subgroup_classfier_b.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pysubgroup as ps\n",
    "import pandas as pd\n",
    "from graph_pipeline.main import *\n",
    "from tqdm.auto import tqdm\n",
    "import numpy as np\n",
    "import multiprocessing as mp\n",
    "from itertools import product\n",
    "import statsmodels.api as sm\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import confusion_matrix, plot_confusion_matrix, classification_report\n",
    "from sklearn.model_selection import GridSearchCV, StratifiedKFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from tqdm.auto import tqdm\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from collections import Counter\n",
    "import pickle\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "from pprint import pprint\n",
    "\n",
    "def filter_on_stats(stat_df, min_replies=10, max_replies=350, language='en', acyclic=True):\n",
    "    filter_array = np.logical_and(stat_df.replies >= min_replies ,stat_df.language == language)\n",
    "    filter_array = np.logical_and(filter_array, stat_df.is_acyclic == acyclic)\n",
    "    filter_array = np.logical_and(filter_array,stat_df.replies <= max_replies)\n",
    "    filter_array = filter_array.to_numpy()\n",
    "    return filter_array\n",
    "\n",
    "def drop_useless(graph_feature_df):\n",
    "    return graph_feature_df.drop(columns=[\"network_diameter\", \"network_radius\", \"closeness_centralization\", \"102\", \"111D\", \"111U\", \n",
    "                                          \"030C\", \"201\", \"120D\", \"120U\", \"120C\", \"210\", \"300\" ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "graph_features_all = pd.read_pickle(\"b_collection_extracted/graph_features_all\")\n",
    "stat_df = pd.read_pickle(\"b_collection_extracted/stat_df\")\n",
    "result_df = pd.read_pickle(\"b_collection_extracted/document_result_df_25tops\")\n",
    "graph_features_all['replies'] = stat_df.set_index('thread_id').replies\n",
    "clean_features = drop_useless(graph_features_all).drop(columns=[\"max_degree\", \"003\"])\n",
    "clean_features = clean_features[filter_on_stats(stat_df)]\n",
    "stat_df = stat_df[filter_on_stats(stat_df)]\n",
    "stat_df = stat_df.join(result_df.dominant_topic, on='thread_id')\n",
    "clean_features = clean_features.join(result_df.dominant_topic).drop(columns=\"board\")\n",
    "clean_features_label = clean_features.join(result_df.labels)\n",
    "clean_features_cluster = clean_features.join(pd.read_pickle(\"b_collection_extracted/result_df_dist_clust\").km.drop(columns=['dominant_topic', 'labels', 'board']))\n",
    "graph_features_clean = clean_features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dominant Topic als abhängige Variable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forests mit dominant topic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c393ff402d09469899fba475bdddfebb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/14 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10 50\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py:691: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': {20: 0.0002852253280091272, 7: 0.0005422993492407809, 11: 0.00030525030525030525, 5: 0.00029342723004694836, 3: 0.0005672149744753262, 10: 0.0005777007510109763, 4: 0.0003003003003003003, 12: 0.00024084778420038535, 9: 0.00039277297721916735, 1: 0.00044464206313917296, 13: 0.0003104625892579944, 17: 0.000625, 15: 0.0010362694300518134, 22: 0.0005636978579481398, 0: 0.0007272727272727272, 24: 0.0015923566878980893, 18: 0.0005959475566150178, 2: 0.0005807200929152149, 19: 0.000496031746031746, 21: 0.0006112469437652812, 16: 0.0004677268475210477, 23: 0.0004048582995951417, 6: 0.00044603033006244426, 8: 0.0005537098560354374, 14: 0.002232142857142857}, 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 150, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.02      0.01      0.01       344\n",
      "           1       0.07      0.03      0.04       562\n",
      "           2       0.06      0.09      0.08       431\n",
      "           3       0.04      0.04      0.04       441\n",
      "           4       0.12      0.09      0.10       833\n",
      "           5       0.07      0.04      0.05       852\n",
      "           6       0.04      0.02      0.03       560\n",
      "           7       0.01      0.01      0.01       461\n",
      "           8       0.04      0.04      0.04       451\n",
      "           9       0.05      0.02      0.03       637\n",
      "          10       0.09      0.18      0.12       432\n",
      "          11       0.10      0.10      0.10       819\n",
      "          12       0.12      0.04      0.06      1038\n",
      "          13       0.06      0.02      0.03       805\n",
      "          14       0.02      0.13      0.03       112\n",
      "          15       0.02      0.03      0.03       241\n",
      "          16       0.06      0.06      0.06       534\n",
      "          17       0.04      0.03      0.04       400\n",
      "          18       0.04      0.09      0.05       420\n",
      "          19       0.06      0.17      0.09       504\n",
      "          20       0.07      0.03      0.04       876\n",
      "          21       0.03      0.03      0.03       409\n",
      "          22       0.05      0.05      0.05       444\n",
      "          23       0.07      0.08      0.07       618\n",
      "          24       0.02      0.08      0.03       157\n",
      "\n",
      "    accuracy                           0.06     13381\n",
      "   macro avg       0.05      0.06      0.05     13381\n",
      "weighted avg       0.06      0.06      0.05     13381\n",
      "\n",
      "10 100\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 200, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.03      0.02      0.02       417\n",
      "           1       0.07      0.03      0.04       667\n",
      "           2       0.06      0.07      0.06       501\n",
      "           3       0.05      0.06      0.05       508\n",
      "           4       0.17      0.13      0.15      1061\n",
      "           5       0.08      0.02      0.03      1031\n",
      "           6       0.04      0.02      0.02       653\n",
      "           7       0.06      0.05      0.05       558\n",
      "           8       0.04      0.03      0.04       519\n",
      "           9       0.05      0.03      0.03       753\n",
      "          10       0.12      0.27      0.16       560\n",
      "          11       0.11      0.10      0.11       982\n",
      "          12       0.09      0.04      0.05      1176\n",
      "          13       0.07      0.03      0.04       906\n",
      "          14       0.01      0.10      0.02       123\n",
      "          15       0.04      0.07      0.05       292\n",
      "          16       0.05      0.07      0.06       606\n",
      "          17       0.05      0.04      0.04       492\n",
      "          18       0.04      0.09      0.06       492\n",
      "          19       0.07      0.19      0.10       625\n",
      "          20       0.10      0.03      0.05      1038\n",
      "          21       0.04      0.05      0.05       472\n",
      "          22       0.03      0.03      0.03       526\n",
      "          23       0.12      0.09      0.11       775\n",
      "          24       0.01      0.06      0.02       181\n",
      "\n",
      "    accuracy                           0.07     15914\n",
      "   macro avg       0.06      0.07      0.06     15914\n",
      "weighted avg       0.08      0.07      0.06     15914\n",
      "\n",
      "10 350\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py:691: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': {16: 0.0003756574004507889, 10: 0.00031466331025802394, 19: 0.00030211480362537764, 13: 0.00025967281225655674, 5: 0.00020296326364927948, 23: 0.0002498750624687656, 11: 0.00022675736961451248, 0: 0.000376081233546446, 4: 0.00010296540362438221, 24: 0.0009074410163339383, 6: 0.00032362459546925567, 9: 0.00027063599458728013, 3: 0.0003861003861003861, 1: 0.000328515111695138, 7: 0.0003322259136212625, 15: 0.00044742729306487697, 2: 0.00032509752925877764, 12: 0.00019113149847094801, 21: 0.0004640371229698376, 22: 0.00041476565740356696, 20: 0.0002003606491685033, 18: 0.0004342162396873643, 17: 0.00028376844494892167, 8: 0.00042408821034775233, 14: 0.0018083182640144665}, 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 175, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.22      0.13      0.16       664\n",
      "           1       0.09      0.03      0.05       761\n",
      "           2       0.27      0.22      0.24       769\n",
      "           3       0.12      0.11      0.11       647\n",
      "           4       0.65      0.47      0.55      2428\n",
      "           5       0.07      0.02      0.03      1232\n",
      "           6       0.07      0.02      0.04       773\n",
      "           7       0.11      0.05      0.07       753\n",
      "           8       0.04      0.04      0.04       589\n",
      "           9       0.09      0.03      0.05       923\n",
      "          10       0.19      0.38      0.26       794\n",
      "          11       0.11      0.15      0.13      1103\n",
      "          12       0.11      0.06      0.08      1308\n",
      "          13       0.07      0.11      0.09       963\n",
      "          14       0.01      0.17      0.03       138\n",
      "          15       0.37      0.34      0.35       559\n",
      "          16       0.04      0.10      0.06       666\n",
      "          17       0.36      0.23      0.28       881\n",
      "          18       0.06      0.10      0.07       576\n",
      "          19       0.11      0.22      0.15       827\n",
      "          20       0.14      0.05      0.07      1248\n",
      "          21       0.03      0.04      0.04       539\n",
      "          22       0.04      0.04      0.04       603\n",
      "          23       0.17      0.18      0.18      1001\n",
      "          24       0.12      0.21      0.16       275\n",
      "\n",
      "    accuracy                           0.16     21020\n",
      "   macro avg       0.15      0.14      0.13     21020\n",
      "weighted avg       0.19      0.16      0.16     21020\n",
      "\n",
      "50 100\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 150, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.02      0.01      0.02        75\n",
      "           1       0.09      0.06      0.07       108\n",
      "           2       0.09      0.07      0.08        72\n",
      "           3       0.06      0.09      0.07        70\n",
      "           4       0.22      0.23      0.22       236\n",
      "           5       0.11      0.02      0.04       185\n",
      "           6       0.13      0.07      0.09        97\n",
      "           7       0.06      0.05      0.05        99\n",
      "           8       0.02      0.01      0.02        70\n",
      "           9       0.07      0.03      0.05       121\n",
      "          10       0.19      0.38      0.25       132\n",
      "          11       0.12      0.15      0.13       169\n",
      "          12       0.03      0.01      0.01       142\n",
      "          13       0.06      0.08      0.07       104\n",
      "          14       0.00      0.00      0.00        11\n",
      "          15       0.12      0.17      0.14        53\n",
      "          16       0.05      0.07      0.06        76\n",
      "          17       0.05      0.04      0.04        96\n",
      "          18       0.06      0.16      0.09        74\n",
      "          19       0.08      0.16      0.10       127\n",
      "          20       0.12      0.07      0.09       168\n",
      "          21       0.01      0.02      0.01        66\n",
      "          22       0.04      0.03      0.04        86\n",
      "          23       0.23      0.26      0.24       162\n",
      "          24       0.05      0.04      0.04        24\n",
      "\n",
      "    accuracy                           0.11      2623\n",
      "   macro avg       0.08      0.09      0.08      2623\n",
      "weighted avg       0.10      0.11      0.10      2623\n",
      "\n",
      "50 150\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py:691: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 150, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.03      0.03      0.03       107\n",
      "           1       0.12      0.06      0.08       139\n",
      "           2       0.08      0.06      0.07       108\n",
      "           3       0.07      0.09      0.08        92\n",
      "           4       0.29      0.34      0.31       365\n",
      "           5       0.08      0.02      0.04       252\n",
      "           6       0.07      0.02      0.03       130\n",
      "           7       0.10      0.07      0.08       145\n",
      "           8       0.03      0.02      0.02        95\n",
      "           9       0.17      0.05      0.08       170\n",
      "          10       0.23      0.47      0.31       194\n",
      "          11       0.11      0.13      0.12       219\n",
      "          12       0.05      0.01      0.02       190\n",
      "          13       0.04      0.06      0.05       130\n",
      "          14       0.00      0.00      0.00        14\n",
      "          15       0.18      0.27      0.22        85\n",
      "          16       0.07      0.08      0.07        98\n",
