anz59WLlyJd27d2fgwIFkZmbyxRdf0LZt\nW1avXm3UBqtXr+bcuXMMGzaMYcOG4eTkZDbuotizZw+enp54enoapHt5eVGpUiV2795t9vxJkybR\nokULpk+fzksvvUTLli2BvM55dnY2ISEh7Ny5kx49ejBu3DhOnDjBJ598wqZNm9izZw+VKlUyW394\neDhr1qyhc+fOhISEcOrUKbp27Yqvr69BuaSkJFq1aoVOp2Po0KF4e3tz7do19uzZw65duwgNDS3G\nu2OGiBTpAJYA6UA5M2XcgAzyOs9FvsbffeS9DYqiKIqiKI+OU9uPioBc0ZU3SL/zvaXkvtvA33uU\noLi4ONE0TT744ANJTEwUGxsbmTFjhoiIpKenS9myZWXChAkiIuLo6CitWrUyOD8tLc2ozitXroib\nm5s899xzBum1atUSOzs7OXTokNmYfHx8RKfTyeLFi43yJk6cKDqdTj799FOD9Hnz5ommaTJ58uRC\n79lUzCIiQUFBUrZsWcnOztan9ejRQ+zs7OT69esGZfv06SPW1tZy9epVERHJyMgQDw8PCQoKktzc\nXIOyc+fOFZ1OJ1u3bjW4lk6nKzDejIwMo7TFixeLTqeTVatW6dPWrVsnmqbJSy+9ZFA2NjZWNE0T\nnU5nkF63bl2pWbOmpKamGqTHxMSIpmmydOlSk/HkS05OFkdHR6levbrcunVLn37z5k3x9/cXJycn\nuXHjhoiIJCUliY2NjfTs2dOgjps3b4qDg4N06dJFn/btt9+KpmmycOFCg7I5OTnSsGFD8fPz06ed\nPn1aNE0Ta2trOXbsmNl4C/LNN9+Yvd8yZcpI06ZNTeY1btxYPD09C71G/r+te6+xYMEC0TRNXnvt\nNYP0/Lbs16+fPi3/Xt966y192oYNG0TTNBk4cKDB+WvWrDFq87Vr14qmaQafmZJQ0M/Q4kylbg7s\nF5FEM53ta8A+oEUx6lcURVEURXnsHTmfN33aWae2arKEq6srnTp1IioqCoD//e9/pKSkMHDgwALP\nuXtqampqKklJSWiaRpMmTdi1a5c+b//+/Rw5coSBAwdSq1Yti2Lp37+/UXpMTAzu7u4MGTLEIP2l\nl17C3d2d1atXF1r33TFnZmaSlJREYmIiISEhpKSkcPToUX1+REQEGRkZBqO9qampxMTEEBoaipub\nG5A3pfzy5cv0799fX1/+0b59e0SEjRs3GsUyduxYkzHa2NgAkJuby40bN0hMTCQoKAgRMXhfv//+\nezRNY8yYMQbnt2/fnho1ahikHTp0iIMHD9KrVy/S09MNYnz66adxcHAwGePdNm3aRGpqKqNGjTIY\nmXd0dGTUqFHcunWLzZs3A3nP1nbs2JHvvvuOlJQUfdlVq1aRnp5ORESEPi06OhonJyc6depkEFdy\ncjIdO3bk9OnTnDx50iCWDh06UL16dbPxFldaWpq+De5la2tbpGnn94qJicHKyorXXnvNIP3ZZ5+l\nfv36Jp/zv9uaNWvQNI1x48YZpHfq1MloZkz+9PzY2Fhu3rxZ7JgtVZyOcSXgjAXlztwpqyiKoiiK\nohRRm07PkqtpWGdn8Uvc1gd3ob97zPgBGjBgACdOnOCXX35hyZIlNG7c2Ow09Pj4eMLDw3F1daVM\nmTK4ublRvnx51q9fb7B39IkTJwCoX7++RXH4+/ubXEk8ISGBgIAAo2dnraysqF69OvHx8YXWnZqa\nyrhx4/D29sbOzg43Nzfc3d154403AAzibt++PeXLlzfY1uebb74hLS2Nfv366dP+/PNPIO/9c3d3\nNzhq1KiBpmlcvnzZIA53d/cCp/9+/fXXNG3aFDs7O1xcXHB3d9e/J3fHl5CQgE6no2rVqkZ13Ntu\n+TFOmTLFKMYKFSqQlpZmFOO9EhIS0DSNmjVrGuXVqlULETFog4iICNLT0/n666/1acuWLcPFxcVg\navTRo0e5efMmFSpUMIotMjISwCi2atWqmY31ftjb25OZmWkyLyMjA3t7+2LXnZCQQKVKlfSd1rvV\nqlWLmzdvcu3atQLPj4+PR6fTmfylwL2/DAkMDCQiIoKoqCjc3Nxo0aIFkZGR+s9CSSvyM8ZADmBr\nQTlb1OJeiqIoiqIoxWJrZ0NqaTscbqdRs3KNwk9QaNeuHZUqVeKtt95iy5YtfPbZZwWWTU1N719X\nYgAAIABJREFUpWXLlqSnpzNmzBhq166t3x/2nXfeYcuWLcWO4346HoXp1asX69ev1z/7Wa5cOays\nrFi3bh1z5841WCzKysqK3r178+GHHxIfH4+fn5++Y9exY0d9ORFB0zTef/99g4XC7nbvc6MF3eO3\n335LeHg4TZo04aOPPqJy5crY2tqSk5NDu3btTC5mZcl2ZHLnlypjx46lffv2JsuU9ArKoaGhuLu7\ns2zZMgYPHszZs2fZtm0bw4cPN3hOWkRwd3fnyy+/1Md5r9q1axu8fpCfkUqVKnHhwgWTeRcuXDB6\n9vhRtmTJEsaPH09sbCw///wzs2fPZvr06Xz44Yclvo1UcTrGp4DmmqbZiIjJX0VommZD3pTrwn/t\npSiKoiiKohjJyMjghrjgQBqZV9PA/2FH9OjT6XT069ePGTNm4ODgQHh4eIFlf/zxRy5evEhUVJTB\n6CnAxIkTDV7nj27t37//vuLz8/Pj2LFj5ObmGowa5+TkcPz4cfz8/Myef+PGDdatW0dERATz5s0z\nyCtoGnFERARz587Vd+62bt3K0KFDDfZ1rlatGiKCvb09wcHB93GHedOK7ezsiIuLM5jOe+zYMaOy\nPj4+5ObmcuLECaMR4runhOfHCHmd/eLG6Ofnh4hw+PBhWrVqZZB3+PBhNE0zaIP8Xyx89NFHnD59\nmi+++ALA6PNSrVo11q9fT5MmTR5oh9dSjRo1YsWKFUad4PPnz/PXX3/RpUuXYtft5+fHhg0bSElJ\nMZoxcOTIEZycnPRT9As6Pzc3l+PHjxuNEBe0+n7NmjWpWbMmY8eOJSUlhcaNG/Paa6+VeMe4OCO6\n3wPlgA/MlHkfcAW+K05QiqIoiqIoj7JIc3nmMovg2rVrWJXNAiDptOnRH8XY0KFDiYyMZP78+Tg6\nOhZYLn/14ntHMDdu3MhvvxluqlKvXj1q1arF4sWL72vrrC5dunD16lUWLlxokL5gwQKuXr1K165d\nzZ5fUMwXL1402q7p7tjr1q1LdHQ0y5cvR0QMno+FvJH28uXLM3PmTIOpzvkyMjIs3qfYysoKTdOM\ntnh6++23jUaGO3bsiIgwZ84cg/T169cbdYwbNGhA7dq1+fTTT0lISDC6bk5OjsnY79a2bVscHBz4\n73//a3A/N2/e5L///S+Ojo60bdvW4Jz892rp0qVER0cTEBBAo0aNDMr069ePnJwco+du8+Wv2P13\n6dWrFyLC3LlzDdLnzJmDpmm88MILxa67S5cu5OTkMHPmTIP02NhY9u3bR+fOnc2e37lzZ0SE9957\nzyA9JiaG48cN92xPTk42GoF3cnLC19eXtLS0AqeLF1dxRoznAoOBYZqm1SNvler8T24AMBB4GrgC\nzDFZg6IoiqIoyj/YWxTcOX7rrZLpHHt6enIKFypwhVP7tlIzvHnhJylUrlyZyZMnF1quRYsWeHh4\nMHbsWBISEvDy8mL//v0sX76cOnXqcOjQIYPyS5YsoU2bNjRu3JhBgwZRu3Ztrl+/ztatWwkNDWXE\niBGFXnPChAmsWrWKESNGsHfvXho0aMDvv//O4sWLqVGjBuPHjzd7vqOjIyEhIURHR2Nra0ujRo04\nffo0CxYswM/Pr8C9kCMiIhg7diyzZs2ievXqRh07e3t7li1bRlhYGAEBAQwcOJCqVaty/fp1/vzz\nT1avXk1MTAyBgYGF3mP37t359ttvadWqFf369SMrK4uYmBjS09ONOjnPPvss7dq14/PPP+fq1au0\nadOG+Ph4Pv/8c+rVq8eBAwcMyi9fvpzWrVtTt25d/UJoaWlpnDx5km+//ZaZM2cajebezdnZmXff\nfZeRI0fSpEkT+vfvj4iwdOlS4uPjWbBgAWXKlDE4p379+tSuXZs5c+Zw8+ZNZsyYYVRvt27dGDBg\nAPPmzeP333+nQ4cOuLm5cf78eXbu3MmpU6eMFt8qqrNnz7J8+XIgb3QbYO3atZw7dw7I65xXrlwZ\nyHtfO3TowOzZs7l+/TrNmjVjx44dLF68mH79+tGsWbNix9G/f3+WLl3KrFmzSEhIIDAwkBMnTjB/\n/nwqVqzI9OnTzZ4fEhJCx44dWbp0qX5xt5MnT7JgwQJq166tvzfIe557zpw5hIWFUbVqVUqXLk1c\nXBwbN26kZ8+eBS4wVmymlqou7AAaAOeAXPKeOb77yAXOA08Vp+6HcaC2a1IURVEUpQjMfXMoya8V\nBzyfEgHZN33tXfWb3mpEHsPvNnFxcaLT6WT27NmFli1TpowEBwcbpB08eFBCQ0PF1dVVnJycpFWr\nVrJ9+3bp37+/WFlZGdVx/Phx6du3r1SsWFFsbGzE09NTwsLCZN++ffoyPj4+Rte527Vr12TEiBFS\nuXJlsba2lsqVK8t//vMfSUxMtOieExMTZciQIeLp6Sl2dnZSt25dWbRokURFRRltq5Tv8uXLYm1t\nLVZWVvrtrEw5fPiw9O3bV7y8vMTGxkY8PDykefPmMn36dElOTtaXCwoKMtiC6F4LFy7Ub29VqVIl\nGTp0qCQnJ4tOpzPapictLU3GjBkjHh4eYm9vL02bNpUtW7ZI9+7dxcHBwajus2fPyrBhw8TX11ds\nbGzEzc1NGjZsKG+88YacP3/ekrdQYmJipHnz5uLo6CiOjo7SvHlzWbt2bYHlP/jgA9HpdFK6dGmz\n14iOjpbAwEBxdnYWOzs78fX1lW7duhlsN3T69GnR6XQydepUi2LNl799kk6nM3nc2+6ZmZny5ptv\niq+vr9ja2oq/v79Mnz7dYDuvwq6n0+lMbgmVlpYmEydOFH9/f7GxsZEKFSpIRESEnD171qBcQfea\nkZEh48aNk4oVK4q9vb00adJENm3aZPTvbv/+/dK/f3+pVq2aODo6irOzs9SvX1/mzJkjt2/ftvSt\nM1LQz1BNirk6oKZpDsAQoB3gDQhwFthA3v7Fls23eARomibFfR8URVEURXn8aOR98TGZp5XM4ssi\nwi7v52h6LpY9476i4Xs979SvISJmVytS322Uf7q6deuSnZ19X1PXFcWUgn6GFmcqNQAikkretOq5\nhZVVFEVRFEVRikbTNGxcLsE5uHn5eOEnKMo/UEZGBra2hhverFu3jkOHDjFq1KiHFJXyOCp2x1hR\nFEVRFEV5sFLFD9hHZtKhQssqyj/R1KlT2bdvH61atcLZ2Zl9+/axZMkS3N3dmTBhwsMOT3mMFLtj\nrGmaM9AHaAa4Az+KyLt38qoDPsDPIpJeAnEqiqIoiqI8MqaYyzOXWUTZNq4A2OWa3ltWUf7pWrZs\nyY4dO3j//fe5ceMGrq6u9OjRg6lTpxrtnawoD1KxOsaaprUHvgDK8v+P2dy9j0AAEAP0BlbeZ4yK\noiiKoiiPlEhzeeYyi8ohb59QuZVRgpUqyqMjNDSU0NDQhx2GohR9H2NN02oD3wJlgE+AnuR1ju/2\nA5AGmN/ISlEURVEURSlQts0ZAEpl/v6QI1EURfl3K3LHGJgI2ADdReQ/IrLq3gIikgXsA9S8H0VR\nFEVRlGIqXbYhANY5Vx9yJIqiKP9uxekYBwH7RGRtIeUuABWLUb+iKIqiKIoCWJctD0BWcuWHHImi\nKMq/W3E6xuWAkxaUswbsilG/oiiKoiiKAuB45880tSexoijKg1ScjnEy4GVBOX/gcjHqVxRFURRF\neaRFmsszl1lEKdkXAShnc7TkKlUURVGMFKdj/BvQSNO0agUV0DStEVAX+KW4gSmKoiiKojyq3jKX\nZy6ziCpVawqAXe6tkqtUURRFMVKcjvE8oDTwjaZpAfdmaprmBywmbwun+fcXnqIoiqIoyuPLzi1v\nu6bsFNuHHImiKMq/W5E7xiKyAfgvUAc4omnaAfI6wW00TdsFHAVqAXNEZHtJBqsoiqIoivI4Efs