      "          17       0.04      0.04      0.04       133\n",
      "          18       0.08      0.25      0.13       100\n",
      "          19       0.12      0.22      0.16       184\n",
      "          20       0.11      0.08      0.09       225\n",
      "          21       0.04      0.03      0.03        93\n",
      "          22       0.03      0.02      0.02       113\n",
      "          23       0.30      0.33      0.32       275\n",
      "          24       0.04      0.03      0.03        35\n",
      "\n",
      "    accuracy                           0.14      3691\n",
      "   macro avg       0.10      0.11      0.10      3691\n",
      "weighted avg       0.13      0.14      0.13      3691\n",
      "\n",
      "50 250\n",
      "{'class_weight': {21: 0.0021598272138228943, 4: 0.0002955082742316785, 0: 0.0009861932938856016, 23: 0.0007173601147776184, 9: 0.0010141987829614604, 17: 0.0005892751915144372, 5: 0.0007930214115781126, 7: 0.0012970168612191958, 20: 0.000856898029134533, 1: 0.0014684287812041115, 18: 0.0019646365422396855, 11: 0.000946969696969697, 19: 0.0010235414534288639, 16: 0.002074688796680498, 10: 0.0009900990099009901, 2: 0.0009615384615384616, 12: 0.001075268817204301, 8: 0.002053388090349076, 22: 0.0017301038062283738, 3: 0.002127659574468085, 24: 0.004807692307692308, 15: 0.0008149959250203749, 13: 0.0016339869281045752, 6: 0.0015479876160990713, 14: 0.010752688172043012}, 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 125, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.31      0.29      0.30       253\n",
      "           1       0.09      0.08      0.09       170\n",
      "           2       0.36      0.38      0.37       260\n",
      "           3       0.12      0.14      0.13       117\n",
      "           4       0.55      0.49      0.52       846\n",
      "           5       0.10      0.03      0.05       315\n",
      "           6       0.07      0.06      0.06       161\n",
      "           7       0.09      0.07      0.08       193\n",
      "           8       0.04      0.02      0.03       122\n",
      "           9       0.05      0.02      0.03       246\n",
      "          10       0.23      0.45      0.31       253\n",
      "          11       0.16      0.18      0.17       264\n",
      "          12       0.02      0.01      0.02       232\n",
      "          13       0.06      0.08      0.07       153\n",
      "          14       0.00      0.00      0.00        23\n",
      "          15       0.47      0.53      0.50       307\n",
      "          16       0.05      0.06      0.05       121\n",
      "          17       0.39      0.40      0.40       425\n",
      "          18       0.09      0.22      0.13       127\n",
      "          19       0.15      0.22      0.17       244\n",
      "          20       0.12      0.09      0.10       292\n",
      "          21       0.03      0.03      0.03       116\n",
      "          22       0.06      0.04      0.05       145\n",
      "          23       0.25      0.32      0.28       349\n",
      "          24       0.13      0.08      0.10        52\n",
      "\n",
      "    accuracy                           0.24      5786\n",
      "   macro avg       0.16      0.17      0.16      5786\n",
      "weighted avg       0.24      0.24      0.24      5786\n",
      "\n",
      "50 350\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 175, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.30      0.34      0.32       322\n",
      "           1       0.13      0.07      0.09       202\n",
      "           2       0.39      0.43      0.41       340\n",
      "           3       0.27      0.27      0.27       209\n",
      "           4       0.74      0.64      0.69      1604\n",
      "           5       0.09      0.02      0.03       386\n",
      "           6       0.25      0.09      0.14       216\n",
      "           7       0.19      0.12      0.14       294\n",
      "           8       0.02      0.01      0.02       140\n",
      "           9       0.06      0.02      0.03       291\n",
      "          10       0.31      0.47      0.38       366\n",
      "          11       0.09      0.12      0.10       290\n",
      "          12       0.10      0.05      0.07       274\n",
      "          13       0.04      0.14      0.07       161\n",
      "          14       0.00      0.00      0.00        27\n",
      "          15       0.47      0.58      0.52       320\n",
      "          16       0.07      0.08      0.07       135\n",
      "          17       0.37      0.43      0.40       485\n",
      "          18       0.07      0.19      0.11       158\n",
      "          19       0.22      0.22      0.22       329\n",
      "          20       0.17      0.11      0.14       378\n",
      "          21       0.03      0.03      0.03       133\n",
      "          22       0.03      0.02      0.02       163\n",
      "          23       0.23      0.38      0.28       387\n",
      "          24       0.58      0.45      0.51       119\n",
      "\n",
      "    accuracy                           0.31      7729\n",
      "   macro avg       0.21      0.21      0.20      7729\n",
      "weighted avg       0.32      0.31      0.31      7729\n",
      "\n",
      "100 150\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 125, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.12      0.06      0.08        33\n",
      "           1       0.04      0.03      0.03        33\n",
      "           2       0.03      0.03      0.03        36\n",
      "           3       0.14      0.18      0.16        22\n",
      "           4       0.32      0.46      0.38       132\n",
      "           5       0.15      0.09      0.11        70\n",
      "           6       0.00      0.00      0.00        34\n",
      "           7       0.09      0.09      0.09        46\n",
      "           8       0.00      0.00      0.00        25\n",
      "           9       0.12      0.06      0.08        50\n",
      "          10       0.29      0.38      0.33        63\n",
      "          11       0.15      0.04      0.06        52\n",
      "          12       0.04      0.02      0.03        49\n",
      "          13       0.00      0.00      0.00        26\n",
      "          14       0.00      0.00      0.00         3\n",
      "          15       0.34      0.35      0.35        34\n",
      "          16       0.03      0.04      0.03        23\n",
      "          17       0.10      0.08      0.09        37\n",
      "          18       0.04      0.07      0.05        27\n",
      "          19       0.24      0.34      0.28        59\n",
      "          20       0.12      0.10      0.11        58\n",
      "          21       0.12      0.11      0.12        27\n",
      "          22       0.09      0.07      0.08        28\n",
      "          23       0.42      0.55      0.48       115\n",
      "          24       0.11      0.08      0.10        12\n",
      "\n",
      "    accuracy                           0.20      1094\n",
      "   macro avg       0.12      0.13      0.12      1094\n",
      "weighted avg       0.18      0.20      0.18      1094\n",
      "\n",
      "100 250\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': {17: 0.0007598784194528875, 9: 0.0019723865877712033, 10: 0.00205761316872428, 23: 0.001321003963011889, 6: 0.0038461538461538464, 5: 0.0018867924528301887, 22: 0.004201680672268907, 4: 0.0004079967360261118, 2: 0.0013280212483399733, 19: 0.0021008403361344537, 15: 0.0009794319294809011, 18: 0.004629629629629629, 1: 0.0038910505836575876, 20: 0.001996007984031936, 0: 0.001392757660167131, 24: 0.008695652173913044, 12: 0.0027247956403269754, 21: 0.005, 11: 0.002583979328165375, 7: 0.00267379679144385, 16: 0.005494505494505495, 13: 0.005076142131979695, 3: 0.005263157894736842, 8: 0.004784688995215311, 14: 0.02127659574468085}, 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 200, 'random_state': 42}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.35      0.44      0.39       179\n",
      "           1       0.07      0.08      0.07        64\n",
      "           2       0.35      0.53      0.43       188\n",
      "           3       0.07      0.06      0.07        48\n",
      "           4       0.66      0.59      0.62       613\n",
      "           5       0.12      0.05      0.07       133\n",
      "           6       0.06      0.05      0.05        65\n",
      "           7       0.06      0.06      0.06        94\n",
      "           8       0.07      0.04      0.05        52\n",
      "           9       0.11      0.02      0.04       127\n",
      "          10       0.35      0.51      0.41       122\n",
      "          11       0.15      0.11      0.13        97\n",
      "          12       0.06      0.03      0.04        92\n",
      "          13       0.00      0.00      0.00        49\n",
      "          14       0.00      0.00      0.00        12\n",
      "          15       0.51      0.65      0.57       255\n",
      "          16       0.04      0.04      0.04        46\n",
      "          17       0.44      0.50      0.46       329\n",
      "          18       0.09      0.13      0.10        54\n",
      "          19       0.17      0.26      0.20       119\n",
      "          20       0.19      0.17      0.18       125\n",
      "          21       0.02      0.02      0.02        50\n",
      "          22       0.09      0.07      0.08        59\n",
      "          23       0.38      0.38      0.38       189\n",
      "          24       0.09      0.04      0.05        28\n",
      "\n",
      "    accuracy                           0.35      3189\n",
      "   macro avg       0.18      0.19      0.18      3189\n",
      "weighted avg       0.33      0.35      0.33      3189\n",
      "\n",
      "100 350\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py:691: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 100, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.36      0.44      0.39       249\n",
      "           1       0.07      0.03      0.04        96\n",
      "           2       0.42      0.53      0.47       268\n",
      "           3       0.34      0.49      0.40       140\n",
      "           4       0.76      0.73      0.74      1371\n",
      "           5       0.14      0.06      0.08       203\n",
      "           6       0.26      0.15      0.19       119\n",
      "           7       0.20      0.18      0.18       194\n",
      "           8       0.21      0.08      0.12        71\n",
      "           9       0.03      0.01      0.02       172\n",
      "          10       0.48      0.52      0.50       235\n",
      "          11       0.07      0.06      0.06       122\n",
      "          12       0.18      0.11      0.14       134\n",
      "          13       0.03      0.04      0.03        57\n",
      "          14       0.00      0.00      0.00        15\n",
      "          15       0.48      0.57      0.52       268\n",
      "          16       0.05      0.05      0.05        60\n",
      "          17       0.44      0.51      0.48       390\n",
      "          18       0.11      0.14      0.12        85\n",
      "          19       0.24      0.34      0.28       204\n",
      "          20       0.22      0.19      0.20       211\n",
      "          21       0.07      0.07      0.07        67\n",
      "          22       0.03      0.03      0.03        78\n",
      "          23       0.27      0.37      0.31       228\n",
      "          24       0.52      0.54      0.53        95\n",
      "\n",
      "    accuracy                           0.42      5132\n",
      "   macro avg       0.24      0.25      0.24      5132\n",
      "weighted avg       0.41      0.42      0.41      5132\n",
      "\n",
      "150 250\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 10, 'max_features': 'auto', 'n_estimators': 100, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.34      0.46      0.39       147\n",
      "           1       0.00      0.00      0.00        31\n",
      "           2       0.32      0.62      0.43       152\n",
      "           3       0.00      0.00      0.00        25\n",
      "           4       0.72      0.58      0.64       483\n",
      "           5       0.07      0.02      0.03        65\n",
      "           6       0.06      0.06      0.06        31\n",
      "           7       0.08      0.12      0.10        48\n",
      "           8       0.06      0.04      0.04        28\n",
      "           9       0.08      0.05      0.06        77\n",
      "          10       0.40      0.43      0.42        60\n",
      "          11       0.11      0.09      0.10        46\n",
      "          12       0.06      0.02      0.03        43\n",
      "          13       0.00      0.00      0.00        23\n",
      "          14       0.00      0.00      0.00         9\n",
      "          15       0.54      0.58      0.56       223\n",
      "          16       0.00      0.00      0.00        23\n",
      "          17       0.50      0.44      0.47       292\n",
      "          18       0.11      0.29      0.15        28\n",
      "          19       0.22      0.36      0.27        61\n",
      "          20       0.19      0.15      0.17        68\n",
      "          21       0.00      0.00      0.00        23\n",
      "          22       0.00      0.00      0.00        32\n",
      "          23       0.27      0.20      0.23        75\n",
      "          24       0.06      0.18      0.09        17\n",
      "\n",
      "    accuracy                           0.38      2110\n",
      "   macro avg       0.17      0.19      0.17      2110\n",
      "weighted avg       0.38      0.38      0.37      2110\n",
      "\n",
      "150 350\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 50, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.30      0.43      0.35       216\n",
      "           1       0.19      0.06      0.10        63\n",
      "           2       0.41      0.58      0.48       232\n",
      "           3       0.34      0.45      0.39       117\n",
      "           4       0.78      0.75      0.77      1241\n",
      "           5       0.16      0.08      0.11       135\n",
      "           6       0.17      0.09      0.12        86\n",
      "           7       0.17      0.16      0.17       149\n",
      "           8       0.04      0.02      0.03        47\n",
      "           9       0.05      0.02      0.03       122\n",
      "          10       0.62      0.51      0.56       173\n",
      "          11       0.05      0.03      0.04        71\n",
      "          12       0.14      0.07      0.09        85\n",
      "          13       0.06      0.03      0.04        31\n",
      "          14       0.00      0.00      0.00        13\n",
      "          15       0.54      0.64      0.59       236\n",
      "          16       0.00      0.00      0.00        38\n",
      "          17       0.43      0.56      0.48       352\n",
      "          18       0.12      0.15      0.13        59\n",
      "          19       0.24      0.31      0.27       145\n",
      "          20       0.21      0.24      0.22       154\n",
      "          21       0.00      0.00      0.00        40\n",
      "          22       0.13      0.04      0.06        51\n",
      "          23       0.16      0.19      0.17       113\n",
      "          24       0.55      0.61      0.58        84\n",
      "\n",
      "    accuracy                           0.46      4053\n",
      "   macro avg       0.23      0.24      0.23      4053\n",
      "weighted avg       0.44      0.46      0.45      4053\n",
      "\n",
      "200 350\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': {5: 0.0026455026455026454, 4: 0.0002244668911335578, 17: 0.0010070493454179255, 20: 0.0022675736961451248, 10: 0.0017667844522968198, 2: 0.002320185614849188, 3: 0.0024154589371980675, 7: 0.002105263157894737, 0: 0.001736111111111111, 8: 0.007751937984496124, 13: 0.0136986301369863, 9: 0.0028169014084507044, 23: 0.0041841004184100415, 19: 0.0022727272727272726, 1: 0.005747126436781609, 18: 0.005747126436781609, 12: 0.004132231404958678, 24: 0.003401360544217687, 16: 0.010309278350515464, 11: 0.005847953216374269, 22: 0.00847457627118644, 6: 0.003875968992248062, 14: 0.03571428571428571, 15: 0.004975124378109453, 21: 0.009708737864077669}, 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 150, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.31      0.51      0.39       144\n",
      "           1       0.10      0.05      0.06        43\n",
      "           2       0.66      0.60      0.63       108\n",
      "           3       0.34      0.46      0.39       103\n",
      "           4       0.85      0.77      0.81      1114\n",
      "           5       0.18      0.06      0.09        95\n",
      "           6       0.43      0.23      0.30        65\n",
      "           7       0.27      0.26      0.27       119\n",
      "           8       0.15      0.06      0.09        32\n",
      "           9       0.04      0.02      0.03        89\n",
      "          10       0.64      0.61      0.63       141\n",
      "          11       0.00      0.00      0.00        43\n",
      "          12       0.33      0.18      0.23        61\n",
      "          13       0.00      0.00      0.00        18\n",
      "          14       0.00      0.00      0.00         7\n",
      "          15       0.65      0.40      0.49        50\n",
      "          16       0.00      0.00      0.00        24\n",
      "          17       0.41      0.58      0.48       248\n",
      "          18       0.13      0.19      0.15        43\n",
      "          19       0.24      0.30      0.27       110\n",
      "          20       0.19      0.29      0.23       110\n",
      "          21       0.00      0.00      0.00        26\n",
      "          22       0.00      0.00      0.00        30\n",
      "          23       0.15      0.32      0.21        60\n",
      "          24       0.52      0.73      0.61        74\n",
      "\n",
      "    accuracy                           0.51      2957\n",
      "   macro avg       0.26      0.26      0.25      2957\n",
      "weighted avg       0.52      0.51      0.51      2957\n",
      "\n",
      "250 350\n",
      "{'class_weight': {4: 0.00032414910858995135, 17: 0.004016064257028112, 0: 0.0035971223021582736, 7: 0.0024691358024691358, 10: 0.0022026431718061676, 13: 0.030303030303030304, 6: 0.004608294930875576, 24: 0.003745318352059925, 12: 0.005952380952380952, 11: 0.009708737864077669, 5: 0.003484320557491289, 2: 0.003134796238244514, 9: 0.005494505494505495, 16: 0.017241379310344827, 3: 0.0027100271002710027, 14: 0.06666666666666667, 23: 0.0064516129032258064, 1: 0.007751937984496124, 20: 0.0029069767441860465, 19: 0.0029239766081871343, 18: 0.0078125, 8: 0.013157894736842105, 15: 0.019230769230769232, 22: 0.013513513513513514, 21: 0.014705882352941176}, 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 150, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.26      0.51      0.34        70\n",
      "           1       0.00      0.00      0.00        32\n",
      "           2       0.73      0.65      0.69        80\n",
      "           3       0.37      0.61      0.46        92\n",
      "           4       0.84      0.81      0.83       771\n",
      "           5       0.16      0.08      0.11        72\n",
      "           6       0.32      0.24      0.27        54\n",
      "           7       0.19      0.14      0.16       101\n",
      "           8       0.08      0.05      0.06        19\n",
      "           9       0.18      0.04      0.07        46\n",
      "          10       0.74      0.60      0.66       113\n",
      "          11       0.00      0.00      0.00        26\n",
      "          12       0.25      0.17      0.20        42\n",
      "          13       0.00      0.00      0.00         8\n",
      "          14       0.00      0.00      0.00         3\n",
      "          15       0.00      0.00      0.00        13\n",
      "          16       0.00      0.00      0.00        15\n",
      "          17       0.24      0.32      0.27        62\n",
      "          18       0.14      0.16      0.15        32\n",
      "          19       0.26      0.40      0.31        86\n",
      "          20       0.27      0.36      0.31        86\n",
      "          21       0.00      0.00      0.00        17\n",
      "          22       0.00      0.00      0.00        19\n",
      "          23       0.19      0.31      0.24        39\n",
      "          24       0.63      0.85      0.73        67\n",
      "\n",
      "    accuracy                           0.53      1965\n",
      "   macro avg       0.23      0.25      0.23      1965\n",
      "weighted avg       0.52      0.53      0.52      1965\n",
      "\n"
     ]
    }
   ],
   "source": [
    "graph_features_all = graph_features_clean\n",
    "scaler = StandardScaler()\n",
    "\n",
    "\n",
    "clf = RandomForestClassifier(random_state=42)\n",
    "\n",
    "cv = StratifiedKFold(\n",
    "    n_splits=5\n",
    ")\n",
    "cv_results_rf_topic_no = {}\n",
    "for min_replies, max_replies in tqdm([(10, 50), (10, 100), (10, 350), (50,100), \n",
    "                                      (50, 150), (50, 250), (50, 350), (100, 150), \n",
    "                                      (100, 250), (100, 350), (150,250), (150, 350), \n",
    "                                      (200, 350), (250, 350)]):\n",
    "    print(min_replies, max_replies)\n",
    "    filter_array = filter_on_stats(stat_df, min_replies=min_replies, max_replies=max_replies)\n",
    "    graph_features = clean_features[filter_array]\n",
    "    graph_features = graph_features.dropna()\n",
    "    X = graph_features.drop(columns=[\"dominant_topic\", \"replies\"]).to_numpy()\n",
    "    y = graph_features.dominant_topic.to_numpy()\n",
    "    scaler.fit(X)\n",
    "    X_transformed = scaler.transform(X)\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X_transformed, y, test_size=0.2, \n",
    "                                                        random_state=42, stratify=y)\n",
    "    count = Counter(y_train)\n",
    "    \n",
    "    param_grid = { \n",
    "    'n_estimators': [50, 100, 125, 150, 175, 200],\n",
    "    'max_features': ['auto'],\n",
    "    'max_depth' : [10, 13, 15, None],\n",
    "    'criterion' :['gini'],\n",
    "    'class_weight': ['balanced', {label:1/count[label] for label in count.keys()}],\n",
    "    'random_state': [42]\n",
    "    }\n",
    "\n",
    "    search = GridSearchCV(\n",
    "        estimator=clf, param_grid=param_grid,\n",
    "        scoring='f1_macro', cv=cv,\n",
    "        n_jobs=30, refit = True\n",
    "    )\n",
    "    search.fit(X_train, y_train)\n",
    "    print(search.best_params_)\n",
    "    key = f\"{min_replies}-{max_replies}\"\n",
    "    cv_results_rf_topic_no[key] = {\"model\":search, \n",
    "                              'X_test': X_test, \n",
    "                              'X_train': X_train, \n",
    "                              'y_test':y_test, \n",
    "                              \"y_train\":y_train\n",
    "                             }\n",
    "    cv_results_rf_topic_no[key]['report'] = classification_report(cv_results_rf_topic_no[key]['y_test'], cv_results_rf_topic_no[key]['model'].predict(cv_results_rf_topic_no[key]['X_test']))\n",
    "    cv_results_rf_topic_no[key]['report_dict'] = classification_report(cv_results_rf_topic_no[key]['y_test'], cv_results_rf_topic_no[key]['model'].predict(cv_results_rf_topic_no[key]['X_test']), output_dict=True)\n",
    "    cv_results_rf_topic_no[key]['best_params'] = search.best_params_\n",
    "    print(classification_report(cv_results_rf_topic_no[key]['y_test'], cv_results_rf_topic_no[key]['model'].predict(cv_results_rf_topic_no[key]['X_test'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"b_collection_extracted/rf_cv_results_rf_topic_no\", \"wb\") as f:\n",
    "    pickle.dump(cv_results_rf_topic_no, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"b_collection_extracted/rf_cv_results_rf_topic_no\", \"rb\") as f:\n",
    "    cv_results_rf_topic_no = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "viz_dict_rf_topic_no = {\"precision\":{},\n",
    "                        \"precision_weighted\":{},\n",
    "                        \"recall\":{},\n",
    "                        \"recall_weighted\":{},\n",
    "                        \"f1\":{},\n",
    "                        \"f1_weigthed\":{}\n",
    "                       }\n",
    "for key in cv_results_rf_topic_no.keys():\n",
    "    df =  pd.DataFrame(cv_results_rf_topic_no[key]['report_dict']).transpose()\n",
    "    viz_dict_rf_topic_no[\"precision\"][key] = df.precision.loc[\"macro avg\"]\n",
    "    viz_dict_rf_topic_no[\"precision_weighted\"][key] = df.precision.loc[\"weighted avg\"]\n",
    "    viz_dict_rf_topic_no[\"recall\"][key] = df.recall.loc[\"macro avg\"]\n",
    "    viz_dict_rf_topic_no[\"recall_weighted\"][key] = df.recall.loc[\"weighted avg\"]\n",
    "    viz_dict_rf_topic_no[\"f1\"][key] = df[\"f1-score\"].loc[\"macro avg\"]\n",
    "    viz_dict_rf_topic_no[\"f1_weigthed\"][key] = df[\"f1-score\"].loc[\"weighted avg\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f332e87c390>"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "plt.figure(figsize=(12,7))\n",
    "sns.lineplot(data=viz_dict_rf_topic_no).set_xlabel(\"Spanne an Antworten\")\n",
    "plt.legend([\"Genauigkeit\", \"Genauigkeit gewichtet\", \"Sensitivität\", \"Sensitivität gewichtet\", \"F1\", \"F1 gewichtet\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Tatsächliche Topic Nummer')"
      ]
     },
     "execution_count": 123,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(12,10))\n",
    "g=plot_confusion_matrix(estimator=cv_results_rf_topic_no[\"10-350\"][\"model\"], X=cv_results_rf_topic_no[\"10-350\"][\"X_test\"], y_true=cv_results_rf_topic_no[\"10-350\"][\"y_test\"],cmap=\"magma\", ax=ax, xticks_rotation=\"horizontal\")\n",
    "g.ax_.set_xlabel('Vorhergesagte Topic Nummer')\n",
    "g.ax_.set_ylabel('Tatsächliche Topic Nummer')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Tatsächliche Topic Nummer')"
      ]
     },