7\nO2Km3bszpqIoilKSijNijIi8DAwn7xni2uTtY+wFNAJuAKNFZFxJBakoiqIoivI4ul3qNgDlbC89\n5EgURVEevqCgIPz8/AzS+vfvj05XrG6tgWLXICKfktcZbgg8D/QCWgAVReSj+45MURRFURTlMedd\nM29JF+usTNLT0x9yNI+mK1euMHToUKpUqYKNjQ3e3t6MHj2aGzdumCx//PhxunTpgqurK46OjgQG\nBrJlyxajcvHx8bRv3x5nZ2f8/f356CPTX29HjRpFgwYNyM3NLdH7UkqepW1vzqpVq2jevDmOjo44\nOTkRGBhIbGysybI6nc7k4eTkZHHZ/GPGjBn6sqmpqbz00ktUqFABDw8Phg8fbvLnw7fffoujoyNn\nzpwp0j0+yjTNePaMpmkm04uqVDGC6QRkiUisiOQCv985FEVRFEVRHgtTzOWZyywiOzcHAEplZlPK\nRj1nfK+rV6/SuHFjLl26xNChQ6lVqxaHDh1i/vz5/Pzzz/zyyy/Y2v7/+xYfH0+zZs2wtrbmtdde\nw8nJic8//5x27drxww8/EBwcDICI0KVLFzIzM5k1axaHDx9m9OjRVK5cmbCwMH19u3btYsGCBezY\nsaNERqyUB8fStjdn1qxZvP766zz55JNMmzYNgOjoaDp06EB0dDS9evUyOicwMJAXX3zRIK106dJG\n5aKjo01ec8qUKcTHx9OxY0d92oQJE/jqq6+YOHEiIsI777xD6dKl+fDDD/VlUlJSGDVqFNOnT8fb\n27vQe1PI+4dflAPIATYU9bxH+ch7GxRFURRFUR4ttzMzJQdNBCQtJUtERO58b1HfbUTk5ZdfFp1O\nJytXrjRI//LLL0XTNJk+fbpBeo8ePaRUqVJy4MABfdqtW7fE29tbnnjiCX3asWPHRNM02bZtmz7t\nmWeekT59+uhfZ2VlSd26dWXs2LElfVuPtKysLMnIyHjYYRSZpW1fkMuXL4uNjY3Uq1dPsrOz9enZ\n2dlSp04dKVeunNy8edPgHE3TZMCAAcWO+fz582JlZSVNmjQxSK9YsaJMnTpV/3rKlCni5eVlUOal\nl16SRo0aSW5ubrGvn+9RavOgoCDx9fU1SOvfv7/odDqL6yjoZ2hxfrWVBFwroX65oiiKoiiKUoCs\n7GyyrPNGl5LPJz7kaB49cXFx2NnZ8fzzzxuk9+zZE1tbW5YsWaJPS0tL47vvvqNVq1bUqVNHn+7g\n4MDgwYM5fvw4e/bsAdBPS3VxcdGXc3V1JTU1Vf961qxZ3Lx5k7fffrvIce/evZsBAwYQEBCAg4MD\nTk5OtGjRgpiYGJPlL1++zKhRo/D398fW1pYKFSoQEhLCjz/+aFDu1KlTDBgwgMqVK2NjY4Onpydd\nunRh3759+jI6nY6BAwcaXWPp0qXodDq2bdumT4uMjESn03HkyBFeeeUVKleujJ2dHbt27QJg5cqV\ndO7cGW9vb2xtbXF3dycsLIyDBw+avI/9+/fTo0cPPDw8sLW1pUqVKvTu3ZuEhASysrIoX748LVu2\nNHnue++9h06nY/v2oq/tW5S2L8iOHTu4ffs2vXv3xsrKSp9uZWVF7969SU5OZs2aNSbPzcrKMvjs\nWGrx4sWICIMHDzZIT09PN/vZ3L59O1FRUSxatKjIU4wLa3OAzZs3065dO1xcXLCzs6NevXp89tln\nJusz1+b5ivo5elCKPJWavH2Ma5d0IIqiKIqiKIohe3t7bljZYsNtrp29SKUaFR52SI+UzMxMg6nS\n+TRNw87Ojvj4eJKSknB1deXAgQNkZmbStGlTo/JNmzZFRNi9ezcNGzYkICAAV1dXpk6dyrvvvsvh\nw4f54Ycf9J3g48ePM336dGJiYrCzsyty3KtXr+bYsWP07NkTb29vEhMTWbp0KV27dmXFihWEh4fr\ny545c4ann36aq1evEhERwVNPPUVqaiq//vormzdvpnXr1gDs2bOH1q1bk5OTw+DBg6lVqxZJSUls\n3bqVHTt20KBBg0LjurcTlf/s5gsvvIC9vT3jxo1D0zQqVqwIwLx583Bzc+Oll17Cw8ODU6dOsWDB\nAlq0aMHvv/+Ov7+/vq7vv/+e7t274+joyJAhQ/D39+fSpUts2LCBQ4cO4evrS0REBLNnz+bEiRNU\nq1bNIJYlS5bwxBNP0KJFiyK/30Vp+4JkZmYCef8m72Vvb4+I8Ouvv/LCCy8Y5H3zzTcsX76cnJwc\n3N3d6dmzJ9OmTTP5nPG9oqKicHBwMPg8ADRr1oxPP/2Uli1bIiLMnz+f5s2bA3D79m1efPFFXnnl\nFYNfAliqsDZfsGABw4YNo1mzZkyaNAkHBwc2bdrEsGHDiI+PZ9asWfq6LGlzKNrn6IEyNYxs7gAC\ngWxgUFHPfVQPHpPpRoqiKIqi/PNcKeMuAnJ8/XERUVOp79atWzfR6XTyxx9/GKTv27dPNE0TnU4n\n+/btExGR//3vf6Jpmnz66adG9Rw5ckQ0TZM33nhDnxYTEyNly5bV1/Pss89KZmamiBhPqy6qtLQ0\no7T09HQJCAiQWrVqGaSHhoaKTqeTTZs2ma2zVq1aYmdnJ4cOHTJbrqDpvVFRUaLT6WTr1q36tMjI\nSNE0TYKDgyUnJ8ei+zh69KjY2NjIiBEjDMq5ubmJh4eHXLx4scDYjh8/LpqmyauvvmqQvn37dtE0\nTd5//32z91aQora9KQcPHhRN0yQsLMwor3PnzqJpmnTu3NkgvWnTpjJ79mxZs2aNLF++XHr16iWa\npkm9evUkNTXV7PU2b94smqbJoEGDjPKOHTsmAQEBotPpRKfTyRNPPCEnTpwQEZHJkydLtWrVij31\n2VybX7x4UWxtbU1+9l9++WUpVaqUJCQkiIjlbZ5f9l6mPkcij95UaoBPgQWapsVqmvaipmntNE0L\nNHXcf9ddURRFURTl8XVL8qZM3k5Oe2DXiIwETTM+IiNLtnxB+cU1evRoNE2jR48exMbGcu7cOWJj\nYwkPD8fa2hrIm0Z79582NjZG9eSPOueXAejcuTPnz5/nt99+48SJE6xbtw5ra2sWLlzI4cOHmTt3\nLhkZGYwaNQpfX19q167NvHnzLIr77lHm9PR0kpKSuHXrFsHBwfz555/cunULgOTkZDZs2EBoaCht\n2rQpsL79+/dz5MgRBg4cSK1atSyKwVKapjF69GiTi4vdfR83b94kMTGRcuXKERAQYDD1dsOGDSQm\nJjJ27Fg8PDwKvFa1atV45plnWLZsmcEq34sWLaJ06dL069evWPdQ1LY3pXbt2rRt25Y1a9bw6quv\ncvToUY4ePcqrr77KDz/8gKZpRnXs3LmTMWPG0KlTJ/r06cOKFSuYPn06Bw4cMFgoy5SFCxeiaRqD\nBg0yyqtevTqHDx/mjz/+YP/+/Rw6dIiqVaty5MgRZs2axWeffYaNjQ2ffPIJderUwcfHh1GjRulH\nvQtTUJuvWrWK27dvM3DgQBITEw2ODh06kJOTw+bNmwH44YcfLGpzsPxz9KAVp2McBwwjb+/idsB8\nYD2wxcTxU4lEqSiKoiiK8giJNJdnLrMYrJzynndNPn+6ZCu+S2QkiBgf5jq6xSlf0u9NixYtWLly\nJbdu3aJDhw54e3vTuXNngoODee655wD0U1bzp8Ca6hxkZGQYlMnn4OBAw4YN9fumXrp0iQkTJjB7\n9mzKlSvHmDFjiI2NJTo6mkmTJjF+/Hi++eabQuO+evUqL774Ih4eHjg4OODm5kb58uX59NNPAbh+\n/ToAJ0+eRESoX7++2fpOnDgBUGi54rp3WnO+ffv20aFDB8qUKYOzszPu7u6UL1+egwcPkpycbBCf\npmkWxffiiy9y+fJlvv/+ewBu3brFqlWr6NixI+7u7sWKvzhtb8rXX39N165d+eCDD6hZsyY1a9bk\nm2++4ZNPPkFELJoePX78eKytrVm3bl2BZZKTk4mJiaFGjRo0a9bMZBkrKytq165NnTp19M88Dxky\nhBdeeIFWrVqxcuVKxo0bx6RJk/jiiy9Yv349Y8eOLTS+fKba/OjRo4gIrVu3xt3d3eAICQlB0zQu\nX74M5H12LW1zSz9HD1pxnjHeBkhJB6IoiqIoivJP8RYFd47feqtkO4BpuADnSL16vuQq/Rfp1q0b\nXbt25eDBg9y8eZOAgADc3Nxo0qQJpUqVomrVqgBUqlQJgAsXLhjVkZ/m6elp9lqjRo2iUaNG9O3b\nFxFh6dKlzJs3T/9857p161i0aBHdu3c3W0/btm05duwYo0eP5qmnnsLZ2RkrKysWL17Ml19++VD2\nRM7Ozi4wz1Sn8dy5czzzzDM4OzszZcoUqlevjoODA5qm8fLLLxdrsSnIa89Ro0axaNEiOnXqxFdf\nfUVaWprRAlRFURJtD+Ds7MyqVau4evUqx48fx9HRkXr16un3MX7iiScKraNUqVJUqlSJa9cKXss4\nOjqazMxMk6PFBfnkk084deqUvsO9ePFiunfvTs+ePQF4/fXXGTVqFB9//LFF9ZlqcxFB0zSWL19e\n4Chw/i+RLPWgPkfFUeSOsYgEPYA4FEVRFEVRFBOy7kz/dKTiQ47k0aVpGnXr1tW/vnTpEvv27aNV\nq1b6qbJ16tTBxsaGnTt3Gp2/c+dONE0zu/jSd999x/r16zl06BAA165dIyMjAy8vL32ZypUrG6wA\nbcqBAwc4cOAAkZGRTJ482SBvwYIFBq+rVq2Kpmns37/fbJ3Vq1cHKLQc5K1gnJSUZJR+6tSpQs+9\n2+rVq0lNTeX7778nMNDw6cnExESDRdGqV6+OiLB//36zU8IBrK2t6devH//973+5ePEiixYtwtPT\nk3bt2hUpvrvdb9vfK3+UNN+6devQNI1nn3220HMzMzM5f/58gSPBkDd13Nramr59+1oUz4ULF5g4\ncSKfffYZZcuWBeD8+fMG91S5cmUyMjK4du0abm5uFtV7r/xR5HLlyhW677OlbV6Uz9GDZvFUak3T\nymqa1lPTtAmapr2snh9WFEVRFEV58NK0vNGunJsZDzmSfwYRYdSoUeTm5vLGG2/o0x0cHOjYsSNx\ncXEG28DcunWLhQsXUr16dRo1amSyzlu3bjFixAgiIyPx8fEB8joH1tbWBnUdPHhQPzpZkPxpr/eO\nCh86dMhouyYXFxdCQ0OJjY012prpbvXq1aNWrVosXryYI0eOmL1+9erV2blzp34KMeRN3Y2KijJ7\nnqX38fnnn3Pp0iWDtJCQENzc3Pjggw+M8kwZMmQI2dnZvPrqq+zatYsBAwYUeduhuxW17VNSUjh2\n7BiJiYVvkbZnzx4WLVpEUFAQTz/9tD7d1C8fACZNmkROTg6dOnUymb93714OHDhAp06dLO7Ajhgx\nghYtWuhHhyFvlPzuez1w4ADW1tbF7hQDPP/881hbWzNlyhSDz0++lJQUbt++DVje5kX5HD1oFo0Y\na5r2PPAZ4HRP+j4gTETOPYDYFEVRFEVRHntWZa4CkJF4+CFH8uhJTU2lcePGhIWF4evry40bN/jy\nyy/5/fffeeedd4xGoGbMmMFPP/1E27ZtGTNmDE5OTixYsICLFy+yfv36Aq/z+uuv4+bmxiuvvKJP\n0+l09OrVi7fffpvc3FwuXLhAbGxsoR3MGjVqUKtWLd59911SU1MJCAjg2LFjLFiwgLp167J3716D\n8h9//DHNmzcnNDRUv11Teno6u3btwtfXlxkzZgB52xm1adOGxo0bM2jQIGrXrs3169fZunUroaGh\njBgxAoCRI0fSp08fWrVqRd++fUlOTmbhwoX4+Pjonw+1RGhoKHZ2dvTp04eRI0fi4uLCL7/8Qmxs\nLP7+/uTk5OjL2tnZsWjRInr06EHt2rUZPHgwVatW5cqVK2zcuJGxY8fSsWNHffn8bZmio6PR6XQM\nGDDAZAw+Pj6cO3fO4FoFKUrbr169mgEDBhiN6k+ePJkTJ07QuHFjnJ2d2bt3L1FRUVSuXJlly5YZ\n1DFt2jR+/fVXWrVqRZUqVbh16xbr169ny5YtNGvWjJEjR5qMM3/RLUunjv/vf//jp59+4vBhw58P\nffr0YdCgQYwZMwZPT0+mTZtmtJVUUXl6ejJ//nyGDBlCjRo16Nu3L97e3ly9epUDBw6wdu1ajhw5\nQpUqVSxu86J8jh44U0tV330A9YDbQC5wE9gLnCBvy6Zc4LfC6njUDx6TLQ0URVEURSkZ5r45lPTX\nip+fDBcBiW3X/079arumfLdv35bevXuLn5+f2NnZSbly5aR9+/ZmtzY6evSodOnSRVxcXMTBwUEC\nAwPlp59+KrD8r7/+KjY2NvL7778b5aWkpMiAAQPEzc1NvL29ZdasWRbFffbsWXn++eelfPny4uDg\nIE2aNJGYmBiJjIwUnU4nZ86cMSj/119/ybBhw8Tb21tsbGzEw8ND2rVrZxT38ePHpW/fvlKxYkWx\nsbERT09PCQsL029Zle/9998XHx8fsbW1lZo1a0pUVFSB2zWZiiffzz//LC1bthQnJydxcXGRjh07\nyuHDhyUoKEj8/PyMyu/evVvCwsLE3d1dbG1txdvbW/r27avf4uduy5cvF03TpG3btgW+j25ublK5\ncuUC8+9ladvnvxdTp041SF+9erU0b95cypUrJ3Z2dhIQECCvv/663Lhxw6iONWvWSPv27cXLy0vs\n7OzE0dFRGjRoIDNnztRv+3Wv9PR0KVu2rPj4+Fh0Pzdu3JBKlSrJhx9+aDJ/5syZ4u3tLW5ubjJg\nwAC5efNmoXUW1uYiIjt27JCuXbtKhQoV9J+z4OBgmTNnjtG9WdLmRfkcmUrr37+/WFlZFXpv+Qr6\nGarl5RVM07SlQF8gGhguIrfupNcH/gf4AK1FJO5+Ouha3vyI0cCLd+q8CnwNTBaRIu1PoGmaHXD4\nTj0fi8ioQspLYe+DoiiKoihKvkgKXnwrMrJkF9/a1rw3gTu+5KfgKQT/GImmaYiI2Xml6ruN8k+3\natUqevbsyVdffcXzzz9vlH/gwAHq169PVFRUsbdxUh5PBf0MteQZ45bARWBIfqcYQET2A2PI27ap\nZQnEOBf4ADgEjCSvUzwKWFuMut4GyqFWz1YURVEU5QGINJdnLrMYcmy9AdBl3C7ZihWi6QaRAAAg\nAElEQVTlEfbxxx/j7u5OWFiYyfyNGzdSv3591SlWSowlzxhXBDaIiKkdobfd+dP8KgOF0DStJnmd\n4W9E5Pm70k8DH2maFi4iX1lY15PAy8B4YPb9xKUoiqIoivKwZdnkPfdZ+rZ6xlj5d7t69SqbN2/m\n559/Zvv27cycOZPSpUubLDtu3DjGjRv3N0eo/JtZMmJsA5hcVk1Ert9V5n70vvPn/7V35/FRVff/\nx9+fSUISgrJFQBAQFKKAWCviV7GKorjV1q3uVqsFxQVx+Vl3AkVqrQr6daFaVKyKuFRaKtiqCFr9\nalFRUBBEWRSLioBACBCSz++PO0mzTCaZZIaZZF7Px2MeQ84598xn5s4M9zPn3HMnVit/RNIWSefV\npxMzC4W3mSnpxUbGBAAAkHSWt5ckqcWO2q97CjQHixYt0rnnnqtp06ZpxIgRVRY7AxIt5usYJ8gA\nBQt5zatc6O7bzOxDSZHXzq/pGkm9JZ2sGC5FBQAAkKrKcoNT4co2dUpyJEBiHXHEETUu2wPsLPVN\njDvVcd3iWuvd/Y1I5dV0lrTW3Usi1K2WdIiZZbr7jto6MLMeCq+F4e5fmln3ejwuAABASstuu6ck\nKaO41sMgAEAj1TcxPjZ8i8Sj1Hs9H6OlpEjnMEvS1kptNkbpY5KkZZIm1OPxAAAAGqxQO29V6i0W\nnNG2a+ay+HUKAKiiPknrKiV+dectknarpS6nUpuIzOw8SUMk/cTdG3QV6MJK/4MNHjxYgwcPbkg3\nAAAgDYxR7YnxmDHxTYw/27RJhZI2rF+tNvFe8hoAIEl1X8d4pwRh9rKCxLZl9enUZvYvSb3cvWMt\n27aQ9KWkdxVcPqrcHpJel/RnSWMVTNX+oZY+uNYfAACoN1PtowZmUjwPK/51z0wddu2JWtp6f/Xe\n8CHXMQaARmjMdYx3hnkKYhlYudDMsiX9SNUW5aomV8Fo84mSPqt0e13B/1nnS1oq6eK4Rw0AAJBg\nee2DsYHQ9gZNigMA1EOqrEo9TdJNkkZJeqtS+XAFie9T5QVm1lNSlrsvCRcVSTo9Qp+7SXpI0ixJ\nf5K0MP5hAwAAJFZpVrDoVuvsNUmOBACar5RIjN39YzN7QNLlZvaCgusQ95F0paQ57j61UvPZkrpK\nyghvu0PSX6r3WWlV6s/dnWsaAwCAJqlNpw6SpKyy7UmOBACar5RIjMOukrRcwSjxCZLWSrpX0uhq\n7Vz1XwwslrYAAAD1Uv3gpEpdtMoGyG3TUpJUtiWVDtsAoHlJlXOM5YEJ7r6vu+e6e1d3/3/uvqVa\nux7uXuf/DO6+0t0z3P2qxEUNAADSUWG0umiVDZDTJrhAR1ZpSR0tAaB5KywsVCgU0qpVqyrKpkyZ\nolAopDfeeKNRfadMYgwAAICayhPj3NAWsdJ0Td9++60uvfRSdevWTdnZ2erevbtGjRqlH36oeTGS\nMWPGKBQK1bhlZGTonnvuqdK2qKhIl1xyiTp27KhOnTrpsssuU3FxcY0+//KXv6hVq1ZauXJlwp4j\n4mPjxo268sortcceeyg3N1f9+vXTpEmTYurjgw8+0M9//nPl5+dX9HHvvfeqrKysRtvBgwfX+n77\n4IMPqrQ98sgjI7Ytvx177LFV2k+YMEE9e/ZUmzZtdOKJJ2r58uU1Hn/VqlXadddd9fzzz8f0HFOZ\nmcms5qL8kcpixZwcAACAFJazawtJUmZpqcpKax58p7PvvvtOAwcO1Jo1a3TppZeqb9+++vjjj/XQ\nQw/pzTff1FtvvaWcnJwq25iZJk6cqPbt21cpP/DAA6v8ff311+uZZ57RTTfdJHfX+PHjlZWVpXvv\nvbeizcaNGzVy5Ejdfvvt6t69u5C6SkpKdPTRR+ujjz7SyJEjtc8++2jWrFm67LLL9O233+q2226r\ns4833nhDxx57rNq0aaNRo0YpPz9fr7zyiq6++motXry4RpJtZtptt900ceLEGj9q9ezZs8rft9xy\ni4YNG1bjMZ955hm99NJL+tnPflZR9txzz+naa6/VFVdcoT59+uiee+7RaaedViPZHjFihIYMGaLT\nT4+0TjFqcPe0v6liJjcAAEDq2Z6R6S751vVbPHzcwrGNu1911VUeCoV82rRpVcqnTp3qZua33357\nlfLCwkIPhUK+cuXKOvvefffdfezYsRV/jx492vfYY48qbS655BI/6KCDvKysrBHPounZvHlzskOI\n2QMPPOBm5g888ECV8tNOO82zs7N91apVdfax//77e15enq9YsaJK+SWXXOKhUMjfeuutKuWDBw/2\nHj16NCruffbZx3Nzc339+vUVZWeffbYfeeSRFX+//vrrHgqFfNmyZRVlU6dO9datW/vXX3/dqMcv\nlyr7PNJn+PHHH/dQKORz586tVx+1fYcylRoAACDFFXuuJGnrhq1JjiS1zJkzR7m5uTrjjDOqlJ95\n5pnKycnRY489FnE7d9emTZtUWlr7taGLi4vVtm3bir/btWunoqKiir//9a9/6fHHH9fkyZNjnsb5\nn//8R9dee60OOOAAtWvXTrm5uerbt6/uvPPOiFNyS0pKdOedd+qAAw5QXl6e2rRpo4MOOkgPPPBA\nlXabNm3SzTffrD59+ig3N1f5+fn6yU9+omnTplW0GTx4cI3RSklauXKlQqGQxo4dW1E2d+5chUIh\nPfHEE3rggQfUt29f5eTk6O6775YkzZs3T7/61a9UUFCgvLw87brrrjrssMM0ffr0iM/7m2++0ciR\nI7XXXnspJydHHTt21NChQ/Xaa69Jkk4++WTl5eVp8+bNNbadN2+eQqGQxo0bV49XuKann35aeXl5\n+vWvf12lfNSoUdq+fXuV1yiSDRs2aMGCBTr88MNrzA648MIL5e51vt9i9eabb2rJkiU69dRT1aZN\nm4ry4uJitWvXruLvtm3byt0r3p/r16/XqFGj9Lvf/U677757TI9Z1z6XpGXLlun8889X586dlZ2d\nrR49euj666/Xli1bavRX2z5/9dVXK9rE+j5KFKZSAwAAxKhQtS/AVVgY/wW4LNelImndf76Lb8dN\n3LZt22pMlZaCKay5ubn64osvtG7duipJhLurf//+2rRpkzIyMjRw4EDdeuutOu6446r0ccghh2jS\npEn6yU9+InfXQw89pEGDBkmStm/fruHDh+uaa67RfvvtF3PcCxYs0PTp03XKKador732UklJiV5+\n+WXdcMMNWr58uR566KGKtiUlJRo6dKjeeOMNDR06VOeff75ycnK0cOFCvfjii7r88sslST/88IMG\nDRqkxYsX6xe/+IUuu+wylZaWav78+XrppZd05plnVrw2sZowYYLWrVunYcOGqVOnTuratask6cUX\nX9SSJUt05plnqnv37vr+++81ZcoUnXrqqXr66ad11llnVfSxcuVKHXroofruu+90wQUX6MADD1RR\nUZHeeecdvfrqqxoyZIiGDRumGTNmaOrUqTWmFU+ePFkZGRm6+OKLY47f3TV//nwdeOCBatGiRZW6\ngQMHysw0b968qH1s27ZNktSyZcsadeVl77zzTo261atXq1WrViouLlbLli117LHHavz48SooKKgz\n7vIfXaon84cccojGjBmjl19+WQUFBRo7dqzat29f0ee1116rvfbaSyNGjKjzMWpT2z5///33NWTI\nELVt21aXXnqpunTpoo8++kj33Xef3n77bc2dO1cZGRmSou/z1157TUcffbSk2N5HCRVpGDndbkqT\n6UYAACA+oh05JOKwYm2rdu6SL3t1CVOpKznttNM8FAr5Rx99VKV8/vz5bmYeCoV8/vz5FeUTJ070\nSy+91J944gmfMWOG33XXXb7HHnt4KBTyKVOmVOljyZIlXlBQ4KFQyEOhkO+zzz7+2Wefubv7bbfd\n5r169fKtW7c2KO7atjv//PM9MzPT16xZU1H2+9//3s3Mb7nllqh9jhgxwkOhkP/pT3+K2q626b0r\nVqxwM/MxY8ZUlM2ZM8fNzNu3b+9r166tsc2WLVtqlBUXF3tBQYH37du3Svnxxx/voVDIX3nllVpj\nKy0t9W7duvnBBx9c43Fat27tP/3pT6M+t9p8//33bmZ+1llnRazv0KGDDxo0qM5+dtttN+/SpUuN\n/TdhwgQ3M2/dunWV8osuushvueUWf/bZZ/2FF17w66+/3nNzc71Nmzb+8ccfR32sjRs3el5enu+1\n11416oqLiyteTzPztm3b+owZM9zd/bXXXvOcnBxftGhRnc8nkrr2ef/+/b1Pnz5eVFRUpXz69Olu\nZlU+R/XZ5+6xvY8SOZU66UlpKtzS5T8PAAAQHzs7Mf5P607uki/926KEJMaja+lodJzb11bfUG++\n+aZnZGR47969febMmb5q1SqfOXOmFxQUeHZ2dsTzPqtbt26d77777t6uXbsaB/s7duzwhQsX+oIF\nC3zHjh3u7v7JJ594dna2z549292Dc1f79evn3bt39yuvvDLmZHn79u2+bt06X7t2rf/5z3/2UCjk\nf//73yvq999/f2/fvr1v27at1j7Kysq8Xbt2NZKISBqSGF9zzTV19rtlyxb//vvv/bvvvqtI0jdt\n2uTuwWscCoX8xBNPrLOf8sSncuI4ZcoUD4VC/uKLL9a5fSRffvmlm5lfcMEFEeu7devmBxxwQJ39\njB8/3kOhkB9//PE+b948X758uT/88MPeunVrz8rK8qysrDr7KH/PDh06NGq7SZMmuZn5+PHja23z\nxRdf+Lx58yrO/926dav36tXLR48e7e7uc+fO9UMPPdQ7d+7sP/vZz+p1HnW0fb5w4UI3M//tb3/r\na9eurXL79ttvvVWrVn7uuee6e2z7vLJo7yN3zjEGAABIa5t3BFOBdxRtT0j/hZI8wq0wzu1rq2+o\nww47TNOmTdPmzZv105/+VN27d9fPf/5zHXXUUTrxxBMlSbvuumvUPsqnhG7YsEFvv/12lbqMjAz1\n69dP++23X8X00GHDhuncc8/VkUceqWnTpum6667TLbfcoqeeekozZ87UtddeW2fcpaWlGjdunAoK\nCpSTk6P27dtrt9120y9/+UtJwTmi5T777DPts88+NaYAV7Z27VqtX79eP/rRj+p87Ibo1atXxPLv\nvvtOw4cPV6dOnZSXl6f8/Hx16NChYnXmDRs2SArOSXX3esV38cUXKxQKafLkyRVlkydPVocOHXTS\nSSc1KP7yqc7l06Gr27p1a8Qp0tXdcMMNuvnmmzV37lwdfPDB6tmzp6677jrdfffdatu2bZ3vNSl4\nzx5++OF6/fXXa41HCp5zZmamLrzwwlrb9OjRQwMGDFBeXp6k4Bq/mZmZuvnmm7Vy5Uodd9xxGjJk\niP7+97+rrKys4jNRH5H2+eLFiyVJo0eP1m677Vbl1rFjR23ZskXffPONpNj2eX3fR4nGOcYAAAAp\nrsUum6Ui6bsvP092KCnntNNO06mnnqqFCxdq06ZNKigoUH5+vg4++GBlZmZq7733rrOPPffcU1KQ\nYEbz4IMP6vPPP9dLL70kSXr00Ud1+umnV5y/e+ONN2rkyJG6//77o/Zz9dVX6/7779fZZ5+tW265\nRR06dFBWVpbef/993XDDDREX4IqX2s4x3rFjR63b1JY0HnPMMVqyZIlGjRqlAw88UK1bt1ZGRoYe\nffRRTZ06tUHPY4899tBxxx2nJ598UnfeeaeWL1+uN998U9dff33FjxOxatu2rXJzc7V69eoaddu3\nb9fatWs1ePDgOvsxM40dO1Y33nijFi5cKHfX/vvvr9LSUg0fPlyHHHJIveLZc889NXfuXK1fv16d\nOnWqUf/xxx/rvffe00knnVTvxbMWLFige+65R7Nnz1ZWVpaeeuopdejQoWIxtYkTJ6pXr1569913\ndfDBB9fZX6R9Hgy2BucwVz8nv1zlBevqKxHvo4YgMQYAAEhx2xSMCGlr7asopzMzU//+/Sv+XrNm\njebPn68jjzwy4uJc1S1dulSS1LFjx1rbrF69WjfddJP++Mc/VqwQ/NVXX2nAgAEVbbp27aqtW7dq\n7dq1ys/Pr7WvJ598UkcccYSeeuqpiHFU1rt3b3366acqKSlRVlZWxP7y8/PVtm1bffTRR7U/ybB2\n7drVuN6tJH3+eWw/uixYsEALFixQYWFhjWsAP/zww1X+3nvvvWVm+vDDD+vV9/DhwzVz5kxNnz5d\n77//vsxMF110UUzxVWZm+vGPf6z58+fXeB3fffddubsOOuigeveXm5urgQMHVvz9/PPPy93rPSK7\ndOlSZWZmVlkUrrJHHnkk4qJbtXF3DRs2TBdddFHFAnGrV69Wly5dKtqUL5715Zdf1isxjqR8FDkj\nI0NHHXVU1Lb13eexvI8SjanUAAAAMRodrS5aZQPtyAqm0bbwuqdqpjt318iRI1VWVqabb765ory0\ntFQbN26s0f7LL7/UQw89pPz8fB166KG19nv55ZfrsMMOqxgdlqTOnTtr4cKFFX8vWLBALVq0iJoU\nS0FiUT76Vq6oqEgTJ06s0fbcc8/VunXrol6myMx09tlna9GiRXr00UejPnbv3r21adMmvffeexVl\n7q4JEybEtGJ1+eht9dG8jz/+uMZldtq2bavjjz9es2bNqrg0UzQnnniidt99d02aNElPPPGEBg0a\npN69e9c7tkjOPvtsFRUV1Ui2Jk6cqKysrCqX/NqxY4eWLFmiL7/8ss5+v//+e910003q0KGDLrnk\nkoryjRs3RhzpfOmll/T2229r6NChEafHb9++XU899ZQ6duxY70T73nvv1VdffaXf//73FWWdO3fW\nZ599ppKSEknBe7O8vKEOOOAA9evXT5MmTdLy5ctr1JeWllacBlDffR7L+yjRGDEGAACIUWG0umiV\nDbSlrJOkj7SjqPZzEtNRUVGRBg4cqFNOOUU9evTQDz/8oKlTp+qDDz7Q+PHjdfjhh1e03bx5s3r0\n6KGTTz5Z++67r9q2batPP/1UkydPVlFRkZ555hllZ2dHfJwXXnhBs2fP1ieffFKl/LzzztPFF1+s\nq6++Wl26dNG4ceN07rnn1hn36aefrocfflhnnXWWjj76aK1Zs0aPPfZYxIT6qquu0owZMzRu3Dj9\n+9//1tChQ5WTk6NPPvlES5cu1T//+U9J0rhx4zR79mwNGzZM//znP3XYYYfJPbhMUWlpqaZMmSIp\nGI29++67dfLJJ+uqq65SixYt9Pzzz6u0tLRGsh7NvvvuW3Ht5aKiIhUUFGjJkiV6+OGH1b9/f73/\n/vtV2t9///0aNGiQjj/++IpL9xQXF+vdd99Vjx499Lvf/a6ibSgU0kUXXaRx48bJzHTHHXdEjOHC\nCy/UE088oTlz5lTZ15EMGzZMjz32mK655hotX75c++67r1566SX99a9/1a233qpu3bpVtF29erX2\n3XdfDR48WLNnz64onzVrlv7whz/omGOOUadOnbRixQpNnjxZGzZs0IwZM6qMAL/++uu65pprdNJJ\nJ6lnz57KzMzUu+++WzHFecKECRHjfPHFF7Vu3TrdcMMNCoXqHsNcuXKlbr31Vj3xxBNVznE+88wz\nNXbsWJ166qk64YQTdP/996ugoKDBo8Xl/vznP2vIkCHq37+/LrroIvXt21dbtmzRsmXL9Je//EV3\n3HFHxbny9dnnsb6PahPLezdqJ+l+E6tSAwCAFPZRwQB3yWf96ndcrqmS7du3+znnnOM9e/b03Nxc\nb9++vR933HERLw+zbds2HzZsmPfv39/btWvnLVq08M6dO/sZZ5zh7733Xq2P8cMPP3jnzp393nvv\njVh/xx13ePfu3T0/P99/9atfVVlBtzbFxcV+/fXX+5577um5ubneu3dvv/POO/21116LeOmobdu2\n+fjx471fv36em5vrbdu29YEDB/qkSZNqxPqb3/zGe/Xq5dnZ2Z6fn++HH364P/fcc1XazZo1yw84\n4ADPycnxLl26+I033uhLly71UCjkY8eOrWg3Z86ciPGUW7VqlZ9xxhneoUMHz8vL84MPPtinT58e\nceVgd/evv/7aR4wY4d27d/fs7Gzv1KmTH3vssRUrfFe2cuVKz8jI8NatW0e8nI97cLmujIwMX7p0\nae0vdrXX58orr/QuXbp4Tk6O9+3b1x988MEa7VasWOGhUMiPOuqoKuWLFi3yE044wTt37uzZ2dne\npUsXv+CCCyI+/uLFi/3MM8/0vffe23fZZRfPycnxvffe26+88kr/+uuva41x6NChnpGRUXFpsLqc\ncMIJfsopp0Ssmzlzpu+3336+6667+pAhQ+r1OtW1z92D/T5ixAjv0aNHxftswIABfvPNN/tXX31V\npW199nks76NErkptHo/suokzM+d1AAAAqWrePsfpoCX/0Kyzx+iEqaPl7lHnvHJsg6ZuzZo16tq1\nq4YNG6YHH3ywRr27V6xUXdf0caAyM4v4Hco5xgAAACnOs4Pz8LJKa65gCzRHDz74oMrKyjR8+PCI\n9e+//76Ki4t1++237+TI0FxxjjEAAECK25EZrC7rxZxjjOZt2rRpWrlype666y4dd9xxtV4Hd8CA\nAdq8efNOjg7NGSPGAAAAMSqMVhetsoG8RXB9XS9eWEdLoGk7++yzNWbMGB1xxBH605/+lOxwkEZI\njAEAAGI0JlpdtMoGKmsRXF83VLIp/p0DKaSsrExFRUWaNWuWdt9992SHgzRCYgwAAJDqsoOFtEq3\ndKujIQCgIUiMAQAAUlxmqyAhtq2lSY4EAJonEmMAAIAUV5KxUZLUMmNRkiMBgOaJxBgAACDFZbZq\nJ0lqUbYxyZEAQPNEYgwAABCj0dHqolU2kLcIzjHesr5z/DsHAHAdYwAAgFgVRquLVtlArfK7B/+o\n52WMc3JyvjGzjvGPBACatpycnG8ilZMYAwAApLockyS1b7msXs2Li4s7JTIcAGhumEoNAACQ4rxF\nMJaRo81JjgQAmicSYwAAgFSXEyTGW9a3S3IgANA8kRgDAACkuFb5wcxo216W5EgAoHkiMQYAAIhR\nYbS6aJUNlNkqR5LUvuXX8e8cACBz92THkHRm5rwOAACgvkxSbUcOZlK8Dyu+evVT7XHMvvquXXt1\nWPe93N3i+wgAkN4YMQYAAEhxGXnBiHHJ5uwkRwIAzROJMQAAQIrL2iVIjDN37EhyJADQPJEYAwAA\npLisVsFI8S5ZG5McCQA0TyTGAAAAKa588a3MUkaMASARSIwBAABiNDpaXbTKBmqxSzBinMVUagBI\nCFalFqtSAwCA1OYu7QhlKUs7ghWxWZUaAOKKEWMAAIAUZyYpI9lRAEDzRWIMAADQBJSFOGwDgETh\nGxYAAKAJKGXIGAAShsQYAACgCdhempPsEACg2SIxBgAAiFFhtLpolY2QkVOSmI4BAKxKLbEqNQAA\niI1Jqu3IwSxYRTrefshro9ZbfmBVagBIAEaMAQAAmoAStUh2CADQbJEYAwAANAHbS3OTHQIANFsk\nxgAAAE1AVu6WZIcAAM0WiTEAAEATsMOZSg0AiUJiDAAAEKPR0eqiVTbCDudyTQCQKKxKLValBgAA\nqW/ZLv219+aFrEoNAAnAiDEAAEATkJO3IdkhAECzRWIMAADQBGwvY1VqAEgUEmMAAIAmYIeTGANA\nopAYAwAANAElZa2SHQIANFskxgAAADEqjFYXrbIRsvO+S0zHAABWpZZYlRoAAMTGJNV25GAmJeKw\nYlGXA9Xn6w9YlRoAEoARYwAAgCagRLskOwQAaLZIjAEAAJqAUm+Z7BAAoNkiMQYAAGgCWuStSXYI\nANBskRgDAAA0Adt3tE52CADQbJEYAwAAxGh0tLpolY1QojaJ6RgAwKrUEqtSAwCA1PfOXmfrf754\nhlWpASABGDEGAABoArJyVic7BABotkiMAQAAmoCSst2SHQIANFskxgAAAE3ADuMcYwBIFBJjAACA\nJsAzspIdAgA0WyTGAAAAMSqMVhetshEyW6xKTMcAABJjAACAWI2JVhetshFKvHNiOgYAkBgDAAA0\nBR5qnewQAKDZIjEGAABoAjwzM9khAECzRWIMAADQBFjGymSHAADNFokxAABAE1BqXZMdAgA0WymT\nGFvgajNbbGbFZrbKzO4ys5b12LaXmY01s/8zs2/NbKOZzTezm+qzPQAAQCxGR6uLVtkInrFLYjoG\nAKROYixpoqS7JX0s6QpJz0oaKelv9dj2IklXSVqmYKHI6yR9KmmcpLfMLDsRAQMAgPRUGK0uWmVj\nZGQkqGMAQEqs4mBmfRQkw8+7+xmVyldIus/MznL3Z6J08Zyk8e6+qVLZw2a2TNJNki6W9GD8IwcA\nANg5PGN1skMAgGYrVUaMzwnfT6xW/oikLZLOi7axu39QLSkuN02SSerX6AgBAACSqMz2SHYIANBs\npUpiPEBSmaR5lQvdfZukDyUd1MB+y1ep+KbhoQEAAKSATJZNAYBESZXEuLOkte5eEqFutaR8M4tp\n2reZhSTdKqlE0tONDxEAACCJQqly2AYAzU+qfMO2lLStlrqtldrE4l5JB0u61d0/a2hgAAAA1RVG\nq4tW2QilWpOYjgEAMndPdgwyswWSdnP33SPUTZN0uqRsd99Rz/5+K+lmSZPc/bJ6tPfRla6tMHjw\nYA0ePLie0QMAgHRjkmo7gjKT4nl4NWfOHM2ZM0fLpr6rvZe+rDGS3N3i9wgAgFRJjF+WNERSy+rT\nqc3sX5J6uXvHevZVKOk2SZPdfVg9t/FUeB0AAEDTsDMT43KvnPKgjpl+efDYJMYAEFepMpV6noJY\nBlYuDF9/+EeqtihXbSolxY/VNykGAABoCixELgwAiZIqifG08P2oauXDJeVKeqq8wMx6mllB9Q7M\n7DYFSfEUd784UYECAAAkQ5mvT3YIANBsxbTSc6K4+8dm9oCky83sBUkzJfWRdKWkOe4+tVLz2Qou\nw5RRXmBmlytYB2OlpNlmdm61h/jG3V9N4FMAAABIqFBm+2SHAADNVkokxmFXSX81VeAAABHBSURB\nVFquYJT4BElrFawsPbpaO1fN03oGhMu6SXo8Qt9zJZEYAwCAuKh+cFKlLlplIxiXawKAhEmJxbeS\njcW3AABAqpt9zp901