     "execution_count": 153,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAp4AAAJJCAYAAAADTTcEAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAADPD0lEQVR4nOzdeXxM5/4H8M9sCRKCJBIStaXIpZUgYimpJYJILK1aeinV9qJNaIrGdi3d9FZVq9dFIpGFIEFSJWiiKorsixqJWkJkEYSIoJLM8/sjPyktJuacZ2aO+b77Oq+XnGY+833Oec6ZJ2fOImOMMRBCCCGEEMKZ3NAFEEIIIYQQ00ADT0IIIYQQohc08CSEEEIIIXpBA09CCCGEEKIXNPAkhBBCCCF6QQNPQgghhBCiF0pDF0AIIYQQYqrOnElGdbVML++lUjXDiy++qJf3ehIaeBJCCCGEGEh1tQydOt/Ry3udyWusl/d5GvqqnRBCCCGE6AUd8SSEEEIIMRgGaDSGLkJvJDHwTEw8hJLiq6JmaliVqHmEEEIIkZaXXuoKFxcXQ5dhUiQx8Cwpvop3ps8TNfPe/SJR8wghPPE68Z5xyjVlUlpXvGrlcRZbDYdMXngsVz7bqlp9kkvuMzOhI550jichhBBCCNELSRzxJIQQQgh5LjHQEU9CCCGEEELERkc8CSGEEEIMhgHMdM43l+wRT3NzFQ4nLcXx5E+Rmv45Fi0eU/f/li57HZk5/0F65krMnOUp6H28vNxwSh2K3LwwzJ8/QWjZ3DIdHW2RkLAKOSc3ITsnGH5+Y7S/6DnJfEDs5SqlWoOC56KoOBpZ2UEiVPcnHn2VR665uQrHj69FesZ6ZOcEYenSKSJUWUsq+wAefYBXvwIAuVyO1LT/Ie6HT0TJk1KtHTs6Ii19Xd1UdmMX/P2F71+ksh/gtb3y2l8RkTED+OWXX9jQoUPZkCFD2IYNG7T+fmTENtbArPXfJutmTqyBWWtm2agtO34imPV/5VX2zvRJLDz8E9bQ/AXWwKw1a+3wj8e+FlBoneRyFcs7s4W1a+fIVCpzlp4RzJyd29XrtfrMBBTM3t6Wubp2YoCCWVpaMvXpCMG5UsnktVylVGv//i7M1bUTy8oOEVwfzzp1z1VqnSwsLBmgZEqlOTt2bB1zd3+pHq/T/zLgtVx59AHdMrWvK0DJ5swZzyIil7C4H76o52ukVKtZvSe5vAG7XLiTvfCCo5bfldJ+gMf2yme7UqvVehj1PN2p346z6tv79DIZQ3v1fsSzpqYGK1asQHBwMPbu3Ysff/wRZ8+e1SmrsvIPAIBKpYBKqQBjDO+8NwgrP48F+//D1levVuhca69enXDuXBEuXChGVVU1dmw/DF/ffjrn8coEgJKSMmRm1i7H27fvIjf3EhwcbEwiE+CzXKVUa1LSSZSV6d7XH4dXX+WVW1l5DwCgUimhVCnr9gFCSGkfwKMP8MgEAAcHG4wY4Y6QTfGiZUqp1ocNHuyC8+eKcelSqeAsKe0HxN5eedVJxKf3gWdOTg7atGmD1q1bw8zMDN7e3khMTNQpSy6X4diJT3Dh0vc4dOg3pKWeR7t2dnjtdXccObocu2I/QocOdjrX2srBBgUFf+4MLhdeRSsHa53zeGX+VZs2dnBxcUJycq7JZPJerlKqVSy86uSVK5fLkZa+HsUl0UhMyEBKinGuK6msf55WfzMTgYFB0EjgSl7etb4x/lVs23aYS7YYpLK9Snq7Yqi9ql0fkxHQ+8DzypUrsLe3r/vZzs4OV65c0SlLo2Ho23sJOjnNQc+e7fGPfzjA3FyJe39UYcArS7E59Bf8b8M7YpUuCRYWDbAjeikCAtahouKOSWXyIqVaTZlGo0HPHjPQ5oWJcHPrhC5d2hq6JPIY3t7uKC29iYyM3w1dila8a1WplPDx6Y2YmCNc8o0Zba+mS7IXFz2svPwOjvxyGkOGvoyiwjL8EJsGAPghLg1durbWObeo8Bpat25R97Ojgy2KCq8LqpVH5gNKpQLRMcsQtTURsbuPmlQmr+UqpVrFxqtO3u0vL6/E4cPZ8PLqKThLavsAKejbtwt8fPrg7LkIbNm6CAMHuiAs/GNDl/VYvGsdNtwNmZlnUVp6U7RMsUllezX17UpK9D7wtLOzQ0lJSd3PV65cgZ3ds38dbmPTGFZWjQAADRqoMGhwV5zJK8aePRkY4PEPAED//p1x9mzJ02KeKjU1D05ODmjb1h4qlRJvjH8Ve/Yc0zmPV+YDQcFzcfr0RaxZs1OUPCll8lquUqpVbLzq5JFrY2MFKysLAECDBmYYMqQ78vIKjLJWqax/XhYtCkHbNpPg1GEy3pz0GX7+OQtvTfnS0GU9Fu9aJ0ww7q/ZAelsr9LerphJfdWu9/t4vvTSS8jPz0dBQQHs7Oywd+9efP3118+cY2ffFBuD3oNCIYNcLseuncnYH5+F48fOYFPoDHzg54XblX/g/ZmbdK61pkaD2f5rsS9+JRQKOTaH7odafVHnPF6ZANCvX1dMnuyJnJzzSEtfDwBYsjgE8fEpz30mwGe5SqnWyC0L4eHRDTY2Vsi/GIXly8MQGrLf6OrklduyZXOEhM6HQiGHXC5DTPQR7N2bbJS18lquPPoAj0xepFQrADRqZI4hQ7pj5oxvRcuUyn6Ax/bKa7si4pMxMS79fEa//PILPv/8c9TU1OC1117DzJkzn/r7WyK3453p80St4d79IlHzCCE8yTjl6n33ZwKktK541crjy8QaDpm88FiufLZVtfoknJ2duWTXu4aTx9HJUfhdDerjTElHg7fXIE8u8vDwgIeHhyHemhBCCCGEGMhzcXERIYQQQog0Gdc5nrdu3YK/vz+GDRuG4cOHIzMzEzdv3sS0adMwdOhQTJs2DeXl5bWVM4ZPP/0Unp6e8PHxwalTp7Tm08CTEEIIIYQAAD777DP0798f+/fvR1xcHDp06ICNGzeiT58+OHjwIPr06YONGzcCAI4cOYL8/HwcPHgQn3zyCZYtW6Y1nwaehBBCCCEGJGMavUzaVFRUIDU1Fa+//joAwMzMDE2aNEFiYiJGjx4NABg9ejQSEhIAoG6+TCaDi4sLbt26hdLSp5+vSgNPQgghhBCCy5cvo3nz5liwYAFGjx6NRYsW4c6dO7h+/TpatKi9T6qtrS2uX6+9R+pfHwpkb2+v9aFANPAkhBBCCDEUPT4ys6ysDGPHjq2btm/f/kgp1dXVUKvVmDhxImJjY9GwYcO6r9UfkMlkkMl0v3OBQa5qf1YaVoV793V7rObzQCYz45LLWBWPVA6ZhFC/kg4prStetUrp1kc8SKkPmJbmzZtj165dT/z/9vb2sLe3R7du3QAAw4YNw8aNG2FtbY3S0lK0aNECpaWlaN68OYC/PxSopKRE60OB6IgnIYQQQojBMECjp0kLW1tb2Nvb4/z58wCA48ePo0OHDhg0aBBiY2MBALGxsRg8eDAA1M1njCErKwuNGzeu+0r+SSRxxJMQQgghhPC3ZMkSzJ07F1VVVWjdujW++OILaDQazJkzBzExMWjVqhXWrFkDoPa+7L/88gs8PT3RsGFDfP7551rzaeBJCCGEEGJIRvIcdQBwdnZ+7NfxYWFhf5snk8mwdOnSZ8qnr9oJIYQQQohe0BFPQgghhBBDeXBVu4l4bo54nj0Xhsys9UhLX4cTyWtFy/XycsMpdShy88Iwf/4Eo8oMCpqDwqKtyMxaVzfvtddeQVb2//DH/R/Ro8eLguo0N1fh+PG1SM9Yj+ycICxdOkVQ3gM8lqmjoy0SElYh5+QmZOcEw89vjCi5QcFzUVQcjazsIFHypJTJM9eYtyt95PLIpHUlfq6U9isALVdey5WIjBlAYGAg6927N/P29q7X74eHRzDA7KnT2XNRzNraRuvv/TkptE5yuYrlndnC2rVzZCqVOUvPCGbOzu3q9VoxM2Wyho+d+g9wY927d2VZ2aF185ydO7JOnZxY4qFvWc+eLz/xtTJZQwYotU4WFpYMUDKl0pwdO7aOubu/pOU1+l+mgILZ29syV9dODFAwS0tLpj4dIUpu//4uzNW1E8vKDhGcJbVMXrnGsl09b7XSuhI/V0r7FVquui1XtVrNecSj3amso6ymYIteJmNor0GOeI4dOxbBwcGGeOtn0qtXJ5w7V4QLF4pRVVWNHdsPw9e3n9FkHk36DWVlFY/My80twJkzhYJqfFhl5T0AgEqlhFKlBGNMUB6PZQoAJSVlyMw8CwC4ffsucnMvwcHBRnBuUtLJvy1jU8nklWvs25VUa6V1JX6ulPYrtFz57Qf5YwDT6GcyAgYZeLq5ucHKykrUTMaA+P2fIznle7zz7nBRMls52KCg4M9njl4uvIpWDtZGl8mTXC5HWvp6FJdEIzEhAykpuYLy9NH+Nm3s4OLihORkYbUSPqS0XUmpVh6k1H7ey9XY9yu0XIlUPDcXF3kMCEBR0XXY2lph/4GVyMstQFLSb4YuS/I0Gg169pgBKysL7Ny5DF26tMWpU/mGLuuJLCwaYEf0UgQErENFxR1Dl0MIeQ7QfoUPWq4PoYuLpKeoqPaB9VevliMu9le4uXUWnll4Da1b/3kHfkcHWxQVXje6TH0oL6/E4cPZ8PLqKSiHZ/uVSgWiY5YhamsiYncfFSWTiE9K25WUauVBSu3nlSuV/QotVyIVz8XAs1Ejc1haNqz7t6dnD1GOyqWm5sHJyQFt29pDpVLijfGvYs+eY0aXyYuNjRWsrCwAAA0amGHIkO7IyysQlMmz/UHBc3H69EWsWbNTlDzCh5S2KynVyoOU2s8rVyr7FVquEsZgNI/M1Ifn4qt2O7tmiNm5FEDtX1Hbon7GgQNpgnNrajSY7b8W++JXQqGQY3PofqjVF40mMyJyPjw8XoaNTRNcyA/HiuWRKCurwJpvZ8LW1gpxPyxDdvZ5eI9YolN+y5bNERI6HwqFHHK5DDHRR7B3b7JOWQ/wWKYA0K9fV0ye7ImcnPNIS18PAFiyOATx8SmCciO3LISHRzfY2Fgh/2IUli8PQ2jIfpPI5JVr7NuVVGuldSV+rpT2K7Rc+e0HibhkTOhlyjoICAhASkoKbty4AWtra/j5+WHcuHFP/P2IiEhMmTJd5CpqRM7jRyYz45LLWBWPVA6ZhBBCiPjU6pNwdnY2bA1ZSejc+Kxe3ivvfm+Dt9cgRzxXr15tiLclhBBCCCEG9Fx81U4IIYQQIkkMdFU7IYQQQgghYqMjnoQQQgghBiIDg4yOeBJCCCGEECIuOuJJCCGEEGJI+r/BkMHQEU9CCCGEEKIXEjriKY37blpZiH9/rPLK06Jn1lJwyJTGeqol45Apnb9a+d0f9j6XXEKkw7T3LdT+Z6TPq9p5rJpnREc8CSGEEEKIXkjoiCchhBBCyPOG6e+IJ48vOp8RHfEkhBBCCCF6QQNPQgghhBCiF/RVOyGEEEKIIWn0dPEUfdUuHi8vN5xShyI3Lwzz508wutzMk8uQdHwBDh/9GImH5wEAln0yCifSFuPIsUCEb3kHTawa6pwfFDwXRcXRyMoO0jnjcaysLLB9x2L8dioYJ38LQu/ewq/aN/Z19YC5uQrHj69FesZ6ZOcEYenSKSJUWYvHMhArMyhoDgqLtiIza13dvJVfvo2Tv21ARsZ/ER2zGFZWFkZRK+9MXrmmXqupt5/XvkUqy1Vq7SciY3pWVFTE/vnPf7Lhw4ezESNGsM2bN2t9TXh4BAMUT5zkchXLO7OFtWvnyFQqc5aeEcycnds99TX1mXTJtbLo+tjp/PkY1q61+yPzRvm8zpo3eZlZWXRlq1ctZKtXLXzsa+tTa//+LszVtRPLyg55hjaaaZ1CQheyt6f7MsCMqVSNmJVVMy2vEX+Z8lpXgFLrZGFhyQAlUyrN2bFj65i7+0taXsOrVvEzZbKGj536D3Bj3bt3ZVnZoXXzPIf2ZUqlBZPJGrIvVs5iX6yc9cTXS6X9UlpXz1OtptF+/e9bjGu5Sqf9arVaDyOfp1OnHWaazO/0MhlDe/V+xFOhUCAwMBD79u3D9u3bsXXrVpw9e1ZQZq9enXDuXBEuXChGVVU1dmw/DF/ffoJr5ZX7wOFDuaipqb2SLS01Hy0dmuqclZR0EmVlFSJVVqtJk0bo3/8lhGzaDwCoqqpGeXmloEypravKynsAAJVKCaVKCSbC0yV41Cpm5tGk3/7WlxJ+yqzrq8kncuHoYGMUtfLMpFqlkym1WgHx9y1SWq6AdNpPxKf3gWeLFi3QpUsXAIClpSXat2+PK1euCMps5WCDgoLSup8vF15FKwdrQZli5zIGxMS+j8Rf5mHK1L5/+/+TJvdG4k9qnWvloV07e1y7Wo5NIR8hNe2/2LBxDho1MheUKYV19TC5XI609PUoLolGYkIGUlJyBWfyqJVX+x9n6rSh2L8/TefXS6n9VKs0Mnnl8tyuxN63SGm5AtJpv36w2tsp6WMyAgY9x/Py5cs4ffo0unXrZsgy9MLb6xsMGvAfjH/tf5j+7gD06duh7v8FzB2KmmoNorfr/mHOg1KpgGt3J2xY/yPcer6Pysp7+Pjj8YYuS680Gg169piBNi9MhJtbJ3Tp0tbQJRlU4ILxqK6uwdatPxu6FEIkzdT3LabeflNmsIFnZWUl/P39sXDhQlhaWgrKKiq8htatW9T97Ohgi6LC60JLFDW3uLgcAHDt2m3s/TEb3Xu0AQBMnOSOocO64l/vhAmuV2yXL1/D5ctXkZKSBwDYtfMoXLs7CcqUwrp6nPLyShw+nA0vr56Cs3jUyrv9ADBlyhB4e/fClMlfCcqRUvupVmlk8srVx3Yl1r5FSsv1Ycbefr1hTD+TETDIwLOqqgr+/v7w8fHB0KFDBeelpubByckBbdvaQ6VS4o3xr2LPnmNGk9uokRksLc3r/j1wUGecPl2MQUOc4TdnMN4cvxF371YJrldsV67cwOWCa+jY0REAMGiQC06rLwnKNPZ19TAbG6u6q7cbNDDDkCHdkZdXYJS18lquDwz16oGP5r6OMaOX4+7dPwRlSan9VKs0MqVWK499i5SWq5TaT8Sn9/t4MsawaNEitG/fHtOmTRMls6ZGg9n+a7EvfiUUCjk2h+6HWn3RaHJtWzRG+JZ3AQBKpRw7o9NwKOE0UrP+DXMzJXbGvQ+g9gKjuR9u16nWyC0L4eHRDTY2Vsi/GIXly8MQGrJfp6yHzZ79X4RHfAwzMyUuXCjB9Le/FpRn7OvqYS1bNkdI6HwoFHLI5TLERB/B3r3JRlmrmJkRkfPh4fEybGya4EJ+OFYsj8T8j9+AubkK+/d/BgBITs7D++9/b/BaeWZSrdLJlFqtPPYtUlquUmq/XjAAzDjOv9QHGRPjMt1nkJaWhjfffBMdO3aEXF57wDUgIAAeHh5PfE1ERCSmTJmqpwqFsbIQfp/LvyqvPC16Zi0ed5Kt4ZDJi4xDpnF8lVEfMpkZl1zG7nPJJUQ6THvfIqX2q9Un4ews/uf2sziddhidWYZe3ivXcrjB26v3I549e/ZEXl6evt+WEEIIIcQ46evJRUbguXlyESGEEEIIMW70rHZCCCGEEENhzGjusakPdMSTEEIIIYToBR3xJIQQQggxJDriSQghhBBCiLjoiKfIeNz6qKllV9EzAeDm7d+45PIg49BVmaRu/SQ+uu0RITxuKceLlPZXPK7Q5nGLJmIINPAkhBBCCDEUBrqdEiGEEEIIIWKjI56EEEIIIQbDTOqRmXTEkxBCCCGE6AUd8SSEEEIIMSQ6x1NaHB1tkZCwCjknNyE7Jxh+fmNEy/bycsMpdShy88Iwf/4Eo8rMyPk3jhybj5+T5iHhcEDd/Hfe64/jqQtw9MTHWLrCxyhqfYDnupo9ewyycjYgM3s9IrYEwtxcJUquXC5Hatr/EPfDJ6LkAcbdrx6Q0nYlpVp5ZfLK5ZEZFDwXRcXRyMoOEiXvAR61WllZYPuOxfjtVDBO/haE3r2dRck9ey4MmVnrkZa+DieS14qSKZV+xWv9m5urcPz4WqRnrEd2ThCWLp0iaj4RCdOze/fusddee435+PiwESNGsG+//Vbra8LDIxigeOJkb2/LXF07MUDBLC0tmfp0BHN2bvfU19RnkstVLO/MFtaunSNTqcxZekaw4FxdMptadnvsdP78Ttb+hb6PzBs5/DWWmLietWjenTW17Mac2vZ74usN0X5d15UM5k+dHFq1Yr+fjWINGzRmMpizqG3L2VtvjXzqawBlvaY5c8aziMglLO6HL+rx+9LpV7zWFdWq/0yp1dq/vwtzde3EsrJDBGcJq9VM6xQSupC9Pd2XAWZMpWrErKya1et12qaz56KYtbXNM7zm+elXuq3/+u2vLSwsGaBkSqU5O3ZsHXN3f+mpv69Wq/Uw8nk69YlEpklcoZfJGNqr9yOeZmZmCAsLww8//IDY2FgkJSUhKytLUGZJSRkyM88CAG7fvovc3EtwcLARXGuvXp1w7lwRLlwoRlVVNXZsPwxf335Gl/mwqdP74dtvEnH/fu09365du61zFo9aea0rAFAqFWjY0AwKhRyNGpmjqOi64EwHBxuMGOGOkE3xIlRYSyr9SkrblZRq5bUPkFKtSUknUVZWITjnYTxqbdKkEfr3fwkhm/YDAKqqqlFeXilGuaKTUr/isf4fqKy8BwBQqZRQqpRgjHF5H6I7vQ88ZTIZLCwsAADV1dWorq6GTCbejWHbtLGDi4sTkpNzBWe1crBBQUFp3c+XC6+ilYO10WQyMMTEzkDiLx9hytQ+AIAOHVqgT5/2OJD4IX7Y+wFcu7c2ilofR8x1VVR0Hd98HYPz+REoKNyKW+WVSPgpQ3Du6m9mIjAwCBoRH2dm7P3qcYx9u3qYsdfKq/1SqpUHHrW2a2ePa1fLsSnkI6Sm/RcbNs5Bo0bmQksFADAGxO//HMkp3+Odd4cLzpNSv+JJLpcjLX09ikuikZiQgZQU4fsB/ljtOZ76mIyAQc7xrKmpwahRo9C3b1/07dsX3bp1EyXXwqIBdkQvRUDAOlRU3BEl05h5e32HQQO+xvjXNuDtd15Bn77toVTK0bRZI3gN/gZLl/yA4M1TDV3mY4m9rpo2tYSPbx+82GEqXnB8E40sGmDSm4MEZXp7u6O09CYyMn4XXJ+USWm7klKtxPgplQq4dnfChvU/wq3n+6isvIePPx4vSrbHgAD0cvsAI70XYeZMX/Tvz+cJdaZGo9GgZ48ZaPPCRLi5dUKXLm0NXRL5C4MMPBUKBeLi4vDLL78gJycHZ86cEZypVCoQHbMMUVsTEbv7qAhVAkWF19C6dYu6nx0dbFFUKOzrWzEzS4rLAdR+nb7vx5Po3qMNiopuYu+eHABAZsYlaDQM1tYWBq/1YTzW1eAhrsjPv4Jr18pRXV2D2N2/ok8fYRcB9O3bBT4+fXD2XAS2bF2EgQNdEBb+seBajb1fPUwq2xUgnVp5tV9KtfLAo9bLl6/h8uWrSEnJAwDs2nkUrt2dBGU+8OBUoKtXyxEX+yvc3DoLy5NQv9KH8vJKHD6cDS+vnoYupX40Gv1MRsCgV7U3adIE7u7uSEpKEpwVFDwXp09fxJo1O0WorFZqah6cnBzQtq09VCol3hj/KvbsOWYUmY0amcHS0rzu368O6oTT6mLE7z2JV/q/CADo0MEWZioFrl/X7ZwkHu0H+Kyrgkul6OXeGQ0b1i6TQYNckHu6QFDmokUhaNtmEpw6TMabkz7Dzz9n4a0pXwqu1Zj71V9JZbsCpFMrr/ZLqVYeeNR65coNXC64ho4dHQHU7ldOqy8JrrVRI3NYWjas+7enZw+cOpUvKFNK/YoXGxsrWFnVHmhp0MAMQ4Z0R16esM8BIj6938ezrKwMSqUSTZo0wb1793Ds2DG8++67gjL79euKyZM9kZNzHmnp6wEASxaHID4+RVBuTY0Gs/3XYl/8SigUcmwO3Q+1+qJRZNq2aIywyLcBAEqlHDtjMnAoMRcqlQLf/Xciko5/jKqqanwwc6vBa30Yr3WVkpKHXTuTkJL2Paqra5CddQ5BQeJdECQmY+5XD5PSdiWlWnlkSq3WyC0L4eHRDTY2Vsi/GIXly8MQGrLfKGudPfu/CI/4GGZmSly4UILpb38tONPOrhlidi4FUHukflvUzzhwIE1QppT6FY/1DwAtWzZHSOh8KBRyyOUyxEQfwd69yYJzuWMwmvMv9UHG9HzJV25uLgIDA1FTUwPGGIYNG4YPPvjgqa+JiIjElClT9VOgEWpqyefcn5u3f+OSy4OMw99IDDWiZ9buQQgh0qAwdAHPgMf+SkrEuwj5YWp1Dpydxbk3q65On0hE55tH9PJeuW0mGLy9ej/i2blzZ8TGxur7bQkhhBBCjBCjZ7UTQgghhBAiNhp4EkIIIYQQvdD7V+2EEEIIIeQhJnRxER3xJIQQQgghekFHPAkhhBBCDMXEbqdERzwJIYQQQoheSOKIp0wmh1zeSNRMjeauqHl/Ev+vlpu3T4ueCQBV1QmiZ6qUQ0TPBACGatEzFYomomfW1NwSPZMQKTFX2XPJ/aOqhEMqn1vYKBVWomdW19wUPZMXudxS9EymuSd6pvFgRvM4S32gI56EEEIIIUQvJHHEkxBCCCHkuUTneBJCCCGEECI+OuJJCCGEEGJIdMSTEEIIIYQQcUl24LkxyB+XC8ORmbm2bt6SJRNxIT8UqWlrkJq2BsOG9RD8PnK5HKlp/0PcD58IznrAy8sNp9ShyM0Lw/z5E0TLPXsuDJlZ65GWvg4nktdqf8FDbt68jfFvrEDXLm/jpa5v4/hxNT6evxFdu7wNV9f38Ppry3Dz5m0AQFVVNaZN+w9cXN7FS13fxpcro57pvYKC56KoOBpZ2UHP9Dp9ZDo6WuPAweXIyv4WmVlr8MEH3gCAsa/1QWbWGty9F4Pu3TsIrlfsPuDoaIuEhFXIObkJ2TnB8PMbIzgT4NNXpVQrr1xTW67m5ir8krQcJ1I+Q2rGSixaMhYAsG79OziR8hmSUz9H5FZ/WFiY61wnj/3KA2J8DtTuW5YhM/sbZGR9g/c/GAEAWLpsAlLTv0Zy6lf4ce8StGzZTOf3MOZ968agD3C5cDMyM7+tm/fyy21xJGklMjK/xe7di9C4cUNB7zF79hhk5WxAZvZ6RGwJhLm5SlCeXmk0+pmMgMEGnjU1NRg9ejT+9a9/6fT68LBEjBy57G/zv/s2Dm4958Ct5xzs358usErA338McnMvCc55QC6X47u1fhjpvRAvdZ2O8RMGwtn5BdHyhwyej549ZqG3u98zve7DD9dhqFdP/HYqBOkZG+Ds/AKGDOmOrOwgZGZuxIsvOtQNMGNijuD+H1XIygpCcso6BAXtRX5+/W91Eh52AN4jFjxTffrKrK7W4OP5YXDpNhv9XwnEjJnD0dnZEepTlzD+jf8gKUkt+D149IHq6hrMm7ceL780Hf36+mHmrFGCM3n1VSnVyiPXFJfrH39UYcSwz9G71yL06bUInp4vw61XB3w8bwt691oEd7eFuFxwHTNmDtW5Vh77lQfE+Byorq7Bx/PD4NrtQwx4ZQFmzByGzs6OWP11HNx6fAR3t3nYty8dCxeN0/k9jHnfGh52CCNHrnhk3voN72PRwgh0d52N2LgT+Ogj3f9YatXKGu/7jULvXn5w7TYDCoUc4ye8KrBqwoPBBp7h4eHo0EH3I0dHj57CjbLbIlb0dw4ONhgxwh0hm+JFy+zVqxPOnSvChQvFqKqqxo7th+Hr20+0fF2Ul1fiaNJJvP32cACAmZkKTZtawnNoTyiVCgCAe29nXC68BgCQyYDKynuorq7B3bv3YWamRJMm9b/PalLSSZSVVYjaBrEyS0puICvrPADg9u17yM29DIdW1sjNLcSZM0WC8wE+faCkpAyZmWcBALdv30Vu7iU4ONgYXZ1Sq5VHrqku18rKPwAAKpUCKpUSjAEVFX/eT7lBQxUY0/08Nx77FUC8z4GSkpvIyroA4MG+pRAOrZo/sgwsLMyNbhmIlXn0qPpvn9kvvtgKSUmnAACJCdkYM6aPoPdQKhVo2NAMCoUcjRqZo6jouqA8vWFMf5MRMMjAs6SkBIcPH8brr78uevbMWd5Iz/gOG4P80bSphaCs1d/MRGBgEDQiHp5u5WCDgoLSup8vF15FKwdrUbIZA+L3f47klO/xzrvD6/26CxeKYWNjhenTv0LPnjPw3ntfo7Ly0Rvsbw49gGHD3AAAr702ABYWDdDacTzat3sTHwaMQ/Pm4t+M3dDatLFFt27tkJJyRtRcnn0AANq0sYOLixOSk3MF5fCuEzD+WnnkmupylctlOJ78GfIL1uFQ4kmkpZ4DAKzf+B4uXPwvOnZqhf+tOyioXh54fA60aWMLl25tkZLyOwBg+YqJOHtuPSZM7I8Vy7eL9j7GTq0ugK+vOwDgtdf7wrG17n8oFRVdxzdfx+B8fgQKCrfiVnklEn7KEKtUIiKDDDw///xzzJs3D3K5uG+/YUM8Onf6F3r2mI2S4jL856vpOmd5e7ujtPQmMjJ+F7FCvjwGBKCX2wcY6b0IM2f6on//rvV6XXV1DTIzf8e//uWDtLT1sLBogP98+efO74vPt0CpVGDSpMEAgJSUXMgVclwq2Ibfz4ZjzTcxOH++mEubDMXCogG2bZ+PuXNDHjkiYewsLBpgR/RSBASsQ0XFHUOX81RSqlVKjHW5ajQMfdwXoWMHf/Rw64B//MMRADDjvY3o0O4D5OUW4fVxvQ1c5aN4fA5YWDRA1Pa5mDt3c92+Zem/o+DUYQa2RSVh5qxhor2XsXvv3bX414zhOJH8NRpbNsT9+1U6ZzVtagkf3z54scNUvOD4JhpZNMCkNweJWC1nGqafyQjofeD5888/o3nz5ujatX6DomdRWnoTGo0GjDFs2nQQbj1f1Dmrb98u8PHpg7PnIrBl6yIMHOiCsPCPBddYVHgNrVu3qPvZ0cEWRYXifB3w4GuFq1fLERf7K9zcOtfrdY6OtnB0tIW7uzMA4LWxA5CZWbujDQs7gL17kxEeEQiZTAYA2LbtELy8ekKlUqJFi2bo07cL0tPFPSpoSEqlAtu3z8O2qCOIi00WPZ9XH1AqFYiOWYaorYmI3X1UcB7PviqVWnnkmvpyLS+/gyO/qOE59OW6eRoNQ0z0cYwa7SYoW2xifw4olQps2z4X26KSHrtv2RaVhNFjjGvwzVNeXiG8RyxDb/ePsH17Es6f1/2xqIOHuCI//wquXStHdXUNYnf/ij59nEWslohF7wPPjIwMHDp0CIMGDUJAQABOnDiBuXPnipJtb//n1YCjRvfGqVMXdc5atCgEbdtMglOHyXhz0mf4+ecsvDXlS8E1pqbmwcnJAW3b2kOlUuKN8a9iz55jgnMbNTKHpWXDun97evbAqVP59XqtvX1zODraIi+vAABw6FAmnJ3b4MD+VHy9agd2x65Ao0YN6n7/hdYt8PPPWQCAysq7SEk+jU6dWgtug7HYsPF95OYW4ttv93DJ59UHgoLn4vTpi1izZqcIVfKrE5BOrTxyTXG52tg0hpVV7XngDRqoMGjwSzhzphjt29vV/Y63d3ecyRPnPGqxiP05sGHjLOTmXsZ33/5YN6+D05/Pth/p44a8vEJBNUuJrW3tM+1lMhkWLByHjRsP6JxVcKkUvdw7o2HD2jsjDBrkgtzTBaLUScSl9xvIf/TRR/joo48AAMnJyQgJCcGqVaueOSciYi4GeHSFjU0TnL8QghUrouDh0RXdurUDY8DF/CuYNWud2OULVlOjwWz/tdgXvxIKhRybQ/dDrdZ9gPyAnV0zxOxcCuD//6qO+hkHDqTV+/Vrvn0fU6Z8gfv3q9G+XUsEb5qLPr0/wB9/VGHYsNq/8N3dnbFu3RzMnDUK70z/Ct1efgeMMbz1lhdefrl9vd8rcstCeHh0g42NFfIvRmH58jCEhux/tgZzyuzbtzP++c9XcfJkPlJSvwYA/HvJFpiZq/DNN+/A1rYJYuMWISf7AkaO1O3WKjz6QL9+XTF5sidycs4jLX09AGDJ4hDEx6fonMmrr0qpVh65prhc7e2bYmPwv6BQyCGXy7BzZzL2x2fhp0NL0KRxQ8hkwMmTlzDbb7POtfLYr4ipb9/OePOfHjh58iKSU78CAPx7yVZMnTYYHTu2gkbDcOnSVfi9v1Hn9zDmfWtERMBDn9nBWLFiGywtG2DmjNrrEWJjTyBsc6LOdaak5GHXziSkpH2P6uoaZGedQ1CQeBcGc2ckX4Prg4wJuYROoAcDzw0bNjz19yIjt+Ktt2aI+t4aDa/z9ngsTgWHTKCqWve/Lp9EpRwieiYvCoX4F0TV1NwSPZMQKTFX2Wv/JR38UaX717BPJuOQCSgVVqJnVtfcFD2TF7ncUvRMprkneiYAnFJnwtnZsF/Jn/71IDpf1M8fSbmu7xq8vQZ9ZKa7uzvc3d0NWQIhhBBCiOEwGM3N3fVBsk8uIoQQQggh0mLQI56EEEIIIabNeG51BACDBg2ChYUF5HI5FAoFdu3ahZs3b+LDDz9EYWEhHBwcsGbNGlhZWYExhs8++wy//PILGjRogJUrV6JLly5PzacjnoQQQgghpE5YWBji4uKwa9cuAMDGjRvRp08fHDx4EH369MHGjbUXwR05cgT5+fk4ePAgPvnkEyxbtkxrNg08CSGEEEIMhcHobyCfmJiI0aNHAwBGjx6NhISER+bLZDK4uLjg1q1bKC0tfUoSDTwJIYQQQshDpk+fjrFjx2L79tqnGF6/fh0tWtQ+TMLW1hbXr9c+SOLKlSuwt//zThb29va4cuXKU7MlcY4nYxpoNH+InSpynvTwufURn1s/AeJf8VdTUyl6JiGmjs9tjwAe+xalUvzbHgFAdfUNLrlSodHcNnQJ0qOnczzLysowduzYup/Hjx+P8ePHP/I7UVFRsLOzw/Xr1zFt2jS0b//ofbplMlndkwx1IYmBJyGEEEIIEaZ58+Z1520+iZ1d7RPFrK2t4enpiZycHFhbW6O0tBQtWrRAaWkpmjdvXve7JSV//rFZUlJS9/onoa/aCSGEEEIMhAFgGqaXSZs7d+7g9u3bdf/+9ddf8eKLL2LQoEGIjY0FAMTGxmLw4MEAUDefMYasrCw0bty47iv5J6EjnoQQQgghBNevX8f7778PAKipqcHIkSMxYMAAvPTSS5gzZw5iYmLQqlUrrFmzBgDg4eGBX375BZ6enmjYsCE+//xzre9BA09CCCGEEENhAAz39PJHtG7dGj/88MPf5jdr1gxhYWF/my+TybB06dJneg/6qp0QQgghhOjFczPwtLKywPYdi/HbqWCc/C0IvXs7C84MCp6LouJoZGUHiVDhn7y83HBKHYrcvDDMnz9BtFwey4BHrTzqfEAulyM17X+I++ETUfLOngtDZtZ6pKWvw4nktaJkAnyWK49MHtsAr+1KStsrr32A2LlSWqYdOzoiLX1d3VR2Yxf8/cc8c46jozUOHPg3MrNWIyPza7z/wXAAwOLF43Du/Hokp/wHySn/gdcwV8E1i72/kso+wNHRFgkJq5BzchOyc4Lh5/fs6+lxePVX/vR0D08jeTqSQQaegwYNgo+PD0aNGvXIZf1CfLNmJg4cSEPXLu+gu+tMnD59SXBmeNgBeI9YIEJ1f5LL5fhurR9Gei/ES12nY/yEgXB2fkGUbLGXAa9aeayrB/z9xyA3V7w8ABgyeD569piF3u5+ouTxWK681hWPbYBHJq9cKa0rHrlSWaYAcObMZfTsMQs9e8xCL7cPcOfOH4iN/fWZc6qra/DxxxFwdQnAgP6LMGOGFzp3dgAArF27F+695sO913wc2J8puGYx91dS2gdUV9dg3rz1ePml6ejX1w8zZ40y2lqJ+Ax2xPOvj2MSokmTRujf/yWEbNoPAKiqqkZ5ufB7NCYlnURZWYXgnIf16tUJ584V4cKFYlRVVWPH9sPw9e0nOJfHMuBRK691BQAODjYYMcIdIZviRcnjhcdy5dWveGwDPDJ55UppXfHIlcoy/avBg11w/lwxLl16+hNUHqek5Caysi4AAG7fvofc3EI4ODQXtT5A/P2VlPYBJSVlyMw8CwC4ffsucnMvwcHBRnAur30LEddz8VV7u3b2uHa1HJtCPkJq2n+xYeMcNGpkbuiyHquVgw0KCv7cGV4uvIpWDtaCc3ksAx618lxXq7+ZicDAIGg04t1snjEgfv/nSE75Hu+8O1yUTB7LlVe/MnVSWldS6QP6qPON8a9i27bDgnPatLGFS7d2SEmpHSTNnOGF1LSvsGHDTDRtaiEoW+z9lVTW/1+1aWMHFxcnJCfnGroUw6Kv2vn76+OYhFAqFXDt7oQN63+EW8/3UVl5Dx9/PF77C58jUlkGvOr09nZHaelNZGT8LkKVf/IYEIBebh9gpPcizJzpi/79u4qaTwgRl0qlhI9Pb8TEHBGUY2FhjqhtH2Hu3M2oqLiLjRsPwtnZD73c5qOk5Aa+/HKKztm89ldSY2HRADuilyIgYB0qKu4YuhyiJwYZeEZFRWH37t0ICgrCli1bkJqaKijv8uVruHz5KlJS8gAAu3YehWt3JzFKFV1R4TW0bv3nzVUdHWxRVHhdcC6PZcCjVl7rqm/fLvDx6YOz5yKwZesiDBzogrDwjwXnFhXVtvfq1XLExf4KN7fOwjM5LFde/crUSWldSaUP8K5z2HA3ZGaeRWnpTZ0zlEoFtm3/CNu2JSEuLgUAUFpaDo2GgTGGkJBE9HTroHM+j/2VVNb/A0qlAtExyxC1NRGxu48auhzDYqAjnrw97nFMQly5cgOXC66hY0dHAMCgQS44rRb3AhOxpKbmwcnJAW3b2kOlUuKN8a9iz55jgnN5LAMetfJaV4sWhaBtm0lw6jAZb076DD//nIW3pnwpKLNRI3NYWjas+7enZw+cOpUvuFYey5VXvzJ1UlpXUukDvOucMEH41+wbNsxAbm4hvvt2b908e/umdf/2HdULp04V6JzPY38llfX/QFDwXJw+fRFr1uw0dClEz/R+A/k7d+5Ao9HA0tKy7nFMs2bNEpw7e/Z/ER7xMczMlLhwoQTT3/5acGbkloXw8OgGGxsr5F+MwvLlYQgN2S8os6ZGg9n+a7EvfiUUCjk2h+6HWn1RcK2A+MuAV6081hUPdnbNELOz9sa4SqUC26J+xoEDaYJzeSxXXuuKxzbAI5NXrpTWFY9cqSzTBxo1MseQId0xc8a3Omf07dsJb/7TAydPXkRyyn8AAP/+dxTGv9EPL3drC8YYLl68ig/e3yhKzWKR0j6gX7+umDzZEzk555GWvh4AsGRxCOLjU4yuVr0xkqOR+iBjTL+3yy8oKPjb45hmzpz51NdERERiypTpIldSI3IeTwpOuTyWAa9axbtg6E88DvhLqV8RIiXi71uUSivRMwGguvoGh1TTGZjok1p9Es7O4t1LWqcaftmPzjm79fJeeUPmGLy9ej/i+aTHMRFCCCGEmBwGMBM64vlc3E6JEEIIIYQYP70f8SSEEEIIIQ8YzxXn+kBHPAkhhBBCiF7QEU9CCCGEEEOiI56EEEIIIYSIi454EkIIIYQYyoMnF5kICQ08edzHURpkMj73xmSMxz0n+dzHUiZrIHomY/dEzySEyDjlir9vqa4uFz2TEPJ0Ehp4EkIIIYQ8h/T7LB+DonM8CSGEEEKIXtDAkxBCCCGE6AV91U4IIYQQYkDMhC5jea6OeMrlcqSm/Q9xP3wiWqaXlxtOqUORmxeG+fMnGFVmUNAcFBZtRWbWurp5K798Gyd/24CMjP8iOmYxrKwsdM8Pnoui4mhkZQfpnPFXjo62SEhYhZyTm5CdEww/vzG61xc0G4VFkcjM+m/dvGXL/4mMjLVIS/sO++JXoGXL5kZR68N4LFfAuPvqw3gtVx618sqlWsXfX/PYrjp2dERa+rq6qezGLvj7i9NfAfGXgVT6lZT2gUR8Bhl43rp1C/7+/hg2bBiGDx+OzMxMUXL9/ccgN/eSKFlA7U7hu7V+GOm9EC91nY7xEwbC2fkFo8kMC0/ASO8lj8xLSMiES7eZ6N79ffz+eyE+DnxD51rDww7Ae8QCnV//ONXVNZg3bz1efmk6+vX1w8xZowS2f+kj875etRPdu/uhZ09/7NubisWLJxpFrQ/jsVyNva8+jMdy5VWrlJarlGoFxN9f89iuzpy5jJ49ZqFnj1no5fYB7tz5A7Gxv4qWL+YykFK/kso+UG8e3E5JH5MRMMjA87PPPkP//v2xf/9+xMXFoUOHDoIzHRxsMGKEO0I2xYtQYa1evTrh3LkiXLhQjKqqauzYfhi+vv2MJvNo0m8oK6t4ZF7CT5moqak9Zp98IheODjY615qUdPJv+UKVlJQhM/MsAOD27bvIzb0EBx1rPJp06m/1VVTcrft3I4sGYAKuFBSz1ofxWK7G3lcfxmO58qpVSstVSrXy2F/z2K4eNniwC86fK8alS6Wi5Im9DKTUr6SyDyR86H3gWVFRgdTUVLz++usAADMzMzRp0kRw7upvZiIwMAgajXgnSrRysEFBwZ87mcuFV9HKwdroMp9k6rSh2L8/jUu2GNq0sYOLixOSk3NFzV3xyWScvxCKiRNfxbJlkaJk8qpVLFLtq2ItV161Smm5SqlWHvtr3t4Y/yq2bTssWp7Yy0BK/YoHqdT5eHo62mmqRzwvX76M5s2bY8GCBRg9ejQWLVqEO3fuCMr09nZHaelNZGT8LlKV0he4YDyqq2uwdevPhi7lsSwsGmBH9FIEBKxDRYWw9f9X/14SgfbtpiEq6jBmvT9ScB7PWk0ZLVfTJMX9tUqlhI9Pb8TEHBElT4rLgBCx6H3gWV1dDbVajYkTJyI2NhYNGzbExo0bBWX27dsFPj59cPZcBLZsXYSBA10QFv6x4FqLCq+hdesWdT87OtiiqPC60WX+1ZQpQ+Dt3QtTJn8laq5YlEoFomOWIWprImJ3H+X2PlFbD2PMGGFfteirVqGk1lfFXq68apXScpVKrbz21zwNG+6GzMyzKC29KUoej2UgpX7Fg1TqfCxWe1W7PiZjoPeBp729Pezt7dGtWzcAwLBhw6BWqwVlLloUgrZtJsGpw2S8Oekz/PxzFt6a8qXgWlNT8+Dk5IC2be2hUinxxvhXsWfPMaPLfNhQrx74aO7rGDN6Oe7e/UO0XDEFBc/F6dMXsWbNTtGznZxa1f3b19cdeXmXBeXxrFVMUuurYi9XXrVKablKpVZe+2ueJkwQ92t2HstASv2KB6nUSQxwH09bW1vY29vj/PnzaN++PY4fPy7KxUU81NRoMNt/LfbFr4RCIcfm0P1Qqy8aTWZE5Hx4eLwMG5smuJAfjhXLIzH/4zdgbq7C/v2fAQCSk/Pw/vvf65QfuWUhPDy6wcbGCvkXo7B8eRhCQ/brlPVAv35dMXmyJ3JyziMtfT0AYMniEMTHpzxzVkTkPHh4vPT/7d+MFcu3YNjwnujY0RFMo8HFS1fx/qz/ag/SQ60P47Fcjb2vPozHcuVVq5SWq5Rq5YHHdgUAjRqZY8iQ7pg541sRquRHSv1KKvtAvTKS8y/1QcaEXParo9OnT2PRokWoqqpC69at8cUXX8DKyuqJvx8REYkpU6aJXIV0VrJMZsYll7H7XHJ5kMkaiJ7J2D3RMwkhMk65PPbZCg6ZAMDjO03pfGZJiVp9Es7OzoatIXEfXkzarpf3Ojs+0ODtNciTi5ydnbFr1y5DvDUhhBBCiHExkvMv9eG5enIRIYQQQggxXvSsdkIIIYQQQ2EAM6FzPOmIJyGEEEII0Qs64kkIIYQQYkh0jichhBBCCCHiooEnIYQQQgjRC0l81S6TySGXNxQ1U6OhZ0PL5ZaiZ2o0laJnAgBj4j+Fie4NSggP0rlIQi4355Kr0dzlkisVPO49LaX7Tj8zBiltNoLREU9CCCGEEKIXkjjiSQghhBDyPGKg2ykRQgghhBAiOjriSQghhBBiSHQ7JUIIIYQQQsQl2YHnxiB/XC4MR2bm2rp5S5ZMxIX8UKSmrUFq2hoMG9ZD0HsEBc9FUXE0srKDhJb7CC8vN5xShyI3Lwzz50/QOScoaA4Ki7YiM2td3bxmzSwRv/8zqE8HIX7/Z2ja9NmuXN8Y9AEuF25GZua3dfO6dWuHpKNfIjXtGxw/sQo93V7UueYH5HI5UtP+h7gfPhGcBQDm5iocP74W6RnrkZ0ThKVLp+iUExQ0G4VFkcjM+u/f/t+cD8egqvpHWFs3EVSrsfcr3pm8cqlW8XN59VUeuY6OtkhIWIWck5uQnRMMP78xOmc97vNly5Z5dZ8tZ34PQmraGsE1i70fNOZ+xePzikedBsEAptHPZAz0PvA8f/48Ro0aVTd1794dmzdvfuac8LBEjBy57G/zv/s2Dm4958Ct5xzs358uqNbwsAPwHrFAUMZfyeVyfLfWDyO9F+KlrtMxfsJAODu/oFNWWHgCRnoveWTe/I/fwKFDWfiH87s4dCgL8z8e90yZ4WGHMHLkikfmff7FW/j0k+1w6/khli+LwhdfvKVTvQ/z9x+D3NxLgnMe+OOPKgwZMg89us9Aj+4z4OXVE+7uzs+cU7tMl/5tvqOjDTw9XXHxYqngWo29X/HMpFqlVSuPvsort7q6BvPmrcfLL01Hv75+mDlrlM7tf9zny5tvflX32bJ793HE7j4uuGYx94PG3q94fF7xqJPwp/eBZ/v27REXF4e4uDjs2rULDRs2hKen5zPnHD16CjfKbnOo8E9JSSdRVlYhamavXp1w7lwRLlwoRlVVNXZsPwxf3346ZR1N+u1v9fn49EZEeAIAICI8Ab6+fZ4t86j6b8uVMYYmTWrvo2pl1QjFRWU61fuAg4MNRoxwR8imeEE5f1VZWXsPTZVKCaVKCcae/SrBo0mnHrvOV339LhYEhuqU+VfG3q94ZlKt0qqVR1/llVtSUobMzLMAgNu37yI39xIcHGx0ytL2+fL66/2wffsRnbIfEHs/aOz9isfnFY86DUajp8kIGPSr9uPHj6N169ZwcHAQLXPmLG+kZ3yHjUH+aNrUQrRcsbRysEFBwZ9HzS4XXkUrB2vR8u3smqKk5AYAoKTkBuzsmgrOnPvRJnyxcirOnQ/Gyi+nYvHiCEF5q7+ZicDAIGg04m4FcrkcaenrUVwSjcSEDKSk5IqS6+PjjqLC68jJuSBKHg88+hWvvkq1SqdWqWrTxg4uLk5IThZnH/CwV17pgtLSmzh7tlhQjtj7QSn2KzE/r6j/S4dBB5579+7FyJEjRcvbsCEenTv9Cz17zEZJcRn+89V00bKlSowjdO/9axjmzQ1Bh/bvYN7cEGzY+IHOWd7e7igtvYmMjN8F1/VXGo0GPXvMQJsXJsLNrRO6dGkrOLNhQ3MELngDy5ZFCi+QEMKdhUUD7IheioCAdaioEP8JdeMnDMD2bUmCMnjuB6VMjM8rSaJzPPXj/v37OHToEIYNGyZaZmnpTWg0GjDGsGnTQbj1FH4RjNiKCq+hdesWdT87OtiiqPC6aPlXrtyEvX0zAIC9fTOUlpYLzpw8eSB2///5TDExv8JNwMVFfft2gY9PH5w9F4EtWxdh4EAXhIV/LLjGh5WXV+Lw4Wx4efUUnNWhgz3atrVDesZa/H52ExwdbZCSukaUI8li4tGvePVVqlU6tUqNUqlAdMwyRG1NROzuo6LnKxRyjB7dB9HRwgaePPaDUuxXYn5eUf+XDoMNPI8cOYIuXbrAxka3c3Ae50EHBoBRo3vj1KmLomWLJTU1D05ODmjb1h4qlRJvjH8Ve/YcEy3/xx9PYPKUIQCAyVOGYM+eE4Izi4vKMGBAVwDAwIEvC/qKadGiELRtMwlOHSbjzUmf4eefs/DWlC8F12hjYwUrq9pTKxo0MMOQId2Rl1cgOPe33y7CodU/8aLTdLzoNB2XL19DL7c5uHLlpuBsMfHoV7z6KtUqnVqlJih4Lk6fvog1a3ZyyR882AV5eZdRKHBAw2M/KMV+JebnleT7vwmd42mwG8jv3bsX3t7eOr8+ImIuBnh0hY1NE5y/EIIVK6Lg4dEV3bq1A2PAxfwrmDVrnfagp4jcshAeHt1gY2OF/ItRWL48DKEh+wVl1tRoMNt/LfbFr4RCIcfm0P1Qq3UbIEdEzoeHx8uwsWmCC/nhWLE8Ev/5MhpR2xZg2rShuHSpFBMnfPFsmREBDy3XYKxYsQ0zZq7D6tXvQKmU4969KsycKWy58tCyZXOEhM6HQiGHXC5DTPQR7N2b/Mw5EZHz4OHx0v8v081YsXwLQkN/ErVWY+9XPDOpVmnVyqOv8srt168rJk/2RE7OeaSlrwcALFkcgvj4lGfOetzny+bQn/DG+P6CLyrixdj7FY/PKx51Ev5kzAAnVdy5cwcDBw5EQkICGjdurPX3IyO34q23Zohag0Yj/rk/vMhkZpLJ1WgqRc/kRSYzFz2TsXuiZxJC+JDLG3HJ1WjuckiVzvmPPD5bGLsveiYAqNUn4ez87LfeE9OpA/vQbk+UXt7r4vsLDd5egxzxbNSoEZKTn/1oFCGEEEIIkS7JPrmIEEIIIYRIi8HO8SSEEEIIMXnMeG51pA90xJMQQgghhOgFHfEkhBBCCDEkOuJJCCGEEEKIuCRxxJMxjaRufyQ2xmo45d7mkisVdOsjQkybKX+u8MTr1kfPKwY6x5MQQgghhBDRSeKIJyGEEELIc4kB+n+Uj+HQEU9CCCGEEKIXdMSTEEIIIcSQNDJDV6A3dMSTEEIIIYToxXMx8HR0tEVCwirknNyE7Jxg+PmNES3by8sNp9ShyM0Lw/z5E4w2EwDOngtDZtZ6pKWvw4nktaJkSqn9PHKDgueiqDgaWdlBouTxygSk036AajX19ktpu+KRy+szSyrrCpBWrfrANPqZjAIzgNDQUDZixAjm7e3NPvzwQ3bv3r2n/n54eAQDFE+c7O1tmatrJwYomKWlJVOfjmDOzu2e+pr6THK5iuWd2cLatXNkKpU5S88IFpyrW6ZZvaaz56KYtbVNvX9fOu03XG7//i7M1bUTy8oOEZzFM1NK7Tf1Wk29/bxq5ZHJK5fHZ5aU1pUx1apWq/U08nmyk/v2svKpk/UyGUN79X7E88qVKwgPD8fOnTvx448/oqamBnv37hWUWVJShszMswCA27fvIjf3EhwcbATX2qtXJ5w7V4QLF4pRVVWNHdsPw9e3n9Fl8iKl9vPKTUo6ibKyCsE5vDOl1H5Tr9XU2w9IZ7vilcvjM0tK60pKteoFAxiT6WUyBgb5qr2mpgb37t1DdXU17t27hxYtWoiW3aaNHVxcnJCcnCs4q5WDDQoKSut+vlx4Fa0crI0u8wHGgPj9nyM55Xu88+5wwXlSaj/P5SoFUmq/qddq6u0njxLrM0tK60pKtRLx6f2qdjs7O7z99tsYOHAgzM3N0a9fP7zyyiuiZFtYNMCO6KUICFiHigrTeyKFx4AAFBVdh62tFfYfWIm83AIkJf1m6LIIIYQ8hql/ZpE/Gc35l3qg9yOe5eXlSExMRGJiIpKSknD37l3ExcUJzlUqFYiOWYaorYmI3X1UhEqBosJraN36z6Oxjg62KCq8bnSZddlFtTlXr5YjLvZXuLl1FpYnofbzXK5SIKX2m3qtpt5+UkvszywprSsp1UrEp/eB57Fjx+Do6IjmzZtDpVJh6NChyMzMFJwbFDwXp09fxJo1O0WoslZqah6cnBzQtq09VCol3hj/KvbsOWZ0mQDQqJE5LC0b1v3b07MHTp3KN7paebWfV65USKn9pl6rqbef1BL7M0tK60pKtRLx6X3g2apVK2RnZ+Pu3btgjOH48ePo0KGDoMx+/bpi8mRPDBzoirT09UhLX4/hw3sJrrWmRoPZ/muxL34lfjsVgpjoX6BWXzS6TACws2uGX46sRnrG/3D8xFrs25eCAwfSjK5WXu3nlRu5ZSGO/vodOnVqjfyLUZj29jCjzJRS+029VlNvP69aeWTyyuXxmSWldSWlWvWBwbRupyRjTP9PCP3uu++wb98+KJVKODs747PPPoOZmdkTfz8iIhJTpkzVX4FGR8Ept4ZTLiGEEGL81OqTcHZ2NmgNJ/ftg+PW7Xp5r5JFgQZvr0Eemenv7w9/f39DvDUhhBBCiPH4/9spmYrn4slFhBBCCCHE+BnkiCchhBBCCAEAGZiGjngSQgghhBAiKhp4EkIIIYQYEGP6meqrpqYGo0ePxr/+9S8AQEFBAcaNGwdPT0/MmTMH9+/fBwDcv38fc+bMgaenJ8aNG4fLly9rzaaBJyGEEEIIqRMeHv7IrS5XrVqFqVOn4qeffkKTJk0QExMDAIiOjkaTJk3w008/YerUqVi1apXWbBp4EkIIIYQYEGMyvUz1UVJSgsOHD+P111///9oYTpw4AS8vLwDAmDFjkJiYCAA4dOgQxowZAwDw8vLC8ePHoe0unRK5uEgGmezJ9/nUBWP3Rc3ji89dX+XyRqJnajT0vGFCiKnjcaGI3m+5bWRM5+IbQ/v8888xb948VFZWAgBu3LiBJk2aQKmsHTLa29vjypUrAIArV66gZcuWAAClUonGjRvjxo0baN68+RPzJTLwJIQQQgh5/jAGvV3VXlZWhrFjx9b9PH78eIwfP77u559//hnNmzdH165dkZyczKUGGngSQgghhJiA5s2bY9euXU/8/xkZGTh06BCOHDmCP/74A7dv38Znn32GW7duobq6GkqlEiUlJbCzswMA2NnZobi4GPb29qiurkZFRQWaNWv21BroHE9CCCGEEAMylqvaP/roIxw5cgSHDh3C6tWr0bt3b3z99ddwd3fHgQMHAAC7d+/GoEGDAACDBg3C7t27AQAHDhxA7969IZM9/egtDTwJIYQQQsgTzZs3D6GhofD09MTNmzcxbtw4AMDrr7+OmzdvwtPTE6GhoZg7d67WLPqqnRBCCCHEgIzxWe3u7u5wd3cHALRu3bruFkoPMzc3x3ffffdMuZI94hkUNAeFRVuRmbWubt5rr72CrOz/4Y/7P6JHjxdFeR8vLzecUociNy8M8+dPMNpMc3MVjh9fi/SM9cjOCcLSpVN0ytkY5I/LheHIzFxbN2/Jkom4kB+K1LQ1SE1bg2HDegiqlUf7g4Lnoqg4GlnZQaLkPSCVWh0dbZGQsAo5JzchOycYfn5jRMnl0X5euaZeq1S2AV59FZBWrQAgl8uRmvY/xP3wiSh5UtkGePVVsT4HCWfMADZv3sy8vb3ZiBEjWGhoqNbfDw+PZDJZw0em/gPcWPfuXVlWdmjdPGfnjqxTJyeWeOhb1rPny397zcMToNA6yeUqlndmC2vXzpGpVOYsPSOYOTu3q9drxc1U1muysLBkgJIplebs2LF1zN39paf+vlze+G/TgAG9Wfce3Vh29ua6eUuXvsc++mjyY3//r5MhlimgYP37uzBX104sKztEcJYUa7W3t2Wurp0YoGCWlpZMfTrCQH1VStvV81WrVLYBHn3V+Gqt3z57zpzxLCJyCYv74Yt6/P7zsw3o1lf5fA6q1Wr+gx4tsvfsY8Vjp+tlMob2PvWIZ01NDSZPnizqQPfMmTOIjo5GdHQ04uLicPjwYVy8ePGZc44m/YaysopH5uXmFuDMmUKxSkWvXp1w7lwRLlwoRlVVNXZsPwxf335Gl/lAZeU9AIBKpYRSpdR6E9fHOXr0FG6U3Ralnsfh1f6kpJN/6w9CSanWkpIyZGaeBQDcvn0XubmX4OBgIyiTV/ultF1JqVapbAM8+qrUagUABwcbjBjhjpBN8aLkSWkb4NFXHxDjc1DvGKDRyPQyGYOnDjwVCgXkcjkqKsTrIOfOncPLL7+Mhg0bQqlUws3NDQcPHhQtX0ytHGxQUFBa9/Plwqto5WBtdJkPyOVypKWvR3FJNBITMpCSkitKLgDMnOWN9IzvsDHIH02bWuicw7P9YpNSrQ9r08YOLi5OSE4Wtv55tV9K25WUauWBd61i9VVAWrUCwOpvZiIwMAgajTgPCJHSNsATz89BIg6t53g2atQIPj4+WLhwIT799NO6SVcdO3ZEeno6bty4gbt37+LIkSMoKSnROY/8SaPRoGePGWjzwkS4uXVCly5tRcndsCEenTv9Cz17zEZJcRn+89V0UXKJ+CwsGmBH9FIEBKxDRQU9RYoYLyn1VbFr9fZ2R2npTWRk/C5CdeRhvD4HeWIwntsp6YPWq9qHDh2KoUOHivaGHTp0wDvvvIPp06ejYcOG6Ny5M+Ry47zGqajwGlq3blH3s6ODLYoKrxtd5l+Vl1fi8OFseHn1xKlT+YLzSktv1v1706aDiI1donOWPtovFinVCgBKpQLRMcsQtTURsbuPCs7j1X4pbVdSqpUHXrWK3VcBadXat28X+Pj0wfDhvdCggRmaNGmEsPCP8daUL3XOlNI2oA9ifw4S8Wgd8Y0ZMwbDhw9Ht27dMGbMmLpJiHHjxmHXrl3YsmULrKys0LZtW0F5vKSm5sHJyQFt29pDpVLijfGvYs+eY0aXCQA2Nlawsqr9CrxBAzMMGdIdeXkFgnMBwN7+z6cQjBrdG6dOPfs5uQ/waj8PUqoVqL1S9PTpi1izZqcoebzaL6XtSkq18sCrVrH7KiCtWhctCkHbNpPg1GEy3pz0GX7+OUvQoBOQ1jbAC8/PQb5kYEw/kzHQesTz0KFD+PLLL1FVVYVDhw7h9OnT+Pbbb7F+/Xqd3/T69euwtrZGUVERDh48iB07djxzRkTkfHh4vAwbmya4kB+OFcsjUVZWgTXfzoStrRXifliG7Ozz8B6h+9G5mhoNZvuvxb74lVAo5Ngcuh9qte6DLl6ZANCyZXOEhM6HQiGHXC5DTPQR7N377M9ZjYiYiwEeXWFj0wTnL4RgxYooeHh0Rbdu7cAYcDH/CmbNWqc96Al4tT9yy0J4eHSDjY0V8i9GYfnyMISG7BeUKaVa+/XrismTPZGTcx5p6bXb5pLFIYiPT9E5k1f7pbRdSalWqWwDPPqq1GrlQUrbAI++Coj3OUj4kjEtl3yNHTsWYWFhmDx5MmJjYwEAI0eOxI8//qjzm06aNAk3b96EUqnEggUL0KdPn6f+fkTEFrz11rs6v9/jMHZf1Dy++PyVIpc3FD1TozHuc7UIIYQ/HvtsIzlBz2D4fA6q1Tlwdnbmkl1f2Xv2o3nQbr281+0v5xi8vVqPeCqVSjRu3PiRedqew6nN1q1bBb2eEEIIIYRIj9aBp5OTE/bs2YOamhrk5+cjIiICrq6u+qiNEEIIIeS5pzGS8y/1QevFRUuWLMHZs2dhZmaGgIAAWFpaYtGiRfqojRBCCCGEPEe0HvFs2LAhPvzwQ3z44Yf6qIcQQgghxGQwBjAjeaqQPmgdeJ48eRIbNmxAYWEhqqur6+bv2bOHa2GEEEIIIeT5onXgOXfuXMyfPx8dO3Y02hu9E0IIIYRIlbE8VUgftA48mzdvjsGDB+ujlqdgYKzKwDXUj0zWQPRMxu6JnklIiybuXHJLb/G4b56CQyYA1HDKJabNhEYRekPL9HmhdeDp7++PRYsWoU+fPjAzM6ubL+ZjNAkhhBBCTJUpXdWudeC5c+dOnD9/HtXV1Y981U4DT0IIIYQQ8izqdXHRgQMH9FELIYQQQoiJMZ7nqOuD1quFunfvjrNnz+qjFkIIIYQQ8hzTesQzKysLo0ePhoODwyPneNLtlAghhBBCyLPQesQzODgYBw4cQEhICNavX183GRNzcxWOH1+L9Iz1yM4JwtKlU0TL9vJywyl1KHLzwjB//gSdMoKCZqOwKBKZWf+tm7ds+T+RkbEWaWnfYV/8CrRs2VxQnUHBc1FUHI2s7CBBORuD/HG5MByZmWsfmT/rfW+cPLkOWVnf44svpgp6DzGWqb5yTbHWJk0aYOPmiTiSPAe/nJiNHm6t0aVrS+w5+C/8dOQDxB+aBZfujkZR6wMdOzoiLX1d3VR2Yxf8/ccIzgWMe13pI5dHplj7q7+SSvsdHW2RkLAKOSc3ITsnGH5+xttXeawrKa1/fWCs9uIifUxGgdXDzZs32enTp9lvv/1WN2kTGBjIevfuzby9vevm3bhxg02dOpV5enqyqVOnsps3b9bn7Vl4eAQDlE+dLCwsGaBkSqU5O3ZsHXN3f0nLaxRaJ7lcxfLObGHt2jkylcqcpWcEM2fndk99jUxm8bep/4BerHv3l1lW9ua6eU2a2NT9289vAlu3bv5jXyuTWdSr1v79XZirayeWlR1Sr9+vbV/jv00DBvRm3Xt0Y9nZm+vmDRzYlx04+A1r0KAZk8sbMzs7x8e+Vi5vzGWZ8lpXpl5riyZ9HzttjfyazXnfn7Vo0pc5WPdnTo6DWOJPG9n4sVNZiyZ92YTXprIjv2x64uv5tN+s3pNc3oBdLtzJXnjBsR6/L411JaV+VZ9Jl/3V89R+e3tb5uraiQEKZmlpydSnI4y2Vh7rypjWv1qtrtc4hKeM2P0sb+j7epmMob1aj3iuWbMGvr6++Oyzz7By5UqsXLkSX375pdYB7dixYxEcHPzIvI0bN6JPnz44ePAg+vTpg40bN+o+Yv6Lysrae12qVEooVUowEe7G2qtXJ5w7V4QLF4pRVVWNHdsPw9e33zPnHE06hbKyikfmVVTcrft3I4sGgutNSjr5t/fQxdGjp3Cj7PYj8/71r+H46j87cf9+7ZOrrl4t1zlfrGWqj1xTrLVxE3P07tsWWyPSAABVVTW4deseGAMaNzYHUHtE9EqJ7n2N13J9YPBgF5w/V4xLl0oFZxnzupJyrWLtrx4mpfaXlJQhM7P22onbt+8iN/cSHBxsjLJWHutKKutfnxiT6WUyBloHnvHx8fjpp58QERFRN4WHh2sNdnNzg5WV1SPzEhMTMXr0aADA6NGjkZCQoFvVjyGXy5GWvh7FJdFITMhASkqu4MxWDjYoKPjzw+ty4VW0crAWnPvAik8m4/yFUEyc+CqWLYsULVdsL3ZshVde+QeO/voVEhI/R4+eTjpn8VqmPHJNsdYXXmiO69fu4Jv/voaDv7yPVd+OQcNGKvx74V4sWTEMab/Nw5IVw/H5ioMGr/VJ3hj/KrZtOyxKljGvK33k8l5XYpJq+9u0sYOLixOSk4V9ZklpXfFg6u2XEq0Dz44dO6KiQpy/TK5fv44WLVoAAGxtbXH9+nVRcgFAo9GgZ48ZaPPCRLi5dUKXLm1Fy+bl30si0L7dNERFHcas90caupwnUioUaNa8MV7pNw+BgaHYuvVjQ5dEOFEo5XipW0uEhyRjqMd/cefOfXwwxwNvvd0LSxfuQ8+uX2HZor1Y/Z0456SJTaVSwsenN2Jijhi6FEK0srBogB3RSxEQsA4VFXcMXQ4xII2eJmOgdeD53nvvYfTo0Zg+fTpmzJhRNwklk8kgk4l/2Le8vBKHD2fDy6un4Kyiwmto3bpF3c+ODrYoKhRvsPxA1NbDGDPGeL8SuFx4HbG7jwMA0lJ/h0ajgY1NE52yeC1THrmmWGtxUTmKi24hM/0yAODHH37DS91aYdzE7ti35xQAYE/sb4IuLuK5XQ0b7obMzLMoLb0pSp4xryt95OprHygGqbVfqVQgOmYZorYmInb3UcF5UlpXPJh6+6VE68AzMDAQ7777Lt599128/fbbdZMurK2tUVpaeyi8tLQUzZsLu5L7ARsbK1hZWQAAGjQww5Ah3ZGXVyA4NzU1D05ODmjb1h4qlRJvjH8Ve/YcE5wLAE5Orer+7evrjry8y6Lk8vDDDyfw6qsvAQBefLEVzMyUuHbtlk5ZvJYpj1xTrPVq6W0UFZajg1Pt+Wb9B3TA73mluFJ8C336tQMAvDKgPS6c132HznO7mjBBvK/ZAeNeV1KulQeptT8oeC5On76INWt2ipInpXXFg5Tbz2Ba53hqvY9ngwYNMGWKOLcnGjRoEGJjY/Hee+8hNjYWgwcPFiW3ZcvmCAmdD4VCDrlchpjoI9i7N1lwbk2NBrP912Jf/EooFHJsDt0PtfriM+dERM6Dh8dLsLFpggv5m7Fi+RYMG94THTs6gmk0uHjpKt6f9V/tQU8RuWUhPDy6wcbGCvkXo7B8eRhCQ/Y/e60RczHAoytsbJrg/IUQrFgRhc2hCQgK9kdm5lrcr6rG9Le/1blOsZapPnJNtdbF83/E9xvfgMpMgUv5Zfjw/Z04sO80VnzhDYVSjj/uVWPenFijqPVhjRqZY8iQ7pg5Q/f++VfGvq6kWqtY+yvetfJqf79+XTF5sidycs4jLb329oRLFocgPj7F6Grlsa6ksv4JHzKm5XLqL774AmZmZhg0aNAjN5Dv0qXLU4MDAgKQkpKCGzduwNraGn5+fhgyZAjmzJmD4uJitGrVCmvWrEHTpk21FhkREYkpU6bVr0X1Jvyq98eRyRqInsnYPdEzAUAubyR6pkZD5ylJRYsm7lxyS28J/6Pv7xQcMgGghlMuIUQK1OqTcHZ2NmgNGXEHYLZ2n17eS7F2hsHbq/WIp1qtBlD7BKMHZDKZ1ivbV69e/dj5YWFhz1AeIYQQQgh5XmgdeEZEROijDkIIIYQQE2Q851/qg9aB5/fff//Y+R988IHoxRBCCCGEkOeX1oFno0Z/ngf4xx9/4PDhw2jfvj3XogghhBBCTEHts9oNXYX+aB14/vXWSdOnT8f06dO5FUQIIYQQQp5PWgeef3X37l2UlJTwqIUQQgghxOTQOZ4P8fHxqfu3RqNBWVkZ3n//fa5FEUIIIYSQ54/Wgef69ev//GWlEtbW1lAqn/lAqcngdc9NHuiem6at9JbuN6vWP7rfJiGEPA+0jiAdHBz0UQchhBBCiEnSgL5qh6urK2Sy2gXx4OFGMpkMNTU1qKqqqruxPCGEEEIIIfXxxIFnZmbmIz9XVlZiy5Yt2L59Ozw9PbkXRgghhBBiCp7+8PLni9av2m/duoWwsDDExsZi5MiRiImJQbNmzfRRGyGEEEIIeY48ceBZVlaG0NBQ7Nu3D6+99hpiY2PRuHFjfdZGCCGEEPJcYzCt2ynJn/Q/Bg0ahL1792LMmDFo2LAhYmJiEBoaWjcZI7lcjtS0/yHuh09Ey/TycsMpdShy88Iwf/4EwXmOjrZISFiFnJObkJ0TDD+/MSJUCQQFz0VRcTSysoNEyQP41Mqr/YBprytA/PYDgLm5CsePr0V6xnpk5wRh6dIpouTyqFVKy9WU2y+l7UpK+ytemTxyefVVXrlEXE8ceE6fPh1jx44FUHt+518nbRYsWIA+ffpg5MiRdfPi4+Ph7e2Nzp074+TJkyKU/yh//zHIzb0kWp5cLsd3a/0w0nshXuo6HeMnDISz8wuCMqurazBv3nq8/NJ09Ovrh5mzRgnOBIDwsAPwHrFAcM7DeNTKq/2mvq54tB8A/vijCkOGzEOP7jPQo/sMeHn1hLu7s1HWKpXlaurtl9J2JaX9Fa9+xSOXx7rimasPGsj0MhmDJw48/fz88MEHHzxx0mbs2LEIDg5+ZF7Hjh2xdu1auLm5Ca/8LxwcbDBihDtCNsWLltmrVyecO1eECxeKUVVVjR3bD8PXt5+gzJKSMmRmngUA3L59F7m5l+DgYCO41qSkkygrqxCc8zAetfJqv6mvKx7tf6CysvbetCqVEkqVsu4uF7riVatUlqupt19K25WU9le8+hWPXB7rimcuEdcTB55Cubm5wcrK6pF5HTp0QPv27bm83+pvZiIwMAgajUa0zFYONigoKK37+XLhVbRysBYtv00bO7i4OCE5OVe0TF541CpmpqmvK57tl8vlSEtfj+KSaCQmZCAlRdgy4L2uxMSjVlNv/8OMfbt6mLHvr3itKyn1V6mqPcdTP5Mx4Dbw1Cdvb3eUlt5ERsbvhi6l3iwsGmBH9FIEBKxDRYVxP0GIR62m3n4p0Wg06NljBtq8MBFubp3QpUtbQ5dEngNS2q6kVCshxu65ePZl375d4OPTB8OH90KDBmZo0qQRwsI/xltTvhSUW1R4Da1bt6j72dHBFkWF14WWC6VSgeiYZYjamojY3UcF5/HEo1Yemaa+rni1/2Hl5ZU4fDgbXl49cepUvs45+qhVLDxqNfX2A9LZrgDp7K94rSsp9VfJYjJo6Kr2P61evRq3bt2q+7m8vBzffPMN16Ke1aJFIWjbZhKcOkzGm5M+w88/ZwkedAJAamoenJwc0LatPVQqJd4Y/yr27DkmODcoeC5On76INWt2Cs7ijUetPDJNfV3xar+NjRWsrCwAAA0amGHIkO7Iyyswylp54FGrqbcfkM52BUhnf8VrXUmpvxJp0DrwPHLkCJo0aVL3s5WVFY4cOcK1KGNRU6PBbP+12Be/Er+dCkFM9C9Qqy8KyuzXrysmT/bEwIGuSEtfj7T09Rg+vJfgWiO3LMTRX79Dp06tkX8xCtPeHiY4k0etvNpv6uuKR/sBoGXL5khIXIWMzA04kfw9EhIysHdvslHWKpXlaurtl9J2JaX9Fa9+xSOXx7rimasPDDK9TMZAxrRcourj44OdO3fCzMwMAHDv3j289tpr2Lt371ODAwICkJKSghs3bsDa2hp+fn5o2rQpPvnkE5SVlaFJkyZwdnbGpk2btBYZERGJKVOmPUOz6sNIzrIlxGB47YRo2yKESINafRLOzsJuESdU6u6fcH/1T3p5r+Ybpxm8vVrP8fTx8cFbb71Vd0/PXbt2YfTo0VqDV69e/dj59Jx3QgghhJBaDIDGhP5e1zrwfO+999CpUyecOHECADBr1iz079+fe2GEEEIIIeT5Uq+r2j08PODh4cG7FkIIIYQQ8hx74sBz4sSJiIqKgqurK2SyP88FY4xBJpMhIyNDLwUSQgghhDzPTOl2Sk8ceEZFRQEAMjMz9VYMIYQQQgh5ftXrq/ZTp04hPT0dMpkMPXr0wD/+8Q/edRFCCCGEmARjudWRPmgdeH7//fc4cOBA3dXogYGBGDZsGGbNmsW9OPKAgkuqjENHZ6gWPZMfHhu6lC5NlFKtfLYBoIZTLjFtpr5vIeTJtA489+zZgx9++AHm5uYAaq9yHzVqFA08CSGEEEIEMrXbKWl9clGLFi3wxx9/1P18//592NnZcS2KEEIIIYQ8f7Qe8WzcuDG8vb3Rr18/yGQy/Prrr3j55Zfx6aefAgAWL17MvUhCCCGEkOcVneP5EE9Pz0eeNtSrl/Dn1BJCCCGEENOjdeA5ZswY3L9/H/n5+QCAdu3aQaVS8a6LEEIIIcQk0DmeD0lOToaXlxdWrFiB5cuXw8vLC6mpqfqord7MzVU4fnwt0jPWIzsnCEuXThElNyh4LoqKo5GVHSRK3gNeXm44pQ5Fbl4Y5s+fIFru2XNhyMxaj7T0dTiRvFaUTD+/UcjMXo+snA3w9x8tSiaP5cojU0r9ildflco20LGjI9LS19VNZTd2wd9/jAiV8tleee0DpFKrVPoVIK19i1T6laOjLRISViHn5CZk5wTDz0/4tsojk3DCtBgzZgw7d+5c3c/nz59nY8aM0fYyFhgYyHr37s28vb3r5q1cuZJ5eXmxkSNHslmzZrHy8nKtOYwxFh4ewQDlUycLC0sGKJlSac6OHVvH3N1f0vIahdapf38X5uraiWVlh9Tr9+szyeUqlndmC2vXzpGpVOYsPSOYOTu30/I6s3pNZ89FMWtrm3r/vgzmT526dunIsrJDWaOGTZhS0ZAdOLCaOXVo+9TXGGq56pb59D4lpX7FI1OK2wBgxuTyBuxy4U72wguO9fh9HrXqP1NqtRpPv+JVq/73LVLqV/b2tszVtRMDFMzS0pKpT0cYLFOtVtdrHMLT8Z0/sf3uC/QyGUN7tR7xrKqqQvv27et+bteuHaqqqrQOaMeOHYvg4OBH5vXr1w8//vgj9uzZg7Zt22LDhg06DJUfr7LyHgBApVJCqVKCMSY4MynpJMrKKgTnPKxXr044d64IFy4Uo6qqGju2H4avbz9R30MsnZ1fQGpKHu7e/QM1NRocOXISo8cIr5XHcuWRCUinX/FqvxS3gcGDXXD+XDEuXSoVnMWjVl7tl1KtUupXUtm3SKlflZSUITPzLADg9u27yM29BAcHG6PLJHw8ceAZGRkJAOjatSsWLVqE5ORkJCcnY/HixejatavWYDc3N1hZWT0y75VXXoFSWXtaqYuLC0pKSoTU/gi5XI609PUoLolGYkIGUlJyRcsWUysHGxQU/PmBeLnwKlo5WIuSzRgQv/9zJKd8j3feHS4479Rv+ej3Shc0b94YDRuaY/hwN7RubStCpdIhlX4lJTy3AQB4Y/yr2LbtsChZPGrl1X4p1cqDlGoFxN+3SKlfPaxNGzu4uDghOVm8fSuPTN6YniZj8MSB586dOwEAy5cvR4cOHRAREYGIiAg4OTlh+fLlgt94586dGDBggOCcBzQaDXr2mIE2L0yEm1sndOnSVrRsqfAYEIBebh9gpPcizJzpi/79tf+B8DS5uQVY9VU04vd/jr37PkV29jnU1GhEqlYaqF9Ji0qlhI9Pb8TEHDF0KYQ8Fe1bAAuLBtgRvRQBAetQUXHHaDOJuLRe1W5mZoa3334bb7/9tmhv+r///Q8KhQK+vr6iZT5QXl6Jw4ez4eXVE6dO5YueL1RR4TW0bt2i7mdHB1sUFV4XJ7uoNufq1XLExf4KN7fOSEr6TVBmaMgBhIYcAAB88ulUFF6+JrhOKTL2fiUlPLeBYcPdkJl5FqWlN0XJ41Err/ZLqVYepFTrw8Tat0ipXwGAUqlAdMwyRG1NROzuo4LzeGXqi4aZzn08n3jEMy8vD927d//b5Orqiu7du+v8hrt27cLhw4exatUqyGTiLGgbGytYWVkAABo0MMOQId2Rl1cgSrbYUlPz4OTkgLZt7aFSKfHG+FexZ88xwbmNGpnD0rJh3b89PXuIMkCyta09XaJ1a1uMHtMPUVE/C86UCin1KynhtQ0AwIQJ4n3NDvCplVf7pVQrD1Kqlce+RUr9Cqi9W8Dp0xexZs1OwVk8M4n4nnjEs2PHjoiNjRX1zY4cOYLg4GBERkaiYcOGouW2bNkcIaHzoVDIIZfLEBN9BHv3JgvOjdyyEB4e3WBjY4X8i1FYvjwMoSH7BWXW1Ggw238t9sWvhEIhx+bQ/VCrLwqu1c6uGWJ2LgVQ+1fftqifceBAmuDcHdFL0Ny6MaqrauDv91+Ul1cKzuSxXHlkSqlf8cjklctrG2jUyBxDhnTHzBnfCs56gEetvNovpVql1K+ksm+RUr/q168rJk/2RE7OeaSlrwcALFkcgvj4FKPKJHzI2BMupRs9erSggWdAQABSUlJw48YNWFtbw8/PDxs3bsT9+/fRtGlTAEC3bt2wYsUKrVkREZGYMmWazrU8nrGcZlsfCi6pMg6P6GKoFj2THx5fbUipX0kJn20AqOGUS0wb7VukQq0+CWdnZ4PWcHxXAq5/qZ9vEzts/qfB2/vEI57Dhg0TFLx69eq/zRs3bpygTEIIIYQQIl1PHHjOmDFDn3UQQgghhJgkRhcXEUIIIYQQIi6tt1MihBBCCCH8mNIdsrUe8bx27RoWLlyId955BwBw9uxZREdHcy+MEEIIIYTozx9//IHXX38dvr6+8Pb2xnfffQcAKCgowLhx4+Dp6Yk5c+bg/v37AID79+9jzpw58PT0xLhx43D58mWt76F14BkYGIhXXnkFpaW1j8xq27YtwsPDhbSLEEIIIYSg9n4FGqafSRszMzOEhYXhhx9+QGxsLJKSkpCVlYVVq1Zh6tSp+Omnn9CkSRPExMQAAKKjo9GkSRP89NNPmDp1KlatWqX1PbQOPG/cuIERI0ZALq/9VaVSWfdvQgghhBDyfJDJZLCwqH24QXV1NaqrqyGTyXDixAl4eXkBAMaMGYPExEQAwKFDhzBmzBgAgJeXF44fP44n3KWzjtZzPBs1aoQbN27UPWUoKysLjRs31r1VOjPle5jxudegKS/RWrQEpIPut0mkhPYt5NkwLvd+1U1NTQ3Gjh2LS5cuYdKkSWjdujWaNGkCpbJ2yGhvb48rV64AAK5cuYKWLVsCqD0w2bhxY9y4cQPNmzd/Yr7WgWdgYCBmzpyJS5cuYcKECbhx4wa+/Va8J4MQQgghhBD+ysrKMHbs2Lqfx48fj/Hjxz/yOwqFAnFxcbh16xbef/99nD9/XtQatA48u3TpgsjISFy4cAGMMbRr1w4qlUrUIgghhBBCTFV9zr8UQ/PmzbFr1656/W6TJk3g7u6OrKws3Lp1C9XV1VAqlSgpKYGdnR0AwM7ODsXFxbC3t0d1dTUqKirQrFmzp+bW62TNnJwc5ObmQq1WY+/evaI/w50QQgghhBhWWVkZbt26BQC4d+8ejh07hg4dOsDd3R0HDhwAAOzevRuDBg0CAAwaNAi7d+8GABw4cAC9e/euOzXzSbQe8Zw3bx4KCgrQuXNnKBS1z0uWyWQYPXq0zg0jhBBCCCEAIDOaczxLS0sRGBiImpoaMMYwbNgwDBw4EE5OTvjwww+xZs0aODs71z0C/fXXX8e8efPg6ekJKysrfPPNN1rfQ+sRz99++w1RUVFYtmwZlixZgiVLlmDx4sXCWycyLy83nFKHIjcvDPPnTzDqXB6Zjo62SEhYhZyTm5CdEww/vzGi5PKoNSh4LoqKo5GVHSRKHq9MnrlS6Vemvl3xypXKdgVIp/28cqW0v5LScpVKpqnp3LkzYmNjsWfPHvz444/44IMPAACtW7dGTEwMfvrpJ3z33XcwMzMDAJibm+O7777DTz/9hJiYGLRu3Vr7mzAt/Pz82JUrV7T92t8EBgay3r17M29v77p533zzDRs5ciTz9fVl06ZNYyUlJfXKCg+PYIDiiZNcrmJ5Z7awdu0cmUplztIzgpmzc7unvqY+E49cXrXa29syV9dODFAwS0tLpj4dYbS19u/vwlxdO7Gs7BDBWTwzeeVKpV/RdiWtWk25r0ptuUplXUmpD+iaqVarn3l8I7ajMQlsW/clepmMob1PPOI5Y8YMzJgxAzdu3IC3tzemT59eN2/GjBlaB7Rjx45FcHDwI/Peeecd7NmzB3FxcXj11Vfx3//+V/vIuB569eqEc+eKcOFCMaqqqrFj+2H4+vYzylxetZaUlCEz8ywA4Pbtu8jNvQQHBxujrDUp6STKyioE5/DO5JUrlX5F25W0ajXlvsozVyr7KyktV6lkEj6eeI7n22+/LSjYzc3tb49OsrS0rPv33bt3tZ6AWl+tHGxQUFBa9/Plwqvo1auzUebyqvVhbdrYwcXFCcnJuYJy9FGrKZJKv6LtSlq18iCl9ktpufIgpeUqlUzCxxMHnr169QJQ+3zOFi1awNzcHEDtVU7Xrl3T+Q2/+eYbxMbGonHjxvToTQ4sLBpgR/RSBASsQ0XFHUOXQwghhBAt9HU7JWOg9eKi2bNnP3JkUi6XY/bs2Tq/4YcffohffvkFPj4+iIyM1DnnYUWF19C6dYu6nx0dbFFUeN0oc3nVCgBKpQLRMcsQtTURsbuPCs7jWaspk0q/ou1KWrXyIKX2S2m58iCl5SqVTMKH1oFnTU1N3dVLQO0D5KuqqgS/sY+PDw4ePCg4BwBSU/Pg5OSAtm3toVIp8cb4V7FnzzGjzOVVK1B7peTp0xexZs1OUfJ41mrKpNKvaLuSVq08SKn9UlquPEhpuUolU5/Y/99SifdkDLTex7N58+ZITEzE4MGDAQAJCQla70r/JPn5+Wjbti0AIDExEe3bt9cp569qajSY7b8W++JXQqGQY3PofqjVF40yl1et/fp1xeTJnsjJOY+09PUAgCWLQxAfn2J0tUZuWQgPj26wsbFC/sUoLF8ehtCQ/UaXyStXKv2Ktitp1WrKfZVnrlT2V1JarlLJJHzIGGNPPbPg0qVLmDt3LkpLS8EYQ8uWLfHll1+iTZs2Tw0OCAhASkoKbty4AWtra/j5+eHIkSO4cOECZDIZHBwcsHz58rrHLj1NREQkpkyZ+kwNI4QQQgh5GrX6JJydnQ1aw9Gdicj/9Ihe3qvH1gkGb6/WI54vvPACduzYgcrKSgCAhYVFvYJXr179t3kP7nRPCCGEEEJMzxMHnnFxcRg1ahRCQ0Mf+/+nTZvGrShCCCGEEFOhMXQBevTEgefdu3cBoO5IJyGEEEIIIUI8ceA5YULtc04fPKeTEEIIIYSIjzHjuOJcH5448Pz000+f+sLFixeLXgwhhBBCCHl+PXHg2aVLF33WQQghhBBichjoHE8AwJgxY/RZByGEEEIIec5pvZ3ShQsXEBISgsLCQlRXV9fNp+esE0IIIYQIZ0rPatc68Jw9ezYmTJiAcePGQS7X+oRNQgghhBBCHkvrwFOpVGLSpEn6qIUQQgghxOSY0AFPPPEQ5s2bN3Hz5k0MHDgQW7ZsQWlpad28mzdv6rFEQgghhBDyPHjiEc+xY8dCJpPhwaPcN23aVPf/ZDIZEhMT+VdHCCGEEEKeG08ceB46dEifdRBCCCGEmBzGAI0J3UBe69VCW7Zswa1bt+p+Li8vx5YtW7gWpQsvLzecUociNy8M8+dPMOpcqWQGBc9FUXE0srKDRMl7wNTXFa9cqWTyyqVapVOrqe+vpFQrr1ypZBIOmBa+vr5/mzdq1ChtL2OBgYGsd+/ezNvb+2//b9OmTaxjx47s+vXrWnMYYyw8PIIBiidOcrmK5Z3Zwtq1c2QqlTlLzwhmzs7tnvqa+kw8cqWSCShY//4uzNW1E8vKDhGcRetKerWaevupVulkAtLaX0mpVqn0AV0z1Wp1vcYhPP0SncDWv7RML5MxtFfrEU+NRlN3nicA1NTUoKqqSuuAduzYsQgODv7b/OLiYvz6669o1arVMw6Rn6xXr044d64IFy4Uo6qqGju2H4avbz+jzJVKJgAkJZ1EWVmF4JyHmfq6klKtpt5+qlU6mYC09ldSqlUqfYBX+4n4tA48X3nlFcyZMwfHjx/H8ePHERAQgP79+2sNdnNzg5WV1d/mf/HFF5g3bx5kMvHOZ2jlYIOCgtK6ny8XXkUrB2ujzJVKJi+mvq545Uolk1cu1SqdWml/xQf1K+msq8fRMP1MxkDrfTznzZuHbdu2ISoqCgDQt29fjBs3Tqc3S0hIQIsWLdC5c2edXk8IIYQQQqRL68BTLpdj0qRJgm8if/fuXWzYsAEhISGCch6nqPAaWrduUfezo4MtigqvG2WuVDJ5MfV1xStXKpm8cqlW6dRK+ys+qF9JZ139FQOgMXQReqT1q/b8/Hz4+/tjxIgRGDx4cN30rC5duoTLly9j1KhRGDRoEEpKSjB27FhcvXpVp8IflpqaBycnB7Rtaw+VSok3xr+KPXuOGWWuVDJ5MfV1JaVaTb39VKt0MnmhWqXTB6S0rkzdE494Tpw4EVFRUViwYAH8/f3x+eefIzw8HLt27YJG8+xj806dOuH48eN1Pw8aNAgxMTFo3ry5bpU/pKZGg9n+a7EvfiUUCjk2h+6HWn3RKHOlkgkAkVsWwsOjG2xsrJB/MQrLl4chNGS/UdYqpeUqlVpNvf1Uq3QyAWntr6RUq1T6AK/26wszkvMv9UHG2OObe/v2bVhaWmLs2LHYtWsXfHx8sGfPHgCom/c0AQEBSElJwY0bN2BtbQ0/P79Hzg19loFnREQkpkyZ+gzNIoQQQgh5OrX6JJydnQ1awy8xiTi57Khe3mtw9BsGb+8Tj3h+9NFH2LBhA8zMzKDRaNCmTRtERkbCzs4OlZWVWoNXr1791P9PT0YihBBCCAE0oCcXYcOGDQCABQsW4M6dO1i8eDFOnTqFuLg4fPnll3orkBBCCCGEPB+0XtXerVs3AIClpSU+++wz3LlzB5aWltwLI4QQQgh53jGY1jmeWq9q/+ijj3D79m3cuXMHI0eOxIgRIx77RCJCCCGEEEKeRuvA8+zZs7C0tERCQgIGDBiAxMRExMXF6aM2QgghhJDnnkZPkzHQOvCsrq5GVVUVEhISMGjQIKhUKn3URQghhBBCnjNaB57jx4/HoEGDcPfuXbi5uaGwsBCNGzfWR22EEEIIIeQ5ovXiooEDB2LKlCl1P7dq1Qqff/4516IIIYQQ6eJxaxwTuvrksZ7j2w0xQGNCq1frEU9/f/9HfpbJZAgICOBWECGEEEIIeT498YjnuXPncPbsWVRUVODgwYN182/fvo0//vhDL8URQgghhDzPGEzrePYTB54XLlzA4cOHUVFRgZ9//rluvoWFBT755BO9FEcIIYQQQp4fTxx4DhkyBEOGDEFmZiZcXV31WRMhhBBCiMmgczwf8o9//ANbtmzBsmXLsGDBgrrJ2Hh5ueGUOhS5eWGYP3+CUedKJdPR0RYJCauQc3ITsnOC4ec3RpRcU19XvHLFzuS1/gFptJ9nrqnXGhQ8F0XF0cjKDhIlj1cmr1xzcxWOH1+L9Iz1yM4JwtKlU7S/qB5MvV/xWq5EZEwLPz8/9s0337DBgwezXbt2sWnTprFPPvlE28tYYGAg6927N/P29q6b991337FXXnmF+fr6Ml9fX3b48GGtOYwxFh4ewQDFEye5XMXyzmxh7do5MpXKnKVnBDNn53ZPfU19Jh65UskEFMze3pa5unZigIJZWloy9ekIo61VSstVKrXyWP9Saj/Vyq/W/v1dmKtrJ5aVHSI4i2em7rlKrZOFhSUDlEypNGfHjq1j7u4vaXmNqfcr7ctUl+WqVqvrNQ7h6dCOBLay03K9TMbQ3ice8ayurgYAXLp0CXPmzEHDhg0xZswYbNiwATk5OVoHtGPHjn3sozWnTp2KuLg4xMXFwcPDQ8CQ+U+9enXCuXNFuHChGFVV1dix/TB8ffsZZa5UMgGgpKQMmZlnAQC3b99Fbu4lODjYGGWtUlquUqmVx/rnVaupryup1ZqUdBJlZRWCc3hn8sytrLwHAFCplFCqlGACH9ZN/aqW2MuViO+JA89x48YBAJTK2tNAmzRpgjNnzqCiogLXr1/XGuzm5gYrKyuRyny6Vg42KCgorfv5cuFVtHKwNspcqWT+VZs2dnBxcUJycq6gHFNfV7xyefcBsdY/IK32U6389y2mSi6XIy19PYpLopGYkIGUFNq3ikHs5aoPDPTIzEeMHz8e5eXlmDNnDmbOnAlvb2+8++67Or/hli1b4OPjgwULFqC8vFznHKI/FhYNsCN6KQIC1qGi4o6hyyF6RuufEPFpNBr07DEDbV6YCDe3TujSpa2hS3ou0HI1fk8ceF6/fh2hoaG4ffs2du3ahd9++w1vvvkm3n33Xdy9e1enN5s4cSJ++uknxMXFoUWLFli5cqXOhT+sqPAaWrduUfezo4Mtigq1H5U1RK5UMh9QKhWIjlmGqK2JiN19VHCeqa8rXrm8ahV7/QPSaj/Vym/fQmqVl1fi8OFseHn1FJRD/epRYi1XfdEw/UzG4IkDT41Gg8rKykemO3fu4M6dO6isrNTpzWxsbKBQKCCXyzFu3DicPHlS58IflpqaBycnB7Rtaw+VSok3xr+KPXuOGWWuVDIfCAqei9OnL2LNmp2i5Jn6upJarWKvf0Ba7ada+e1bTJmNjRWsrCwAAA0amGHIkO7IyysQlEn9is9yJeJ74n08bW1t8cEHH4j6ZqWlpWjRovavnISEBLz44oui