NRhLL4FAAnAT48AAABNQJlvTnYIANBskRgDAAA0AaHMXZMdAgA0WyTGAAAA\nTQDnGANA4vANCwAA0BRwHWMASBgSYwAAgBgVRquLVtkoJYnqGADSHokxAABAjMZEq4tW2QihjBaJ\n6RgAQGIMAADQFFhGRrJDAIBmi8QYAACgKTDOMQaARCExBgAAaAKMxbcAIGFIjAEAAJoAEmMASBwS\nYwAAgBiNjlYXrbIxmEoNAAlDYgwAABCjwmh10SobwTI4bAOAROEbFgAAoAlgKjUAJA6JMQAAQBNA\nYgwAiUNiDAAA0BRwjjEAJAyJMQAAQBPAiDEAJA6JMQAAQIwKo9VFq2wEFt8CgMThGxYAACBGY6LV\nRatsDKZSA0DCkBgDAAA0AW3266oP9huY7DAAoFkyd092DElnZs7rAAAA6ssk1XbkYCYl8rDCzOTu\nDB8DQBwxYgwAAAAASGskxgAAAACAtEZiDAAAEKPR0eqiVQIAUhLnGItzjAEAQNPBOcYAEH+MGAMA\nAAAA0hqJMQAAAAAgrZEYAwAAAADSGokxAAAAACCtkRgDAADEqDBaXbRKAEBKYlVqsSo1AACIjUmq\n7cjBTErkYQWrUgNA/DFiDAAAAABIayTGAAAAAIC0RmIMAAAAAEhrJMYAAAAAgLRGYgwAABCj0dHq\nolUCAFISq1KLVakBAEDTwarUABB/jBgDAAAAANIaiTEAAAAAIK2RGAMAAAAA0hqJMQAAAAAgrZEY\nAwAAxKgwWl20SgBASmJVarEqNQAAiI1Jqu3IwUxK5GEFq1IDQPwxYgwAAAAASGskxgAAAACAtEZi\nDAAAAABIayTGAAAAAIC0RmIMAAAQo9HR6qJVAgBSEqtSi1WpAQBA08Gq1AAQf4wYAwAAAADSGokx\nAAAAACCtkRgDAAAAANIaiTEAAAAAIK2RGAMAAMSoMFpdtEoAQEpiVWqxKjUAAIiNSartyMFMSuRh\nBatSA0D8MWIMAAAAAEhrJMYAAAAAgLRGYgwAAAAASGskxgAAAACAtEZiDAAAEKPR0eqiVQIAUhKr\nUotVqQEAQNPBqtQAEH+MGAMAAAAA0hqJMQAAAAAgrZEYAwAAAADSGokxAAAAACCtkRgDAADEqDBa\nXbRKAEBKYlVqsSo1AACIjUmq7cjBTErkYQWrUgNA/DFiDAAAAABIayTGAAAAAIC0RmIMAAAAAEhr\nJMYAAAAAgLSWUomxBa42s8VmVmxmq8zsLjNruTO2BwAAqI/R0eqiVQIAUlJKrUptZvdKulLSC5Je\nlrSvpJGS3nD3oxO1PatSAwCApoJVqQEg/jKTHUA5M+sj6QpJz7v7GZXKV0i6z8zOcvdnErU9AAAA\nACA9pdJU6nPC9xOrlT8iaYuk8xK8PQAAAAAgDaVSYjxAUpmkeZUL3X2bpA8lHZTg7QEAAAAAaSiV\nEuPOkta6e0mEutWS8s0s2tTvxm4PAAAAAEhDqZQYt5S0rZa6rZXaJGp7AACAeimMVhetEgCQklIp\nMd4iKbuWupxKbRK1PQAAQL2MiVYXrRIAkJJSaWrx15L2NbOsCNOhuyiYJr0jUdsXVvp5d/DgwRo8\neHAssSNJ5syZw75qwth/TRv7r+li3zUtc+bM0Zw5c5IdBgA0a6mUGM+TdIykgZLeKi80s2xJP5I0\nJ5HbFzLvqUni4K5pY/81bey/pot917RU/8F+DEPSABB3qTSVelr4flS18uGSciU9VV5gZj3NrKCh\n2wMAAAAAUC5lRozd/WMze0DS5Wb2gqSZkvpIulLSHHefWqn5bEldJWU0cHsAAAAAACRJ5u7JjqGC\nmZmCEd/hkvaUtFbSM5JGu/uWSu2WS+rq7pkN2T7C46bOiwAAAFAHd7dkxwAAzUlKJcYAAAAAAOxs\nqXSOMQAAAAAAOx2JMQAAAAAgrZEYAwAAAADSGokxAAAAACCtNcvE2AJXm9liMys2s1VmdpeZtdwZ\n26PhGvPam1kvMxtrZv9nZt+a2UYzm29mN7Hvdo54fnbMLNfMvjCzMjO7LxHx4r/ise/MrG14m8/C\nfXxrZrPNbFAiY0dc/t/LC39XLgh/d35nZm+Z2QWJjh2Smd1oZs+a2efh77wvGtjPL83sAzPbYmZr\nzOwRM8uPd7wA0Bw1y8RY0kRJd0v6WNIVkp6VNFLS33bS9mi4xrz2F0m6StIySWMkXSfpU0njJL1l\nZtmJCBhVxPOz81tJ7SWxdP7O0ah9Z2bdJH0g6XxJz0kaIel2ScsldUlAvKiqwfsvfKnDlxV8b74r\n6RoFn7+QpMfM7HcJihn/dbukIxX8/7W+IR2Y2dWSHg9vP1LSJElnSXrdzHLjEyYANGPu3qxukvpI\nKpX0bLXyKySVSTorkdtzS+q++7GkXSKU/zbc72XJfo7N+RbPz054X5YouC55maT7kv38mvMtHvtO\n0puSVkrqkOznk263OHx3/k+43V3VyjMlfS5pXbKfY3O/Sdqz0r8XSvoixu3bS9os6f8UvhRnuPyn\n4X17Q7KfIzdu3Lil+q05jhifE76fWK38EUlbJJ2X4O3RcI167d39A3ffFKFqmiST1K/RESKauHx2\nzCwU3mampBfjFh2iadS+M7PDJQ2S9Ht3/9bMMhmh2qka+9nbNXz/n8qF7r5D0lpJRY0NENG5+4pG\ndnGKpFxJ/+vuFbNs3P3vkr4Qxy4AUKfmmBgPUPDr6LzKhe6+TdKHkg5K8PZouES99l3D9980PDTU\nQ7z23zWSeisY7cLO0dh9d7yCKe9fmdkMScWSisxsiZmdm4B4UVVj99+/JW2QdL2ZnW5mXc2sIDyF\n+seSRicgZsTXgPD9OxHq3pG0D2ttAEB0zTEx7ixprbuXRKhbLSnfzDITuD0aLu6vfXj08VYF03Kf\nbnyIiKLR+8/MekgqlDTG3b+Mf4ioRWP3XYGCWRmPSGqj4DzjX0naJunPLOCUcI3af+6+QdJJCs5N\nfVbBlPjFCs4TP83dH41/yIizzuH71RHqViv4fHaOUAcACGuOiXFLBQdjkWyt1CZR26PhEvHa3yvp\nYEm3uvtnDQ0M9RKP/TdJweIzE+IVFOqlsftul/D9RklHuvsz7j5F0uEKRiLHxyVK1CYen70iBQt3\n/UHBtNyLFXwWp5rZkHgEiYRqKVXMEqiOYxcAqIfmmBhvkVTb6sM5ldokans0XFxfezP7raTLJf3R\n3e9sZGyoW6P2n5mdJ2mIpBHuXhrn2BBdYz97xQqmUk8Nn5cqqWIk8m+SOplZQTwCRUSN/eztJ+lt\nSf9w99+4+1/d/TFJP5G0RtIj4ZWrkbq2SFItV1/g2AUA6qE5JsZfK5g2lhWhrouC6WY7ItTFa3s0\nXNxeezMrlHSzpMnufln8QkQUDd5/ZtZCwaVmZkr61sz2MrO9JO0ZbtI6XNY6AXGj8Z+9r8L3ayLU\nlS/o1LYR8SG6xu6/qxUk1s9XLnT3YkkvSequ/34WkZq+Dt9HujRaFwU/XH0doQ4AENYcE+N5Cp7X\nwMqF4V9Rf6Rqi5MkYHs0XFxe+3BSfJukx9x9WJxjRO0as/9yJe0m6URJn1W6va7ggO58SUsVTO9E\n/DX2s/dvBecw7hGhrnzxu28bGSNq19j9V37uaUaEusxq90hN8xR8Bg+JUPc/kpa4OyPGABBFc0yM\np4XvR1UrH67g4Pup8gIz6xlhel+9t0fcNXbfycxuU5AUT3F3kqidqzH7r0jS6ZJ+Eb4vv41QcLA3\nK/z33xISORr72ZsuaZOk8yqvfGtmu0v6uYKD8i/iHjXKNXb/LVLwObuwcqGZtZF0soJFuZbFMV40\nQqVVwyv/kPFXBac0XFF52ruZnSSpp6Qnd3KYANDkWKXL3TUbZnafgnNLpyuYmtlH0pWS3nT3IZXa\nrZDU1d0zGrI94q8x+87MLpf0vwpWVL1NweVLKvvG3V9N6BNIc4397EXor7uk5ZLud/eRiYobcfne\nHKZg8bRFkh5VMDX3UkmdJJ3o7q/thKeRthr53dlN0gcKVhR/WtJbktpL+rWCadSXufsfd84zSU/h\nNRa6K/iB4gpJWZLuCVevdPcnK7Wdo2Bhuz3dfVWl8msULJ42V9JUBTM4rlHwf+JARowBILrmmhib\ngl/Ohys4L2qtpGckja78H4OZLVdwgJDZkO0Rf43Zd2b2mKRfRul+rrsflYi4EWjsZy9Cf90lfaEg\nMb4qUXEjPvvOzE6WdL2k/RT8MPW2gktvRbq2KuIoDv/v9VDwg+IQSR0VjD5+KGmCu/91ZzyHdGZm\nrytIdiOp8n9XuO1PJPWsnBiH636p4JzxAgWrxM+QdKO7r01I4ADQjDTLxBgAAAAAgPpqjucYAwAA\nAABQbyTGAAAAAIC0RmIMAAAAAEhrJMYAAAAAgLRGYgwAAAAASGskxgAAAACAtEZiDAAAAABIayTG\nAAAAAIC0RmIMoNkysxVmVlbtVmxmK81smpkdnuwYY2Vm48LP46Zq5UPC5f9MVmwAAABNVWayAwCA\nBPLw7R+S1oTL2ko6QNIvJP3CzK5293uTFF9DlD8nAAAAxAmJMYB0cIe7v1H+h5llSLpX0mWS7jCz\n59z966RFBwAAgKRiKjWAtOPupZKuk7RJUgtJQ5MbEQAAAJKJxBhAWnL3rZKWhv/sGKmNmZ1jZq+Y\n2Voz2xY+Z3mSmXWtrV8z62ZmE81skZltNrMfzOwTM7vfzPat1vYYM3vQzD4MP8bW8GM8ZmYF8Xu2\nAAAAiIap1ADSWevw/TeVC80sU9Lzkn4mqUjS++E2+0kaLul0Mxvi7h9V2+54SdMk5UlaLWlWuKqn\npEskfS1pcaVNHlaQlH8iaY6CHyv7Sbog/BhHu/u78XiiAAAAqB2JMYC0ZGZ9JfWQtF1S9ZWc71CQ\nFL8m6Tx3/6bSdiMlTZT0jJn1cXcPl++p/ybFN0r6Q3lduL6rpPbVHucqSa+7+6ZqsY2Q9ICkP0r6\nUaOeKAAAAOrEVGoAacXM2oRHdl+QZJKuqrzwlpnlS7pc0gZJZ1ROiiXJ3e9TsMp1b1U9N/k6Sa0k\nPenud1ZOisPbfenuH1Yr+1v1pDhc/pCkf0vaz8x6NfzZAgAAoD4YMQaQDuaYWfWyrZKOd/dXqpUf\nJSlb0ix3X19Lf3MlHSvpEAVJssJ/u6TJsQQWHkk+QVKBpF0lZYSrdgvf95b0WSx9AgAAIDYkxgDS\nwcsKrmNskjpJOlxSrqQnzGyQu39RqW3P8P3JZlYWpU/Xf5NXSeoWvv+0vkGZ2ThJv1H02Tu71rc/\nAAAANAyJMYB0UP06xh0VnFe8n6SnFIz8lisfsV0sqa6Fr/7d0IDM7ExJNymYsn21gsW3/uPu28P1\n0ySdriCZBwAAQAKRGANIO+7+jZmdIWmBpIFmdo67Px2u/jJ8/6G7XxRDt6sUjDYXSPq2Hu1PVzDq\n/Bt3nxKhfu8YHhsAAACNwOJbANKSuy+R9KCCEdlCMyv/PnxF0g5JQ82sVQxd/iPc16/r2b5d+P6r\n6hVm1k9S/xgeGwAAAI1AYgwgnd0uaZOkvSSdL0nu/h9JkxRcWmlGpFWhzaylmZ1rZpUvv3S3gmse\nn2dm/69Sol2+TTczO6BS0acKEulh4esml7frKOlx8f0MAACw03DgBSBtuftaSXcpSFBvqZTMXivp\neQWLdH1iZv82s2nh27uS1kl6QlLrSn0tl3SWguT495JWmNlzZva8mb0v6QtJx1d6+AmSfpD0c0nL\nwn3PkLRMUpakGar9/GLOOwYAAIgjEmMAzZ3XUX+PpG8UnB98gSS5e4m7nyHpZEkvSeqiIIE9UlKO\npCfDf6+o8kDuLymYAv2AgstBnShpiKQWku5XkGyXt10m6ceSnlWQ6J6k4PzkByQNUjCSXVvs0crr\ner4AAACoxtw5hgIAAAAApC9GjAEAAAAAaY3EGAAAAACQ1kiMAQAAAABpjcQYAAAAAJDWSIwBAAAA\nAGmNxBgAAAAAkNZIjAEAAAAAaY3EGAAAAACQ1kiMAQAAAABp7f8DtaAsP2wdl64AAAAASUVORK5C\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12e3143a10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from scipy import interp\n",