5NbUaDDbfy32xa+EQiHH5tD9UKsvGmWuVDIBoF+/rpg82RM5OeeRlr4eALBkcQji41OMrlYpLVep1Mpj/fOq1dTXldRqjdyyEB4e3WBjY4X8i1FYvjwMoSH7jS6TV27Lls0REjofCoUccrkMMdFHsHdvsqBM6ld8lqu+GMnBSL2QsSdc8jV69GjExsbqHBwQEICUlBTcuHED1tbW8PPzQ0pKCnJza0/0dXBwwIoVK+oGok8TERGJKVOm6lwLIYQQoj8yDpmmNDR5HB7LFFCrc+Ds7Mwlu74ORSfi2OJf9fJer8WOM3h7n3jEc/PmzYKCV69e/bd5D66UJ4QQQggh/39Vuwn9XfHEczybNm2qxzIIIYQQQsjzTuvtlAghhBBCCBHDE79qJ4QQQgghvMnAOJ3DaozoiCchhBBCCNELOuJJCCGEEGIoRnRzd32gI56EEEIIIUQvJHTEU+zzH0zozwu94nWeCq0v00b9ikgJ9SvxPb/LlG6nRAghhBBCCAcSOuJJCCGEEPL8MaEDnnTEkxBCCCGE6Acd8SSEEEIIMSA6x5MQQgghhBCRPRcDT3NzFY4fX4v0jPXIzgnC0qVTRMv28nLDKXUocvPCMH/+BKPNDAqei6LiaGRlB4mSxyuT1hWf5QpIp/08cqlf8cuVSiaPXEdHWyQkrELOyU3IzgmGn98YEaqU1j6AV65UMvWB6fE/o8A4CQwMZL1792be3t6PzA8PD2deXl5sxIgR7Msvv6xXVnh4BAOUT50sLCwZoGRKpTk7dmwdc3d/SctrFFonuVzF8s5sYe3aOTKVypylZwQzZ+d29XqtPjMBBevf34W5unZiWdkhgrOEZT59Pem2rrSvL1NfV1Jqv2651K+el1ql1H57e1vm6tqJAQpmaWnJ1KcjTGofIKU+oGumWq3mMQR6Jge3J7DA9sv1MhlDe7kd8Rw7diyCg4MfmXfixAkkJibihx9+wN69ezF9+nTR3q+y8h4AQKVSQqlSgjEmOLNXr044d64IFy4Uo6qqGju2H4avbz+jywSApKSTKCurEJzDOxOgdcVjuUqp/bxyTb1fSaVWKbW/pKQMmZlnAQC3b99Fbu4lODjYCK5VKvsAXrlSydQnDdPPZAy4DTzd3NxgZWX1yLyoqCi89957MDMzAwBYW1uL9n5yuRxp6etRXBKNxIQMpKTkCs5s5WCDgoLSup8vF15FKwdhNfPIlBpaV+KTUvt55Zp6v5JKrVJq/8PatLGDi4sTkpOF9ysepLRcpZJJ+NDrOZ75+flIS0vDuHHj8M9//hM5OTmiZWs0GvTsMQNtXpgIN7dO6NKlrWjZRFy0rggP1K8ILxYWDbAjeikCAtahouKOocshzyGmp8kY6HXgWVNTg/LycuzYsQPz58/HnDlzRPk67GHl5ZU4fDgbXl49BWcVFV5D69Yt6n52dLBFUeF1o8uUKlpX4pFS+3kvV1PtV1KpVUrtBwClUoHomGWI2pqI2N1HBefxIqXlKpVMwodeB552dnbw9PSETCbDyy+/DLlcjhs3bgjOtbGxgpWVBQCgQQMzDBnSHXl5BYJzU1Pz4OTkgLZt7aFSKfHG+FexZ88xo8uUElpXfEip/TxyqV9Jp1YptR+ovQL99OmLWLNmp+AsnqS0XKWSSfjQ6w3khwwZguTkZPTu3RsXLlxAVVUVmjVrJji3ZcvmCAmdD4VCDrlchpjoI9i7N1lwbk2NBrP912Jf/EooFHJsDt0Ptfqi0WUCQOSWhfDw6AYbGyvkX4zC8uVhCA3Zb3SZtK74LFcptZ9HLvUr6dQqpfb369cVkyd7IifnPNLS1wMAliwOQXx8iqBcqewDeOVKJVNfGIznwh99kDGxv+v+fwEBAUhJScGNGzdgbW0NPz8/jBo1CgsXLkRubi5UKhXmz5+PPn36aM2KiIjElCnTRK7QhNayXsk45dL6Mm3Urwgh4lOrT8LZ2dmgNRzckYiDH/+ql/eavm+cwdvL7Yjn6tWrHzt/1apVvN6SEEIIIURaGMDnEKBxei6eXEQIIYQQQoyfXs/xJIQQQgghj9IYugA9oiOehBBCCCFEL+iIJyGEEEKIgZjaVe10xJMQQgghhKC4uBiTJ0/GiBEj4O3tjbCwMADAzZs3MW3aNAwdOhTTpk1DeXk5AIAxhk8//RSenp7w8fHBqVOntL6HhI54ij1GrhE5jx8Zr9UkE//vDsbui55JCN32SDoUiiZccmtqbnHJ5UPBIVM6n1n8bn/2/DKWq9oVCgUCAwPRpUsX3L59G6+99hr69euHXbt2oU+fPnjvvfewceNGbNy4EfPmzcORI0eQn5+PgwcPIjs7G8uWLUN0dPRT34OOeBJCCCGEELRo0QJdunQBAFhaWqJ9+/a4cuUKEhMTMXr0aADA6NGjkZCQAAB182UyGVxcXHDr1i2UlpY+9T1o4EkIIYQQYkAaPU3P4vLlyzh9+jS6deuG69evo0WLFgAAW1tbXL9+HQBw5coV2Nvb173G3t4eV65ceWquhL5qJ4QQQgghuiorK8PYsWPrfh4/fjzGjx//t9+rrKyEv78/Fi5cCEtLy0f+n0wmg0ym++kUNPAkhBBCCDEgfZ3j2bx5c+zateupv1NVVQV/f3/4+Phg6NChAABra2uUlpaiRYsWKC0tRfPmzQEAdnZ2KCkpqXttSUkJ7OzsnppPX7UTQgghhBAwxrBo0SK0b98e06ZNq5s/aNAgxMbGAgBiY2MxePDgR+YzxpCVlYXGjRvXfSX/JM/NwPPsuTBkZq1HWvo6nEheK1qul5cbTqlDkZsXhvnzJxht5uzZY5CVswGZ2esRsSUQ5uYqnXKCguagsGgrMrPW1c1r1swS8fs/g/p0EOL3f4amTS2fkqAdj/bzyHV0tEVCwirknNyE7Jxg+PmNEaHKWmLXGhQ8F0XF0cjKDhKhuj9JZV0B0loGUlquYmQ6OlrjwMHlyMr+FplZa/DBB94Aavct+/YtxalT32PfvqVo2tRC5zp5rX9euTw+s6TUrwBALpcjNe1/iPvhE6PO5I3BeM7xTE9PR1xcHE6cOIFRo0Zh1KhR+OWXX/Dee+/h119/xdChQ3Hs2DG89957AAAPDw+0bt0anp6eWLJkCZYuXVqPBnMSGBjIevfuzby9vevmzZ49m/n6+jJfX182cOBA5uvrW6+s8PAIBpg9dTp7LopZW9to/b0/J4XWSS5XsbwzW1i7do5MpTJn6RnBzNm5Xb1eK2amDOZPnRxatWK/n41iDRs0ZjKYs6hty9lbb43U+jqZrOHfpv4D3Fj37l1ZVnZo3byVX77P5n88hclkDdn8j6ewL1bOeuxrZbKGBlmmvHLt7W2Zq2snBiiYpaUlU5+OMNpa+/d3Ya6unVhWdojg+qS4rqS0DKS0XHXJVCia/W1ycGjPevTowRSKZszKqiU7nbuFden6EvvyP/7s48C3mULRjH0c+DZb+aX/Y1+vUDQzyPrXPVf7Z5DYn1nG1a+U9ZrmzBnPIiKXsLgfvqj3a8TOVKvVvIZB9Ra/LYHNcFyul8kY2svtiOfYsWMRHBz8yLw1a9YgLi4OcXFxGDp0KDw9PXm9vSh69eqEc+eKcOFCMaqqqrFj+2H4+vYzukwAUCoVaNjQDAqFHI0amaOo6LpOOUeTfkNZWcUj83x8eiMivPbWCRHhCfD17aNznbzazyO3pKQMmZlnAQC3b99Fbu4lODjYGGWtSUkn/7behJLSugKkswyktFzFyiwpuYGsrPMAgNu37yE39zIcWlnDx6cXIiMOAwAiIw7D17eXzrXyWP88c8UmpX4FAA4ONhgxwh0hm+IFZ/HMJOLjNvB0c3ODlZXVY/8fYwzx8fEYOXKkaO/HGBC//3Mkp3yPd94dLkpmKwcbFBT8eT+qy4VX0crB2ugyi4qu45uvY3A+PwIFhVtxq7wSCT9lCMp8mJ1dU5SU3ABQ+wFiZ9dU5ywe7eeZ+0CbNnZwcXFCcnKu4CzetYpFqutKTFLZB/DK5ZHZpo0tunVrh5SUM2jR4tF9S4sWTQVlS4nYn1lS6lcAsPqbmQgMDIJG86w3+dFvpr4wxvQyGQODnOOZlpYGa2trtG3bVrRMjwEB6OX2AUZ6L8LMmb7o37+raNnGrmlTS/j49sGLHabiBcc30ciiASa9OYjb+xlL59UXC4sG2BG9FAEB61BRccfQ5RAiWRYWDbBt+3zMnRuCioq7f/v/prRvMeXPLG9vd5SW3kRGxu9GnUn4MMjA88cffxT1aCeAuq+Wr14tR1zsr3Bz6yw8s/AaWrf+8+osRwdbFBXq9hU2z8zBQ1yRn38F166Vo7q6BrG7f0WfPs6CMh925cpN2Ns3AwDY2zdDaWm5zlk82s8zV6lUIDpmGaK2JiJ291HBeQC/WsUmtXXFg1T2AbxyxcxUKhXYvn0etkUdQVxsMgCgtPTRfcvVq7rvW6RG7M8sKfWrvn27wMenD86ei8CWrYswcKALwsI/NrpMvWGARk+TMdD7wLO6uho//fQTRowYIVpmo0bmsLRsWPdvT88eOHUqX3BuamoenJwc0LatPVQqJd4Y/yr27DlmdJkFl0rRy70zGjY0BwAMGuSC3NMFgjIf9uOPJzB5yhAAwOQpQ7Bnzwmds3i0n2duUPBcnD59EWvW7BSc9QCvWsUmtXXFg1T2AVKodcPG95GbW4hvv91TN+/HPan45+RXAQD/nPwq9uxJEVSvVPD4zJJSv1q0KARt20yCU4fJeHPSZ/j55yy8NeVLo8skfOj9BvLHjh1D+/btH3nEklB2ds0Qs3MpgNq/qrdF/YwDB9IE59bUaDDbfy32xa+EQiHH5tD9UKsvGl1mSkoedu1MQkra96iurkF21jkEBel2cnVE5Hx4eLwMG5smuJAfjhXLI/GfL6MRtW0Bpk0bikuXSjFxwhc618qj/bxy+/XrismTPZGTcx5p6esBAEsWhyA+XtiHI49aI7cshIdHN9jYWCH/YhSWLw9DaMh+o6uTZ65UloGUlqtYmX37dsY///kqTp7MR0rq1wCAfy/Zgq++2oWtW+di2tTBuHTpKiZN+lrnWnmsf165PD6zpNSvyKPY/0+mQsY4nVQTEBCAlJQU3LhxA9bW1vDz88O4ceMQGBiIbt26YeLEifXOioiIxJQp00WusEbkPH5kvP4+kIl/wJux+6JnEkKkQ6FowiW3puYWl1w+FBwypfOZBej+OEV9U6tz4Ows3qlpuojflohdH4lzKpc2AQlvGLy93I54rl69+rHzV65cyestCSGEEEIkx1jOv9SH5+bJRYQQQgghxLjp/RxPQgghhBDyJzriSQghhBBCiMjoiCchhBBCiIHUXtVuOoc86YgnIYQQQgjRCzriSQghhBBiQKZ0jqeEBp5SuoeZuBivtrNqDqE87l8HABoOmSa0pROiJ/zutyn+vkXG6X6TDDz2rVJC+1byZBIaeBJCCCGEPH/4PMrHONE5noQQQgghRC/oiCchhBBCiIEwABoTOj2BjngSQgghhBC9oIEnIYQQQgjRi+dm4Onl5YZT6lDk5oVh/vwJRp3LI9PcXIXjx9ciPWM9snOCsHTpFMGZQcFzUVQcjazsIBEq/NPZc2HIzFqPtPR1OJG8VtRsuVyO1LT/Ie6HT0TJ47EMeC1XqfRVXrlUqzRqdXS0RULCKuSc3ITsnGD4+Y0RocpaPPYts2ePQVbOBmRmr0fElkCYm6sEZ0ppHyClWnltV9yx2ouL9DEZBcZJYGAg6927N/P29q6bp1ar2bhx45ivry8bM2YMy87OrldWeHgEAxRPnORyFcs7s4W1a+fIVCpzlp4RzJyd2z31NfWZeOTqlqms12RhYckAJVMqzdmxY+uYu/tLWl7z9Fr793dhrq6dWFZ2yDO00UzrdPZcFLO2tqnX7/451W8ZzJkznkVELmFxP3xRj9/X3h7dloH+M42nr1KtVOvTJ3t7W+bq2okBCmZpacnUpyPqmSn+vkUG86dODq1asd/PRrGGDRozGcxZ1Lbl7K23Rmp9nba2SGUfIKVadc1Uq9W8hkH19uPWBDbR9t96mYyhvdyOeI4dOxbBwcGPzPvqq6/w/vvvIy4uDrNnz8ZXX30lynv16tUJ584V4cKFYlRVVWPH9sPw9e1nlLm8agWAysp7AACVSgmlSgkm8M+bpKSTKCurEKM0vXBwsMGIEe4I2RQvWiaPZcAjU0p9lWo17VpLSsqQmXkWAHD79l3k5l6Cg4ONoEyelEoFGjY0g0IhR6NG5igqui44Uyr7AEA6tfL8bNUHjZ4mY8Bt4Onm5gYrK6tH5slkMlRWVgIAKioq0KJFC1Heq5WDDQoKSut+vlx4Fa0crI0yl1etQO3XzGnp61FcEo3EhAykpOSKkis2xoD4/Z8jOeV7vPPucNFyV38zE4GBQdBojGXz0h8p9VWqlWp9oE0bO7i4OCE5WZx9ldj7lqKi6/jm6xicz49AQeFW3CqvRMJPGSJUKj7e60pMUuyrRDx6vZ3SwoULMX36dHz55ZfQaDTYtm2bPt/+uafRaNCzxwxYWVlg585l6NKlLU6dyjd0WX/jMSAARUXXYWtrhf0HViIvtwBJSb8JyvT2dkdp6U1kZPwOD4+XRaqUEMKLhUUD7IheioCAdaiouCNKptj7lqZNLeHj2wcvdpiKmzdvY9uORZj05iBs3XJIlHoJeUDoN5RSoteLi6KiorBgwQL88ssvWLBgARYtWiRKblHhNbRu/efRU0cHWxQVCv86hEcur1ofVl5eicOHs+Hl1VPUXLE8+Krq6tVyxMX+Cje3zoIz+/btAh+fPjh7LgJbti7CwIEuCAv/WHCuVEipr1KtVKtSqUB0zDJEbU1E7O6jgvMeEHvfMniIK/Lzr+DatXJUV9cgdvev6NPHWYxSRaePzxaxSKmvEvHpdeC5e/duDB06FAAwfPhw5OTkiJKbmpoHJycHtG1rD5VKiTfGv4o9e44ZZS6vWm1srGBlZQEAaNDADEOGdEdeXoHgXLE1amQOS8uGdf/29OwhylHZRYtC0LbNJDh1mIw3J32Gn3/OwltTvhScKxVS6qtUK9UaFDwXp09fxJo1OwVnPcBj31JwqRS93DujYUNzAMCgQS7IPW18+1WA37riQUp9VR8YAA3Tz2QM9PpVe4sWLZCSkgJ3d3ecOHECbdu2FSW3pkaD2f5rsS9+JRQKOTaH7odafdEoc3nV2rJlc4SEzodCIYdcLkNM9BHs3ZssKDNyy0J4eHSDjY0V8i9GYfnyMISG7BeUaWfXDDE7lwKoPeqxLepnHDiQJiiTJx7LgEemlPoq1Wratfbr1xWTJ3siJ+c80tLXAwCWLA5BfHyKoFwe+5aUlDzs2pmElLTvUV1dg+yscwgKEn7xolT2AVKqlVf7ifhkjNOJBQEBAUhJScGNGzdgbW0NPz8/tGvXDp9//jmqq6thbm6OpUuXomvXrlqzIiIiMWXKVB5lSoSMUy6PVa/gkAnwuR7PSP78I4TUg/j7FhmnfStDNZdcIj61+iScnQ17+sSPUQkI9ftFL+/1adIkg7eX2xHP1atXP3b+rl27eL0lIYQQQggxYnr9qp0QQggh5P/au/OwKsr+f+DvswAuR1H2RzQVDeLRBE00V5IEFMSFLCpXzMcnM9FwCcEyLZf8WlqUuyiikAoGoYCGe2ogCqKyFAQqIqCQpIiy3b8/+HnKR1Bg5j6c8XxeXnNdcjjnPZ+5Z+45w33OzJDH6dBJ7c/PLTMJIYQQQoh2oxFPQgghhJBmwgDU6NA5BzTiSQghhBBCNIJGPAkhhBBCmpEufcdTQgeeYl9Ko1rkPJ54DUyL3wb8Lk+iQ72SEEnjsw+QycS/nJK+0kj0TAB4WHmLQ6qU3rMIqR991E4IIYQQQjRCQiOehBBCCCHPHzq5iBBCCCGEEJHRiCchhBBCSDNhDKjRobOLaMSTEEIIIYRoxHNz4JmVHYzklI1IOr8evyYEipK5Zet85N/ch5SLW0TJ45X5iNht0LGjKeLj1yD10jZcTN2K2bPHiVAlMGfOOKSkbkLyxY0I2e0HAwM9wZm82pVHLq92dXV1wJW07cjIDMbChW+LksmrXXnUyiOTV65U2pXn/koul+Nc0gZE/fR5kzO2bJmLG/mhSE5Zr35s1ZfTcOnyJly48D32hS+GoWHrRmUaGOjhxKml+DVxOc5dWIWATzwf+/3/fTUJhbe3NrnmRwwNW2PP3sW4fGUrLl3egldftRWcKaU+IKV+xR/T2D9twO3Ac9GiRRgwYABGjRqlfiwjIwNeXl7w8PDA+++/j3v37ok6z+GvL0TfVz7Aq/1ni5K3M/gQ3N0WiZLFM/OfxGyDqqpqLFiwEb1efg+DBs7GzA/GwNb2BUGZHToYY9bsMXi132z0tnsfCoUcXm+/JrhWXu3KI5dHu8rlcnwbOBuj3P3xcs/34PX2MMGZAJ/l51Err+XX9Xblub/y8RmHjIxrgjKCd8ZjlPsnjz0WH58Me7uZ6NNnFn7//QY+9nurUZkPH1bCbcQKvNovAAP6BcDZuRcc+nUDAPTu0xXt2zfuQLY+a9fNxKFDSejZYzr69J6J9HRhbSGlPiClfkXEx+3A09PTE1u3Pv5XYUBAAObNm4fo6GgMHz78id9rm1OnLqGk5K7WZ/JSUFCC5OQsAMC9e+XIyLgGS0sTwblKpQItW+pDoZCjVSsD5OcXC87k1a48cnm0a79+NsjOzkdOzk1UVlZh757jGD16kOBaeSw/j1p5Lb+utyuvfmVpaQI3t/4I2hYrKOeXU5efqC/+52RUV9cAABJ+zUDHJvStsrKHAAA9PQX09JRgDJDLZVi+8h0s9v9BUM0A0LZtKwwZ8jKCtsUBACorq1BaWiYoU0p9QEr9SlNqNDRpA24Hng4ODjA0NHzssdzcXDg4OAAABg0ahMOHD4s2P8aA2LgVSEj8DtP/M1K0XCnh2QadO5vD3r47EhIyBOXk5xdj7Vfh+CM3BNdvhOKv0jLE/3xBpCqlR6x27WBpguvXi9Q/5924hQ6WxkLL44JHrbyWX9fblZev186En98W1NTwfSuc6u2CuLikRr9OLpfhbMJy5F5fj6NHLiHpXDben+mCmAMXUFBwR3BdXbta4PatUmwLmodzSd9j0+a5aNXKQFCmlPqAlLZVIj6NfsfzxRdfxJEjRwAAcXFxuHnzpmjZjkN90c/hQ4xyD8DMmaMxZEhP0bKlglcbtG7dAnv3LYGv73rcvXtfUFa7dip4jB6AF7tNxQsdJ6BV6xZ4d4KTKHVKjZjtSohUuLv3R1HRHVy48DvX+fgt8kJVVTVCQ481+rU1NQwD+gfAupsPXnHohkGDbTDujX7YsF6cwRKlUoHefbpj08YDcOg7C2VlD/Dxx16iZBPpYai9jqcmJm2g0QPP5cuXIzQ0FJ6enigrK4O+vr5o2Y8+rr11qxRRkafh4PCSaNlSwaMNlEoF9oV/hrDQI4j88RfBea8P743c3ELcvl2KqqpqRP54GgMGCP9SvdSI3a75N26jUycz9c8dLU2Rf0P4Vxh44FErr+XX9XblYeDAHvDwGICs7BDsDg3AsGH2CN75sajzmDx5ONzd+2HypP8TlFNaeh8nT6RhqOO/0c3KHJfSvkJa5lq0aqWP1CtfNTk3L+828vJuITExEwCwP+IX9O7TXVCtUuoDUtlWCR8aPfDs1q0bgoKCsH//fri7u6NTp06i5LZqZQCVqqX6/87Or+DKlVxRsqWCVxts2Tof6elXsW5dhOAsALh+rQj9+r+Eli1rP1ZycrJHRvp1UbKlROx2PXcuE927W6JLFwvo6SnxltdriI4+I0q22HjUymv5db1deQgICEKXzu+ie7dJmPDuchw7loIpk78ULd/F9RXMmz8e48YuRXn5w0a/3sSkDQwNWwEAWrTQg9PrLyP5Qg6sunyIf9t8hH/bfIT79yvQq8e8JtdYWPgn8q7fhrV1RwC1+8H0NGEnF0mpD0hlW9UkxphGJm2g0QvIFxcXw9jYGDU1NdiwYQPeflucSyiYm7dHeMQSALUjST+EHcOhQ43/Xs//2rXbH46OdjAxMUTu1TAsXRqM7UFxWpcJ8GmDQYN6YtIkZ6Sm/oGk8xsBAJ8sDkJsbGKTMxMTM7E/4hQSk75DVVU1LqZkY8sWYScYAPzalUcuj3atrq7BHJ9AxMSugkIhx47tcUhLuyqoToDP8vOoldfy63q78upXYgnZtRCOjr1gYtIWObk7sWzpLiz8+C0YGOghLm45ACAhIROzZn3X4EwLi3bYvPW/UCjkkMtliIhIQFxsiui1z5nzPXaGfAx9fSVycgrw3rSmj6AC0uoDUupXRHwyxukQ2NfXF4mJifjzzz9hbGyM2bNn4/79+wgNDQUAODs7Y968eZDJZM/MCgnZhcmT3xO5wmqR83hScMoVvw1knP6WYajikksIEduz9+lNSpUJv97v/9JXGomeCQAPK29xSJXSe5Z0pKVdgq1t837dKyr0Z3z7wc8amdd3Z72bfXm5jXh+/fXXdT4+ZcoUXrMkhBBCCCFajO7VTgghhBDSTB6d1a4rnptbZhJCCCGEEO1GB56EEEIIIUQj6KN2QgghhJBmw8C05oaW/NGIJyGEEEII0Qga8SSEEEIIaUa6dHKRRA48ZZDJxL2WJWN0TTSZTLxblj7CWIXomfzwuN4gr52HlGolRHw89i0PKwtFzwQAfT1T0TMrKotEz5QWXtezJpomkQNPQgghhJDnD11OiRBCCCGEEA5oxJMQQgghpNkw1NBZ7YQQQgghhIhLsgeeW7bMxY38UCSnrFc/turLabh0eRMuXPge+8IXw9CwtaB5dOxoivj4NUi9tA0XU7di9uxxQsvGlq3zkX9zH1IubhGc9b8MDVtjz97FuHxlKy5d3oJXX7VtfH11tOsbbwxGysUNeFhxAK+88qLgOl1dHXAlbTsyMoOxcOHbgvN45RoY6OHs2UCcv7ARF1O3YMmSySJUyWcb4FWrVNYVwK9v8aiVRyaP/RXAbxuQy+U4l7QBUT99LlqmNtdqYKCHk6c+x6+JK5F0YTUWf/IGAOD9911w6crXuP8gFMbGbQTVKaX+yqvWrOxgJKdsRNL59fg1IVC0XN6YrEYjk1ZgnOTn57OJEyeykSNHMjc3N7Zjxw7GGGN//vknmzp1KnN2dmZTp05ld+7ceWbWzp27mEzW8rFpyFAH1qdPT5Zycbv6MWeXgUypbM1kspZs5aoP2MpVHzzxukcToHjmZGFhynr3tmGAgqlUKpaWHsJsbbs26LX1TUOG2LPevW1YysWgRrxOv0FT0HZ/Nu290QzQZ3p6rZihYfunPr+udqmrXW1trZmNTXd25Og3rG/fXvW2aUPaVS7XY5m/7WZdu3ZkenoG7PyFrYLbtOm5ymdOrVurGKBkSqUBO3NmPevf/+VnvIbXNqD5WrVrXfFqV83Xymv5eeyvePUrQMnmzvViIbs+YVE/rWzga6RTq77ev+qc2rfrwvT1/sVatezIzp7dwgYNGsIc+r7Kund/mWX/sZf9y8K63tc+T/21aZkNex/Myg5jxsYmDX5+Wloar8OgBtu/O4692u5DjUzasLzcRjwVCgX8/PwQExODPXv2IDQ0FFlZWdi8eTMGDBiAw4cPY8CAAdi8eXOT8n85dRklJXcfeyz+52RUV9ce0Sf8moGOliaClqGgoATJyVkAgHv3ypGRcQ2WAjNPnbr0RN1iaNu2FYYMeRlB2+IAAJWVVSgtLWt0Tl3tmpFxHb/9dkOUOvv1s0F2dj5ycm6isrIKe/ccx+jRg7Q2t6zsAQBAT08JpZ4SjDHBmby2AbFrldq64tGuPGrltfw89le8arW0NIGbW38EbYsVnPWIFGotK3sIANDTU0BPTwEwhosXr+La1duCs6XUX3nVKlUMtWe1a2LSBtwOPM3MzNCjRw8AgEqlgpWVFQoLC3HkyBGMHTsWADB27FjEx8dzmf9UbxfExSWJlte5szns7bsjISFDtEwxde1qgdu3SrEtaB7OJX2PTZvnolUrg+Yu6wkdLE1w/frf16PLu3ELHSyNtTZXLpcj6fxG3CzYhyPxF5CYqJ3rHxC/VqmtKx541KqJ5Rdrf8Wr1q/XzoSf3xbU1Ij30Z8UapXLZfg1YQWuXt+II0cu4dy5bMGZj0ipv/LsA4wBsXErkJD4Hab/Z6QomURcGvmOZ15eHtLT02FnZ4fi4mKYmZkBAExNTVFcXCz6/PwWeaGqqhqhocdEyWvdugX27lsCX9/1uHv3viiZYlMqFejdpzs2bTwAh76zUFb2AB9/7NXcZUleTU0N+r7yPjq/8A4cHGzQo0eX5i6pXlKqlfCj7fsrd/f+KCq6gwsXfm/uUp5J7Fprahhe7e+PF7t9iL4O3fDvf3cUJZf8zXGoL/o5fIhR7gGYOXM0hgzp2dwlNUDtWe2a+KcNuB94lpWVwcfHB/7+/lCpVI/9TiaTQSYT944skycPh7t7P0ye9H+i5CmVCuwL/wxhoUcQ+eMvomTykJd3G3l5t5CYmAkA2B/xC3r36d7MVT0p/8ZtdOpkpv65o6Up8m8I/+ODV+4jpaVlOH78Ilxd+4qWyYtYtUp1XYmJR608l1/s/RWPWgcO7AEPjwHIyg7B7tAADBtmj+CdHwstVVK1lpbex8kTaXB2sROc9YiU+ivPPpCfX5tz61YpoiJPw8HhJVFyiXi4HnhWVlbCx8cHHh4ecHFxAQAYGxujqKh2iL2oqAhGRkaizc/F9RXMmz8e48YuRXn5Q1Eyt2ydj/T0q1i3LkKUPF4KC/9E3vXbsLau/Qvaycke6WnXmrmqJ507l4nu3S3RpYsF9PSUeMvrNURHn9HKXBMTQ/WVEVq00Mfw4X2QmXldcK088KhVSuuKFx618lx+sfdXPGoNCAhCl87vonu3SZjw7nIcO5aCKZO/fO5rNTFpA0PDVgCAFi304PT6y/gtM19Qff8kpf7Kq9ZWrQygUrVU/9/Z+RVcuZIrOFcTNPMNT+0Y8eR2AXnGGAICAmBlZQVvb2/1405OToiMjMSMGTMQGRmJ119/vUn5IbsWwtGxF0xM2iIndyeWLd2FhR+/BQMDPcTFLQcAJCRkYtas75q8DIMG9cSkSc5ITf0DSec3AgA+WRyE2NjEJmfu2u0PR0c7mJgYIvdqGJYuDcb2oLgm5/3TnDnfY2fIx9DXVyInpwDvTfuq0Rl1tWtJyV2s+2YmTE0NEfXTZ7h48Q+4u33SpBqrq2swxycQMbGroFDIsWN7HNLSrjYpi3fuv/5lhKDtC6FQyCGXyxC+7yQOHkwQXCuPbYBHrVJaVwCfduVRK6/l57G/4lUrD9peq4VFO2zZOhPy/99H90f8itjYZMz8wBW+vqNgbtEOiedW4dChFHwws/GXBJNSf+VVq7l5e4RHLAFQO/r/Q9gxHDok3rkeRBwyJsZpunVISkrChAkTYG1tDbm8dmDV19cXvXr1wty5c3Hz5k106NAB69atQ7t27Z6aFRKyG1Om/EfU+hirEDWPLwWXVJlM/Fxptau4X/OoxeusQSnVSnQbj20V4LO98qlVX89U9MyKyqJnP+m5xud9MC0tBba2jb/mtZj2h8ZhxaxIjcwr5MycZl9ebiOeffv2RWZmZp2/Cw4