    "\n",
    "n_experiments=10\n",
    "\n",
    "fig = plt.figure(figsize=[10,8])\n",
    "\n",
    "\n",
    "## ROC CURVE\n",
    "\n",
    "# 1) micro averaged ROC curve is straightfowrward -concatenate true labels -> concatenate probabilities -> draw.\n",
    "\n",
    "all_labels = np.concatenate(roc_labels)\n",
    "all_probas = np.concatenate(roc_probas)\n",
    "fpr,tpr,thresh = roc_curve(all_labels,all_probas[:,1])\n",
    "plt.plot(fpr,tpr,linewidth=2,label=\"micro average over 10 folds\",color=\"blue\")\n",
    "\n",
    "print \"Area under the microaveraged ROC curve is: \", roc_auc_score(all_labels,all_probas[:,1])\n",
    "\n",
    "\n",
    "# for macro average, we need to gather the TP and FP rates for each experiment individually\n",
    "\n",
    "FP_rates=[]\n",
    "TP_rates=[]\n",
    "\n",
    "areas=[] #we also keep track of the AUC to later validate\n",
    "counts=[] #also just for validation purposes\n",
    "\n",
    "# 2) draw dotted lines for each experiment, collect the values\n",
    "for i in range(n_experiments):\n",
    "    #print np.shape(roc_labels[0]),np.shape(roc_probas[0][:,1])\n",
    "    fpr,tpr,thresh = roc_curve(roc_labels[i],roc_probas[i][:,1])\n",
    "    FP_rates.append(fpr)\n",
    "    TP_rates.append(tpr)\n",
    "    \n",
    "    #print roc_auc_score(roc_labels[i],roc_probas[i][:,1])\n",
    "    plt.plot(fpr,tpr,color=\"gray\",linestyle=\"dotted\")\n",
    "    areas.append(roc_auc_score(roc_labels[i],roc_probas[i][:,1]))\n",
    "    counts.append(len(roc_labels[i]))\n",
    "\n",
    "\n",
    "\n",
    "# 3) Compute macro-average ROC curve and ROC area\n",
    "\n",
    "# All experiments have values in different X-coordinates, to average we need to\n",
    "# interpolate between known values.\n",
    "# To do this, we first aggregate all used false positive rates (X-coordinates)\n",
    "all_fpr = np.unique(np.concatenate([FP_rates[i] for i in range(n_experiments)]))\n",
    "\n",
    "# Then interpolate all ROC curves at these points\n",
    "mean_tpr = np.zeros_like(all_fpr)\n",
    "for i in range(n_experiments):\n",
    "    mean_tpr += interp(all_fpr, FP_rates[i], TP_rates[i])\n",
    "\n",
    "# Finally average it and compute AUC\n",
    "mean_tpr /= n_experiments\n",
    "plt.plot(all_fpr,mean_tpr,linewidth=2,label=\"macro average over 10 folds\",color=\"red\")\n",
    "\n",
    "plt.ylabel(\"True positive rate (recall)\",fontsize=22)\n",
    "plt.xlabel(\"False positive rate\",fontsize=22)\n",
    "plt.plot([0,1],[0,1], label=\"random model\",color=\"green\")\n",
    "plt.xticks(fontsize=18)\n",
    "plt.yticks(fontsize=18)\n",
    "plt.legend(loc=4,fontsize=18)\n",
    "\n",
    "print \"Area under the macroaveraged ROC curve is: \", auc(all_fpr, mean_tpr)\n",
    "\n",
    "\n",
    "#\n",
    "#-----------------------------------------------------------------------------\n",
    "# PRECISION - RECALL curve\n",
    "#-----------------------------------------------------------------------------\n",
    "#\n",
    "\n",
    "plt.figure(figsize=[10,8])\n",
    "ax = plt.subplot(111)\n",
    "\n",
    "# 1) again micro-averaging is easy:\n",
    "precision_vir,recall_vir,thresh = precision_recall_curve(all_labels,all_probas[:,1])\n",
    "ax.plot(recall_vir,precision_vir,color=\"blue\",label=\"Micro average over 10 folds \\n for virus class\",linewidth=2)\n",
    "\n",
    "\n",
    "# save virus and non-virus precision and recalls experiment per experiment in here\n",
    "experiment_v_precs=[]\n",
    "experiment_v_recs=[]\n",
    "experiment_nv_precs=[]\n",
    "experiment_nv_recs=[]\n",
    "\n",
    "# 2) draw dotted lines for each experiment, collect the values\n",
    "for i in range(n_experiments):\n",
    "    prec_v,rec_v,thresh = precision_recall_curve(roc_labels[i],roc_probas[i][:,1])\n",
    "    prec_nv,rec_nv,thresh = precision_recall_curve(np.abs(roc_labels[i]-1),roc_probas[i][:,0])\n",
    "    experiment_v_precs.append(prec_v)\n",
    "    experiment_v_recs.append(rec_v)\n",
    "    experiment_nv_precs.append(prec_nv)\n",
    "    experiment_nv_recs.append(rec_nv)\n",
    "\n",
    "    ax.plot(rec_v,prec_v,color=\"gray\",linestyle=\"dotted\")\n",
    "\n",
    "#draw the last line again, simply to make an entry to the legend\n",
    "ax.plot(rec_v,prec_v,color=\"gray\",linestyle=\"dotted\",label=\"Results on each data set\")\n",
    "\n",
    "\n",
    "\n",
    "# 3) Compute macro-average precision-recall curve and AUC\n",
    "\n",
    "# All experiments have values in different X-coordinates, to get an average we need to\n",
    "# interpolate between known values.\n",
    "\n",
    "# To do this, we first aggregate all used recall rates (X-coordinates)\n",
    "all_v_rec = np.unique(np.concatenate([experiment_v_recs[i] for i in range(n_experiments)]))\n",
    "\n",
    "# Then interpolate all ROC curves at these points\n",
    "mean_precisions = np.zeros_like(all_v_rec)\n",
    "for i in range(n_experiments):\n",
    "    # NOTICE: to use interpolate, the x-axis values need to be sorted increasingly,\n",
    "    # so we need to reverse the arrays\n",
    "    mean_precisions += interp(all_v_rec, experiment_v_recs[i][::-1], experiment_v_precs[i][::-1])\n",
    "\n",
    "\n",
    "# Finally average it and compute AUC\n",
    "mean_precisions /= n_experiments\n",
    "plt.plot(all_v_rec,mean_precisions,linewidth=2,label=\"Macro average over 10 folds\",color=\"red\")\n",
    "\n",
    "\n",
    "print \"\\n\\nArea under the micro-averaged PRECISION-REC curve is:\", auc(recall_vir,precision_vir)\n",
    "print \"Area under the macro-averaged PRECISION-REC curve is: \", auc(all_v_rec, mean_precisions)\n",
    "\n",
    "\n",
    "ax.set_xlabel('Recall',fontsize=22)\n",
    "ax.set_ylabel('Precision',fontsize=22)\n",
    "ax.set_ylim([-0.01, 1.05])\n",
    "ax.set_yticks([0.0,0.2,0.4,0.6,0.8,1.0])\n",
    "ax.tick_params(axis='both', labelsize=18)\n",
    "ax.set_xlim([0.0, 1.0])\n",
    "#plt.title('Precision-Recall example: AUC={0:0.2f}'.format(average_precision[0]))\n",
    "\n",
    "# 4) we also add lines to the plot do demonstate the trade-off at certain values \n",
    "id74 = np.min(np.where(precision_vir>0.74)[0])\n",
    "print \"at precision 0.74  \", precision_vir[id74],\" we have recall of: \",recall_vir[id74]\n",
    "id90 = np.min(np.where(precision_vir>0.9)[0])\n",
    "print \"at precision 0.9  \", precision_vir[id90],\" we have recall of: \",recall_vir[id90]\n",
    "id95 = np.min(np.where(precision_vir>0.95)[0])\n",
    "print \"at precision 0.95  \",precision_vir[id95],\" we have recall of: \",recall_vir[id95]\n",
    "ax.axhline(precision_vir[id90],linestyle=\"--\",label=\"90% accuracy, \"+str(np.round(recall_vir[id90],3))+\"% recall\")\n",
    "ax.axvline(recall_vir[id90],linestyle=\"--\")\n",
    "ax.axhline(precision_vir[id95],color=\"cyan\",linestyle=\"--\",label=\"95% accuracy, \"+str(np.round(recall_vir[id90],3))+\"% recall\")\n",
    "ax.axvline(recall_vir[id95],color=\"cyan\",linestyle=\"--\")\n",
    "ax.legend(loc=\"center left\",bbox_to_anchor=(1,0.5),fontsize=18)\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## TRAIN ON GenBank DATA, TEST ON metagenomics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "genbank data shape:  (631982, 59) (631982,) \n",
      " genbank counts of classes  [541999  89983]\n",
      "metagenomics data shape:  (188601, 59) (188601,) \n",
      " metagenomics counts of classes  [182936   5665]\n",
      "#################################################\n",
      "report at 0.5: \n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "  not virus       0.97      1.00      0.98    182936\n",
      "      virus       0.39      0.03      0.05      5665\n",
      "\n",
      "avg / total       0.95      0.97      0.96    188601\n",
      "\n"
     ]
    }
   ],
   "source": [
    "GB_features = loadtxt(\"data/new_data/genbank/matrix.txt\",delimiter=\",\")\n",
    "GB_labels = loadtxt(\"data/new_data/genbank/values.txt\",dtype=int)\n",
    "\n",
    "print \"genbank data shape: \",GB_features.shape, GB_labels.shape, \"\\n genbank counts of classes \", np.bincount(GB_labels)\n",
    "\n",
    "\n",
    "#TEST ON MetaGenomics DATA\n",
    "folder=\"data/new_data/concat_al2\"\n",
    "if \"concat_al2\" in folder:\n",
    "    features1 = loadtxt(folder[:-10]+\"forward_al2/matrix.txt\")\n",
    "    labels1 = loadtxt(folder[:-10]+\"forward_al2/values.txt\",dtype=int)\n",
    "        \n",
    "    features2 = loadtxt(folder[:-10]+\"reverse_al2/matrix.txt\")\n",
    "    labels2 = loadtxt(folder[:-10]+\"reverse_al2/values.txt\",dtype=int)\n",
    "\n",
    "    features = np.concatenate([features1,features2])\n",
    "    labels = np.concatenate([labels1,labels2])\n",
    "         \n",
    "else:\n",
    "    print \" NOT SUPPORTED!! We do LOO only on concat_al2 right now\"\n",
    "    assert(False)\n",
    "        \n",
    "print \"metagenomics data shape: \",features.shape, labels.shape, \"\\n metagenomics counts of classes \", np.bincount(labels)\n",
    "    \n",
    "\n",
    "\n",
    "#Train model on whole DNA dataset, then test on new\n",
    "rf = RandomForestClassifier(n_estimators=100, n_jobs=3)\n",
    "rf.fit(GB_features, GB_labels)\n",
    "\n",
    "\n",
    "\n",
    "print \"#################################################\"\n",
    "\n",
    "pred_new = rf.predict(features)\n",
    "prob_new = rf.predict_proba(features)\n",
    "report = classification_report(labels, pred_new, target_names=[\"not virus\", \"virus\"])\n",
    "\n",
    "print \"report at 0.5: \\n\", report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under the ROC curve is:  0.50964108438\n",
      "\n",
      "\n",
      "Area under the micro-averaged PRECISION-REC curve is: 0.0580687879753\n",
      "at precision 0.74   1.0  we have recall of:  0.0\n",
      "at precision 0.9   1.0  we have recall of:  0.0\n",
      "at precision 0.95   1.0  we have recall of:  0.0\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoQAAAIECAYAAACABz+XAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ndDihEAZBEARBEHQ4fwM2JrDrzzZapwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12e17e5dd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoQAAAH7CAYAAABVK47GAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f12e2d51150>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from scipy import interp \n",
    "fig = plt.figure(figsize=[10,8])\n",
    "\n",
    "fpr,tpr,thresh = roc_curve(labels,prob_new[:,1])\n",
    "plt.plot(fpr,tpr,linewidth=2,label=\"Microaveraged performance \\n of GenBank model on metagenomics\",color=\"blue\")\n",
    "\n",
    "print \"Area under the ROC curve is: \", roc_auc_score(labels,prob_new[:,1])\n",
    "\n",
    "plt.ylabel(\"True positive rate (recall)\",fontsize=22)\n",
    "plt.xlabel(\"False positive rate\",fontsize=22)\n",
    "plt.plot([0,1],[0,1], label=\"random model\",color=\"green\")\n",
    "plt.xticks(fontsize=18)\n",
    "plt.yticks(fontsize=18)\n",
    "plt.legend(loc=4,fontsize=18)\n",
    "\n",
    "\n",
    "\n",
    "plt.figure(figsize=[10,8])\n",
    "ax = plt.subplot(111)\n",
    "\n",
    "# 1) again micro-averaging is easy:\n",
    "precision_vir,recall_vir,thresh = precision_recall_curve(labels,prob_new[:,1])\n",
    "ax.plot(recall_vir,precision_vir,color=\"blue\",label=\"Micro average (over experiments)\\n for virus class\",linewidth=2)\n",
    "\n",
    "\n",
    "print \"\\n\\nArea under the micro-averaged PRECISION-REC curve is:\", auc(recall_vir,precision_vir)\n",
    "\n",
    "ax.set_xlabel('Recall',fontsize=22)\n",
    "ax.set_ylabel('Precision',fontsize=22)\n",
    "ax.set_ylim([-0.01, 1.05])\n",
    "ax.set_yticks([0.0,0.2,0.4,0.6,0.8,1.0])\n",
    "ax.tick_params(axis='both', labelsize=18)\n",
    "ax.set_xlim([0.0, 1.0])\n",
    "#plt.title('Precision-Recall example: AUC={0:0.2f}'.format(average_precision[0]))\n",
    "\n",
    "# 4) we also add lines to the plot do demonstate the trade-off at certain values \n",
    "id74 = np.min(np.where(precision_vir>0.74)[0])\n",
    "print \"at precision 0.74  \", precision_vir[id74],\" we have recall of: \",recall_vir[id74]\n",
    "id90 = np.min(np.where(precision_vir>0.9)[0])\n",
    "print \"at precision 0.9  \", precision_vir[id90],\" we have recall of: \",recall_vir[id90]\n",
    "id95 = np.min(np.where(precision_vir>0.95)[0])\n",
    "print \"at precision 0.95  \",precision_vir[id95],\" we have recall of: \",recall_vir[id95]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "report at 0.5: \n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "  not virus       0.97      1.00      0.98    182936\n",
      "      virus       0.39      0.03      0.05      5665\n",
      "\n",
      "avg / total       0.95      0.97      0.96    188601\n",
      "\n"
     ]
    }
   ],
   "source": [
    "report = classification_report(labels, pred_new, target_names=[\"not virus\", \"virus\"])\n",
    "print \"report at 0.5: \\n\", report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "## Train on Metagenomics and test on GenBank"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "genbank data shape:  (631982, 59) (631982,) \n",
      " genbank counts of classes  [541999  89983]\n",
      "metagenomics data shape:  (188601, 59) (188601,) \n",
      " metagenomics counts of classes  [182936   5665]\n",
      "#################################################\n",
      "report at 0.5: \n",
      "             precision    recall  f1-score   support\n",
      "\n",
      "  not virus       0.86      1.00      0.92    541999\n",
      "      virus       0.80      0.00      0.00     89983\n",
      "\n",
      "avg / total       0.85      0.86      0.79    631982\n",
      "\n"
     ]
    }
   ],
   "source": [
    "GB_features = loadtxt(\"data/new_data/genbank/matrix.txt\",delimiter=\",\")\n",
    "GB_labels = loadtxt(\"data/new_data/genbank/values.txt\",dtype=int)\n",
    "\n",
    "print \"genbank data shape: \",GB_features.shape, GB_labels.shape, \"\\n genbank counts of classes \", np.bincount(GB_labels)\n",
    "\n",
    "\n",
    "#TEST ON MetaGenomics DATA\n",
    "folder=\"data/new_data/concat_al2\"\n",
    "if \"concat_al2\" in folder:\n",
    "    features1 = loadtxt(folder[:-10]+\"forward_al2/matrix.txt\")\n",
    "    labels1 = loadtxt(folder[:-10]+\"forward_al2/values.txt\",dtype=int)\n",
    "        \n",
    "    features2 = loadtxt(folder[:-10]+\"reverse_al2/matrix.txt\")\n",
    "    labels2 = loadtxt(folder[:-10]+\"reverse_al2/values.txt\",dtype=int)\n",
    "\n",
    "    features = np.concatenate([features1,features2])\n",
    "    labels = np.concatenate([labels1,labels2])\n",
    "         \n",
    "else:\n",
    "    print \" NOT SUPPORTED!! We do LOO only on concat_al2 right now\"\n",
    "    assert(False)\n",
    "        \n",
    "print \"metagenomics data shape: \",features.shape, labels.shape, \"\\n metagenomics counts of classes \", np.bincount(labels)\n",
    "    \n",
    "\n",
    "\n",
    "#Train model on whole DNA dataset, then test on new\n",
    "rf = RandomForestClassifier(n_estimators=500, n_jobs=3)\n",
    "rf.fit(features, labels)\n",
    "\n",
    "\n",
    "\n",
    "print \"#################################################\"\n",
    "\n",
    "pred_new = rf.predict(GB_features)\n",
    "prob_new = rf.predict_proba(GB_features)\n",
    "report = classification_report(GB_labels, pred_new, target_names=[\"not virus\", \"virus\"])\n",
    "\n",
    "print \"report at 0.5: \\n\", report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Area under the ROC curve is:  0.554730378516\n",
      "\n",
      "\n",
      "Area under the micro-averaged PRECISION-REC curve is: 0.159295432153\n",
      "at precision 0.74   0.741610738255  we have recall of:  0.00245601947034\n",
      "at precision 0.9   0.907284768212  we have recall of:  0.00152250980741\n",
      "at precision 0.95   0.951048951049  we have recall of:  0.00151139659714\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoQAAAIECAYAAACABz+XAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff2fefe2090>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoQAAAH7CAYAAABVK47GAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7ff2fc9feb10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from scipy import interp \n",
    "fig = plt.figure(figsize=[10,8])\n",
    "\n",
    "fpr,tpr,thresh = roc_curve(GB_labels,prob_new[:,1])\n",
    "plt.plot(fpr,tpr,linewidth=2,label=\"Microaveraged performance \\n of MetaGenomics model on GenBank data\",color=\"blue\")\n",
    "\n",
    "print \"Area under the ROC curve is: \", roc_auc_score(GB_labels,prob_new[:,1])\n",
    "\n",
    "plt.ylabel(\"True positive rate (recall)\",fontsize=22)\n",
    "plt.xlabel(\"False positive rate\",fontsize=22)\n",
    "plt.plot([0,1],[0,1], label=\"random model\",color=\"green\")\n",
    "plt.xticks(fontsize=18)\n",
    "plt.yticks(fontsize=18)\n",
    "plt.legend(loc=4,fontsize=18)\n",
    "\n",
    "\n",
    "\n",
    "plt.figure(figsize=[10,8])\n",
    "ax = plt.subplot(111)\n",
    "\n",
    "# 1) again micro-averaging is easy:\n",
    "precision_vir,recall_vir,thresh = precision_recall_curve(GB_labels,prob_new[:,1])\n",
    "ax.plot(recall_vir,precision_vir,color=\"blue\",label=\"Micro average for virus class\",linewidth=2)\n",
    "\n",
    "\n",
    "print \"\\n\\nArea under the micro-averaged PRECISION-REC curve is:\", auc(recall_vir,precision_vir)\n",
    "\n",
    "ax.set_xlabel('Recall',fontsize=22)\n",
    "ax.set_ylabel('Precision',fontsize=22)\n",
    "ax.set_ylim([-0.01, 1.05])\n",
    "ax.set_yticks([0.0,0.2,0.4,0.6,0.8,1.0])\n",
    "ax.tick_params(axis='both', labelsize=18)\n",
    "ax.set_xlim([0.0, 1.0])\n",
    "#plt.title('Precision-Recall example: AUC={0:0.2f}'.format(average_precision[0]))\n",
    "\n",
    "# 4) we also add lines to the plot do demonstate the trade-off at certain values \n",
    "id74 = np.min(np.where(precision_vir>0.74)[0])\n",
    "print \"at precision 0.74  \", precision_vir[id74],\" we have recall of: \",recall_vir[id74]\n",
    "id90 = np.min(np.where(precision_vir>0.9)[0])\n",
    "print \"at precision 0.9  \", precision_vir[id90],\" we have recall of: \",recall_vir[id90]\n",
    "id95 = np.min(np.where(precision_vir>0.95)[0])\n",
    "print \"at precision 0.95  \",precision_vir[id95],\" we have recall of: \",recall_vir[id95]\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}