O5jVbQgghhBDJYDKgRlsu7q4Bkr1zESGEEEIIkRZuI56EEEIIIeRZmNZc6kgTaMSTEEIIIYRoBI14EkIIIYQ0k9pbZtKIJyGEEEIIIaKiA09CCCGEkGbDtOYC8osWLcKAAQMwatQo9WN37tyBt7c3XFxc4O3tjdLS0tqqGcMXX3wBZ2dneHh44MqVKw1aWol81M4kdn1IsfEZgmdMd4b26yal61hKqVai22hbpWtu8lDd3AXoBE9PT0ycOBEff/z3rWE3b96MAQMGYMaMGdi8eTM2b96MBQsW4OTJk8jNzcXhw4dx8eJFfPbZZ9i3b98z50EjnoQQQgghzagG1RqZnsXBwQGGhoaPPXbkyBGMHTsWADB27FjEx8c/9rhMJoO9vT3++usv9S3Rn4YOPAkhhBBCSJ2Ki4thZmYGADA1NUVxcTEAoLCwEBYWFurnWVhYoLCw8Jl5EvmonRBCCCHk+cP+/3c8NaGkpASenp7qn728vODl5dXg18tkMshkwm41SweehBBCCCE6wMjICPv372/Ua4yNjVFUVAQzMzMUFRXByMgIAGBubo6CggL18woKCmBubv7MPPqonRBCCCGkGdXIajQyNYWTkxMiIyMBAJGRkXj99dcfe5wxhpSUFLRp00b9kfzTPDcHnlu2zkf+zX1IubhF1FxXVwdcSduOjMxgLFz4ttZmGhjo4ezZQJy/sBEXU7dgyZLJWpkJ8Fn+jh1NER+/BqmXtuFi6lbMnj1OlFwetfLKlUomr1wemVLar/DKFTuTV18F+LWrXC7HuaQNiPrpc9EypbCuAGn1AV7rX5f4+vri7bffRk5ODoYOHYp9+/ZhxowZOH36NFxcXHDmzBnMmDEDAODo6IhOnTrB2dkZn3zyCZYsWdKwmTBO8vPz2cSJE9nIkSOZm5sb27FjB2OMsZiYGObm5sZsbGxYampqg7J27gxhgOKp05Ah9qx3bxuWcjHomc9t6CSX67HM33azrl07Mj09A3b+wlZma9u1GTKVDZpat1YxQMmUSgN25sx61r//yw1+rXiZmm9TQMEsLExZ7942DFAwlUrF0tJDmmldSWm7ouV/1iSV/YqU2pVHX216rQ3bD86d68VCdn3Con5a2cDXPB/rSkp9oKmZaWlpvA6DGmxvaDR70ehNjUzasLzcRjwVCgX8/PwQExODPXv2IDQ0FFlZWbC2tkZgYCAcHBxEnd+pU5dQUnJX1Mx+/WyQnZ2PnJybqKyswt49xzF69CCty3ykrOwBAEBPTwmlnhKMMa3L5LX8BQUlSE7OAgDcu1eOjIxrsLQ00cpapbJd6fryA9LZr/DK5ZHJo6/yqhUALC1N4ObWH0HbYgVnPSKVdQVIpw/wfG8l4uJ24GlmZoYePXoAAFQqFaysrFBYWIhu3brBysqK12xF1cHSBNev/31Nqrwbt9DB0ljrMh+Ry+VIOr8RNwv24Uj8BSQmZmhdJs/lf6RzZ3PY23dHQoJ21iqV7UrXl58Xate/idVXAX61fr12Jvz8tqCmRryzjqW4rsSk68uv6zTyHc+8vDykp6fDzs5OE7PTWTU1Nej7yvvo/MI7cHCwQY8eXbQyk6fWrVtg774l8PVdj7t37zd3OYSQekihr7q790dR0R1cuPB7c5dCnmMM0JpbZmoC9wPPsrIy+Pj4wN/fHyqVivfsRJV/4zY6dfr7DK2OlqbIv1GsdZn/q7S0DMePX4Sra1+ty+S5/EqlAvvCP0NY6BFE/viL4DxetUplu9L15eeF2lX8vgrwqXXgwB7w8BiArOwQ7A4NwLBh9gje+fGzX/gMUlpXPOj68us6rgeelZWV8PHxgYeHB1xcXHjOiotz5zLRvbslunSxgJ6eEm95vYbo6DNalwkAJiaGMDRsDQBo0UIfw4f3QWbmda3L5LX8QO3Zl+npV7FuXYQoebxqlcp2pevLzwu1q/h9FeBTa0BAELp0fhfdu03ChHeX49ixFEyZ/KVW1qrrfUBKy/8khhpWrZFJG3C7gDxjDAEBAbCysoK3tzev2ajt2u0PR0c7mJgYIvdqGJYuDcb2oDhBmdXVNZjjE4iY2FVQKOTYsT0OaWlXtS4TAP71LyMEbV8IhUIOuVyG8H0ncfBggtZl8lr+QYN6YtIkZ6Sm/oGk8xsBAJ8sDkJsbKLW1SqV7UrXlx+Qzn6FVy6PTB59lVetvEhlXQHS6QNSWv+6TsbEOPW5DklJSZgwYQKsra0hl9cOrPr6+qKiogKff/45SkpK0LZtW9ja2mLbtm1PzQoJ2YXJk6fyKFMihN2eSrO4bE6EEMIBr30r7QelIi3tEmxtbZu1hr1h0fCbtVUj8zp4elWzLy+3Ec++ffsiMzOzzt85Ozvzmi0hhBBCCNFSdK92QgghhJBmw8CgHd+/1ITn5paZhBBCCCFEu9GIJyGEEEJIM2EAarTkGpuaQCOehBBCCCFEI2jEkxBCCCGkGWnLXYU0gUY8CSGEEEKIRkhoxFPs663RddaoDXhcw0/X25QQKeE19qI7ZygTETAGpiV3FdIEGvEkhBBCCCEaIaERT0IIIYSQ5wud1U4IIYQQQggHdOBJCCGEEEI04rk48DQw0MPZs4E4f2EjLqZuwZIlk0XLdnV1wJW07cjIDMbChW9rbSYAyOVynEvagKifPhclb8vW+ci/uQ8pF7eIkvcIr+UXO5e2Kz6Zur5d8crklSt2ZseOpoiPX4PUS9twMXUrZs8eJ0KVtXgsf1Z2MJJTNiLp/Hr8mhAoSiYgjXXFM1cqmZpRe8tMTUxagXGSn5/PJk6cyEaOHMnc3NzYjh07GGOMrVq1irm6urJRo0axDz74gJWWlj4za+fOEAYonzq1bq1igJIplQbszJn1rH//l5/xGsUzJ7lcj2X+tpt17dqR6ekZsPMXtjJb264Neq24mU9f9kfT3LleLGTXJyzqp5UNfM3Tax0yxJ717m3DUi4GCVpm3m3Ks12f7+2qedbV879dUa1PmywsTFnv3jYMUDCVSsXS0kOacfn1nzllZYcxY2OTBj337+n5WFdSqrWpmWlpabwOgxosbPeP7F/tHDUyacPychvxVCgU8PPzQ0xMDPbs2YPQ0FBkZWVh0KBBOHDgAKKjo9GlSxds2rRJlPmVlT0AAOjpKaHUU4IxJjizXz8bZGfnIyfnJiorq7B3z3GMHj1I6zIBwNLSBG5u/RG0LVZw1iOnTl1CScld0fIAfsvPK1eXtytebarr25Wu11pQUILk5CwAwL175cjIuAZLSxNBmbxq5UUq60pKtUpp/deFsRqNTNqA24GnmZkZevToAQBQqVSwsrJCYWEhBg8eDKWy9mR6e3t7FBQUiDI/uVyOpPMbcbNgH47EX0BiYobgzA6WJrh+vUj9c96NW+hgaax1mQDw9dqZ8PPbgpoa7diw6sNr+Xnl6vJ2xatNeZDSdkW1/q1zZ3PY23dHQoJ29isAYAyIjVuBhMTvMP0/IwXnAdJaV1KpVUr7K12nke945uXlIT09HXZ2do89HhERgaFDh4oyj5qaGvR95X10fuEdODjYoEePLqLkSoG7e38UFd3BhQu/N3cpzx1d3q4I4al16xbYu28JfH3X4+7d+81dTr0ch/qin8OHGOUegJkzR2PIkJ7NXRJ57jDUaOifNuB+4FlWVgYfHx/4+/tDpVKpH9+wYQMUCgVGjx4t6vxKS8tw/PhFuLr2FZyVf+M2OnUyU//c0dIU+TeKtS5z4MAe8PAYgKzsEOwODcCwYfYI3vmxoExeeCw/z9xHdHG74t2mYpLSdkW1AkqlAvvCP0NY6BFE/viL4DyAY7vm12bculWKqMjTcHB4SXimhNaVVGqV0v5K13E98KysrISPjw88PDzg4uKifnz//v04fvw41qxZA5lM+G0LTUwMYWjYGgDQooU+hg/vg8zM64Jzz53LRPfulujSxQJ6ekq85fUaoqPPaF1mQEAQunR+F927TcKEd5fj2LEUTJn8paBMXngsP69cXd+ueK0rHqS0XVGttVc2SE+/inXrIgRnPcKj1latDKBStVT/39n5FVy5kquVter6diWl/dX/YgAYq9bIpA243bmIMYaAgABYWVnB29tb/fjJkyexdetW7Nq1Cy1bthRlXv/6lxGCti+EQiGHXC5D+L6TOHgwQXBudXUN5vgEIiZ2FRQKOXZsj0Na2lWty+Rl125/ODrawcTEELlXw7B0aTC2B8UJyuS1/DxydX274rWudH270vVaBw3qiUmTnJGa+geSzm8EAHyyOAixsYlaV6u5eXuERywBUDtK+0PYMRw6lCQok1etur5dSem9VdfJmBin6dYhKSkJEyZMgLW1NeTy2oFVX19ffPHFF6ioqEC7du0AAHZ2dli2bNlTs0JCdmHyZO+nPqfxuCw2J8JHhesmpTbggUe76nqbEiIlCk652jGyRJ4tLe0SbG1tm7WGsND98JmpmU8pT/66o9mXl9uIZ9++fZGZmfnE446OjrxmSQghhBBCtBi3A09CCCGEEPJs2nKNTU14Lm6ZSQghhBBCtB+NeBJCCCGENBumPfdR1wAa8SSEEEIIIRpBI56EEEIIIc2IvuNJCCGEEEKIyCQ04in2MbKUvk/B59qQMpm+6JmMVYieyQ+Pv7uktF0Rouuov/JA7y3kaSR04EkIIYQQ8nxhjNFH7YQQQgghhIiNRjwJIYQQQppRDWjEkxBCCCGEEFHRiCchhBBCSLOh73hKUlZ2MJJTNiLp/Hr8mhAoWq6rqwOupG1HRmYwFi58W3Delq3zkX9zH1IubhGhOvFzt2yZixv5oUhOWa9+bNWX03Dp8iZcuPA99oUvhqFha0HzELtNeeZKZbuSUiavXF59S9fbVSqZUlr/vHK1uV2l/N5CRMY4yc/PZxMnTmQjR45kbm5ubMeOHYwxxtauXctGjRrFRo8ezby9vVlBQcEzs3buDGGA/lOnrOwwZmxs8szn/T0pnjnJ5Xos87fdrGvXjkxPz4Cdv7CV2dp2bdBr65uGDLFnvXvbsJSLQYJyxMiVyVo+MQ0Z6sD69OnJUi5uVz/m7DKQKZWtmUzWkq1c9QFbueqDOl8rk7VsljZteu6ztxOpbFdSyeSZy6Nv6Xq7SiVTSutfF9pVSu8taWlpvA6DGmz3rr1M1dJaI5M2LC+3EU+FQgE/Pz/ExMRgz549CA0NRVZWFqZPn47o6GhERUXhtddew/fff8+rBMH69bNBdnY+cnJuorKyCnv3HMfo0YMEZZ46dQklJXdFqlD83F9OXX4iJ/7nZFRX134MkPBrBjpamjQ5n0eb8szlgUetUsnkmcujb+l6u0olE5DO+ueVq+3tKtX3FiI+bgeeZmZm6NGjBwBApVLBysoKhYWFUKlU6ueUl5dDJpOJMj/GgNi4FUhI/A7T/zNSlMwOlia4fr1I/XPejVvoYGksSrZUTfV2QVxcUpNfz6tNeeVKZbuSSibPXB50vV2lksmLrq8rTdLW9xZNYGBgqNHIpA00cnJRXl4e0tPTYWdnBwBYu3YtIiMj0aZNG+zcuVOUeTgO9UV+fjFMTQ0Rd2gVMjOu49Spy6Jkk1p+i7xQVVWN0NBjzV2KxtB2RQghfOnie4su435yUVlZGXx8fODv768e7fzoo49w4sQJeHh4YNeuXaLMJz+/GABw61YpoiJPw8HhJeGZN26jUycz9c8dLU2Rf6NYcK4UTZ48HO7u/TB50v8JyuHVptxyJbJdSSWTZy4Put6uUsnkRdfXlSZo+3uLpjBWo5FJG3A98KysrISPjw88PDzg4uLyxO89PDxw+PBhwfNp1coAKlVL9f+dnV/BlSu5gnPPnctE9+6W6NLFAnp6Srzl9Rqio88IzpUaF9dXMG/+eIwbuxTl5Q8FZfFqUx65UtqupJLJM5cHXW9XqWTyouvrijcpvLcQ8XH7qJ0xhoCAAFhZWcHb21v9eG5uLrp06QIAOHLkCKysrATPy9y8PcIjlgAAlEoFfgg7hkOHmv5dkUeqq2swxycQMbGroFDIsWN7HNLSrgrK3LXbH46OdjAxMUTu1TAsXRqM7UFxgmsVKzdk10I4OvaCiUlb5OTuxLKlu7Dw47dgYKCHuLjlAICEhEzMmvVdk+rk0aa8cqW0XUklk2cuj76l6+0qlUxAOuufV662t6tU31s0RVtGIzVBxhhjPIKTkpIwYcIEWFtbQy6vHVj19fVFeHg4cnJyIJPJYGlpiaVLl8Lc3PypWSEhuzB58nsiV1gtcp70yGT6omcyViF6Jj8KDpm0XRFCdJuU3lvS0i7B1taWS3ZD7d69F/9572ONzOt8ckyzLy+3Ec++ffsiMzPziccdHR15zZIQQgghRGKY1pxxrgnPzZ2LCCGEEEKIdqMDT0IIIYQQohEauY4nIYQQQgipA9Otk4toxJMQQgghhGgEjXgSQgghhDQTBkYjnoQQQgghhIhNQiOedH1EsUnrmps80DZFCOFBt68RzOO9pb2ql+iZ2oVGPAkhhBBCCBGVhEY8CSGEEEKeP/QdT0IIIYQQQkRGI56EEEIIIc2GbpkpSVu2zkf+zX1IubhF1FxXVwdcSduOjMxgLFz4tuA8qdQJAB07miI+fg1SL23DxdStmD17nCi5PGrllavrter68vPK5ZEppX2Lrq8rAMjKDkZyykYknV+PXxMCRcnUxXY9n7oYJ84swLFT8/Dz8Y8AAFu2T8KxU/Nw7NQ8nE9djGOn5olSMxEJ4yQ/P59NnDiRjRw5krm5ubEdO3Y89vtt27Yxa2trVlxc/MysnTtDGKB46jRkiD3r3duGpVwMeuZzGzrJ5Xos87fdrGvXjkxPz4Cdv7CV2dp2FZQplToBBbOwMGW9e9swQMFUKhVLSw8RnMurVh65ul6rri+/1GqVyr5FN9aV/jOnrOwwZmxs0qDn1k663a7tVb3rnP74I4J1e2Fwvb//dt0S9sWyj+v9fVpaGq/DoAYLCQllCoWRRiZtWF5uI54KhQJ+fn6IiYnBnj17EBoaiqysLADAzZs3cfr0aXTo0EG0+Z06dQklJXdFywOAfv1skJ2dj5ycm6isrMLePccxevQgQZlSqRMACgpKkJxcu87u3StHRsY1WFqaaGWtPHJ1vVZdX36p1SqVfQutKz6oXes2Zpwdfgy/wCWbNA23A08zMzP06NEDAKBSqWBlZYXCwkIAwMqVK7FgwQLIZDJesxdFB0sTXL9epP4578YtdLA0bsaK6qaJOjt3Noe9fXckJGQIyuFVK49cXa9V15efV65U9iuAtJZfSrUCAGNAbNwKJCR+h+n/GSk4T1fblYFhX+R/EX/iI0ya+upjvxsw0Aq3bt3DH3/cFlQvfwy112nVxNT8NHJyUV5eHtLT02FnZ4f4+HiYmZnhpZde0sSsiQhat26BvfuWwNd3Pe7evd/c5RBCiOQ5DvVFfn4xTE0NEXdoFTIzruPUqcvNXZbkjHL9DgU3S2FiosK+yPeR9VsRzp75AwAwbnxv7KfRTq3D/eSisrIy+Pj4wN/fHwqFAps2bcKcOXN4z1YU+Tduo1MnM/XPHS1NkX+juBkrqhvPOpVKBfaFf4aw0COI/PEXwXm8auWRq+u16vry88qVyn4FkNbyS6lWAMjPr825dasUUZGn4eAgbDBGV9u14GYpAOD27XuIOXAJvV95AQCgUMjh7tELkftTBNWqKYzVaGTSBlwPPCsrK+Hj4wMPDw+4uLjg2rVryMvLw5gxY+Dk5ISCggJ4enri1q1bPMtosnPnMtG9uyW6dLGAnp4Sb3m9hujoM81d1hN41rll63ykp1/FunURouTxqpVHrq7XquvLL7VaeZDS8kup1latDKBStVT/39n5FVy5kquVtWpzu7ZqpY/WKgP1/19zskZGWgEAwPE1a2T9VoSb+aWCaiXi4/ZRO2MMAQEBsLKygre3NwDAxsYGZ8+eVT/HyckJ4eHhMDIyEjy/Xbv94ehoBxMTQ+ReDcPSpcHYHhQnKLO6ugZzfAIRE7sKCoUcO7bHIS3tqk7UCQCDBvXEpEnOSE39A0nnNwIAPlkchNjYRK2rlUeurteq68svtVqlsm+hdQWYm7dHeMQSALWfKv0QdgyHDiVpZa3a3K6mZirs2DUNAKBUyrE//AKOHqk9D2HcG/bYH0Efs2sjGWOM8QhOSkrChAkTYG1tDbm8dmDV19cXjo6O6uc09MAzJGQXJk+eyqNMQgghRGQKDpnacWJIc2mv6sUl93Tibtja2nLJbqhdu3ZjypT/amRely+fa/bl5Tbi2bdvX2RmZj71OUePHuU1e0IIIYQQomXolpmEEEIIIc2FAdCSE3804bm5ZSYhhBBCCNFuNOJJCCGEENJMGGovhK8raMSTEEIIIYRoBI14EkIIIYQ0GwaAvuNJCCGEEEKIqCQx4vnyyz2RlnapucsghBBCyHPk4cOHzV1CLT6XVG+SkydPYvny5aipqcGbb76JGTNmiJoviQNPe3v75i6BEEIIIeS5Vl1djWXLlmH79u0wNzfH+PHj4eTkhO7du4s2D/qonRBCCCGkGTEN/XuW1NRUdO7cGZ06dYK+vj7c3d1x5MgRUZeVDjwJIYQQQggKCwthYWGh/tnc3ByFhYWizkMSH7UTQgghhDyPXn21P9LSUjUyr1u3biEgIED9s5eXF7y8vDQy70fowJMQQgghpJm8+OKLGpuXra0thg4dWu/vzc3NUVBQoP65sLAQ5ubmotZAH7UTQgghhBC8/PLLyM3NxfXr11FRUYGDBw/CyclJ1Hk8NweeJ0+ehKurK5ydnbF582ZRMhctWoQBAwZg1KhRouQBwM2bNzFp0iS4ubnB3d0dwcHBouQ+fPgQ48ePx+jRo+Hu7o5vv/1WlFyg9iy3sWPH4r///a8oeU5OTvDw8MCYMWPg6ekpSiYA/PXXX/Dx8cGIESMwcuRIJCcnC8r7448/MGbMGPXUp08f7NixQ5Rad+zYAXd3d4waNQq+vr6iXNIjODgYo0aNgru7u6A669ru79y5A29vb7i4uMDb2xulpaWCM2NjY+Hu7o6XXnoJly417XJpdeV++eWXGDFiBDw8PDBr1iz89ddfgjPXrVun3manTZvW6O88PW1fEhQUBBsbG5SUlDQqs77cwMBADBkyRL3dnjhxQpRaQ0JCMGLECLi7u2P16tWi1Dp37lx1nU5OThgzZozgzPT0dLz11lvq/UtqauM/wqwrNyMjA15eXvDw8MD777+Pe/fuNSqzvn2/kL5VX6aQvlVfptB+VV+ukL71rPdTIX1LVymVSnz66aeYPn063NzcMHLkSPFHZNlzoKqqir3++uvs2rVr7OHDh8zDw4P9/vvvgnMTExPZ5cuXmbu7uwhV1iosLGSXL19mjDF29+5d5uLiIkqtNTU17N69e4wxxioqKtj48eNZcnKy4FzGGAsKCmK+vr5sxowZouQNGzaMFRcXi5L1TwsXLmR79+5ljDH28OFDVlpaKlp2VVUVGzhwIMvLyxOcVVBQwIYNG8bKy8sZY4z5+PiwiIgIQZmZmZnM3d2d3b9/n1VWVrIpU6aw3NzcJmXVtd1/+eWXbNOmTYwxxjZt2sRWr14tODMrK4tlZ2eziRMnstTUVNFqPXXqFKusrGSMMbZ69WpRar179676/8HBweyTTz4RnMkYY/n5+WzatGnstddea1KfqCv322+/ZVu3bm101tMyz549y6ZMmcIePnzIGGPs9u3bouT+08qVK1lgYKDgTG9vb3b8+HHGGGPHjx9nEydOFKVWT09PlpCQwBhjbN++fWzt2rWNyqxv3y+kb9WXKaRv1ZcptF/Vlyukbz3t/VRo3yL8PBcjnrxO/3dwcIChoaEIFf7NzMwMPXr0AACoVCpYWVmJcsaYTCZD69atAQBVVVWoqqqCTCYTnFtQUIDjx49j/PjxgrN4unv3Ls6dO6euU19fH23bthUt/+zZs+jUqRMsLS1FyauursaDBw9QVVWFBw8ewMzMTFBednY2evXqhZYtW0KpVMLBwQGHDx9uUlZd2/2RI0cwduxYAMDYsWMRHx8vOLNbt26wsrJqUo1Pyx08eDCUytqvr9vb2z/2faWmZqpUKvX/y8vLG9236tuXrFy5EgsWLGhyX+Wxj6orMywsDDNmzIC+vj4AwNjYWJTcRxhjiI2NbfSnS3VlymQylJWVAajdLzSlb9WVm5ubCwcHBwDAoEGDGt2/6tv3C+lb9WUK6Vv1ZQrtV/XlCulbT3s/Fdq3CD/PxYGnJk7/5yEvLw/p6emws7MTJa+6uhpjxozBwIEDMXDgQFFyV6xYgQULFkAuF3dTee+99+Dp6Yk9e/aIkpeXlwcjIyMsWrQIY8eORUBAAO7fvy9KNgAcPHhQtK9cmJubY9q0aRg2bBgGDx4MlUqFwYMHC8q0trbG+fPn8eeff6K8vBwnT55s9BvD0xQXF6vfwE1NTVFcXCxaNk8RERFP/SJ9Y6xduxaOjo6Ijo7GnDlzBOfFx8fDzMwML730kgjVPW737t3w8PDAokWLGv21iLrk5uYiKSkJb775JiZOnNikj6+fJikpCcbGxujSpYvgLH9/f6xevRqOjo748ssv4evrK7xA1J4A8mhAIy4uDjdv3mxy1j/3/WL1LbHfT56WKbRf/W+uGH3rn5k8+xYR7rk48JSisrIy+Pj4wN/f/7G/+IRQKBSIiorCiRMnkJqait9++01Q3rFjx2BkZISePXuKUt8jYWFh+PHHH7Flyxbs3r0b586dE5xZVVWFtLQ0vPPOO4iMjETLli1F+65vRUUFjh49ihEjRoiSV1paiiNHjuDIkSM4deoUysvLERUVJSizW7dumD59Ot577z1Mnz4dL730kuh/LDwik8kkMYqwYcMGKBQKjB49WpS8jz76CCdOnICHhwd27dolKKu8vBybNm0S5QD2f73zzjv4+eefERUVBTMzM6xatUpwZnV1NUpLS7F3714sXLgQc+fOBRPxFn8HDhwQ7Q+7sLAwLFq0CCdOnMCiRYseu3SMEMuXL0doaCg8PT1RVlamHv1trKft+5vat3i8n9SXKbRf1ZUrtG/9M1OhUHDrW0Qcz8WBpyZO/xdTZWUlfHx84OHhARcXF9Hz27Zti/79++PUqVOCci5cuICjR4/CyckJvr6++PXXXzF//nzB9T1aN8bGxnB2dhZl9MTCwgIWFhbqv6BHjBiBtLQ0wblA7YlrPXr0gImJiSh5Z86cQceOHWFkZAQ9PT24uLgIPhEKAN58803s378fu3fvhqGhoSijR48YGxujqKgIAFBUVAQjIyPRsnnYv38/jh8/jjVr1oh+kOzh4dHkrzE8cu3aNeTl5alPqikoKICnpydu3boluD4TExMoFArI5XK8+eabTT5x65/Mzc3h7OwMmUyGXr16QS6X488//xScC9T+0fjzzz/Dzc1NlLwff/xRvV8dOXKkaKOz3bp1Q1BQEPbv3w93d3d06tSp0Rl17fuF9i0e7yf1ZQrtV8+qtSl9638zefYtIo7n4sBTE6f/i4UxhoCAAFhZWcHb21u03JKSEvVZhg8ePMCZM2cEf39u3rx5OHnyJI4ePYqvv/4ar776KtasWSMo8/79++qzQe/fv4/Tp0+LcsacqakpLCws8McffwCo/U5mt27dBOcCtR+zu7u7i5IFAB06dMDFixdRXl4OxphotT76iC4/Px+HDx+Gh4eH4MxHnJycEBkZCQCIjIzE66+/Llq22E6ePImtW7diw4YNaNmypSiZubm56v8fOXJEcN+ysbHB2bNncfToURw9ehQWFhbYv38/TE1NBVYK9UEMUPtxvhj9a/jw4UhISAAA5OTkoLKyEu3btxecC0C9r/rn16WEMDMzQ2JiIgDg119/Fe0PsEf9q6amBhs2bMDbb7/dqNfXt+8X0rd4vJ/Ulym0X9WXK6Rv1ZXJs28RcciYmJ+XNKMTJ05gxYoVqK6uxhtvvIGZM2cKzvT19UViYiL+/PNPGBsbY/bs2XjzzTcFZSYlJWHChAmwtrZWfxTq6+sLR0dHQbkZGRnw8/NDdXU1GGMYMWIEPvzwQ0GZ/5SQkICgoCBs2rRJUM7169cxa9YsALUf340aNUqUdQXUXkYlICAAlZWV6NSpE1auXCn4xIv79+9j2LBhiI+PR5s2bUSpEwC+/fZbxMTEQKlUwtbWFsuXL2/yR3ePvPvuu7hz5w6USqX6cjBNUdd2P3z4cMydOxc3b95Ehw4dsG7dOrRr105QZrt27fD555+jpKQEbdu2ha2tLbZt2ya41s2bN6OiokJdn52dHZYtWyYo8+TJk8jJyYFMJoOlpSWWLl3aqE9VnrUvcXJyQnh4eKNHu+rKTUxMREZGBgDA0tISy5Yta9QJNnVljhkzBv7+/sjIyICenh4WLlzY6O2rvjbw8/ODnZ0d3nnnnUbl1ZfZtWtXrFixAlVVVTAwMMCSJUsa/XWhunLv37+P0NBQAICzszPmzZvXqFG/+vb9vXr1anLfqi+zoqKiyX2rvswvvvhCUL+qLzc8PLzJfash76dN7VuEn+fmwJMQQgghhGi35+KjdkIIIYQQov3owJMQQgghhGgEHXgSQgghhBCNoANPQgghhBCiEXTgSQghhBBCNIIOPAmRiEmTJj1xU4AdO3ZgyZIlDc7o3bu32GVpRF5eHqKjoxv8/D///BNjxozBmDFjMGjQIAwZMkT9c0VFRYNzwsLC1NdYbAgbG5vH7hS0bds2BAYGNvj1hBDyvFM2dwGEkIYZNWoUYmJiMGTIEPVjMTExWLBgQYNeX11dLWj+VVVVUCqbZ5dx48YNHDhwoMEXxW/fvr36NqSBgYFo1aoV3nvvvUbPt7HXltTX18fhw4cxY8YMrb5uYHV1NRQKRXOXQQjRQXTgSYhEuLq6Yt26daioqIC+vj7y8vJQVFSEvn374sCBA9i0aRMYY3B0dFQfjPbu3RteXl44c+YMPv30UwDA2rVrcezYMbRo0QLr16+HiYkJSkpKsGTJEuTn5wMA/P398corryAwMBDXrl3D9evX0aFDByxevBjz5s1DUVER7O3tcebMGURERMDIyAhRUVEICQlBZWUl7Ozs1COxAQEBuHz5MmQyGd544w1MnToVe/fuxZ49e1BZWYnOnTtj9erVaNmyJa5du4b58+ejvLwcTk5O2LlzJ5KTk/HVV18hOzsbY8aMwbhx4zBp0iSsWbMGiYmJqKiowIQJExp0J5mzZ8/iyy+/RHV1NXr27ImlS5dCX18fTk5OGDFiBE6dOgUDAwN89dVX6Ny582MHrVevXsWSJUtQUlIChUKBb775Bi+88MJj+UqlEl5eXggODsZHH3302O/8/Pzw2muvYcSIEep1k5ycjISEBAQGBqJNmzb47bffMHLkSFhbW2Pnzp14+PAhvv/+e7zwwgvw8/ODgYEB0tPTUVxcjBUrViAyMhIpKSmws7NTj7T+8ssvCAwMREVFhfpGCq1bt4aTkxNGjhyJM2fOYPr06aLejYsQQhqKPmonRCLatWuHXr164eTJkwBqRztHjhyJoqIirFmzBsHBwYiMjMSlS5cQHx8PoPbOS7169cJPP/2Evn374v79+7Czs1P/vHfvXgDA8uXLMWXKFERERCAwMBCLFy9Wzzc7Oxs7duzA119/je+++w6vvvoqDh48CFdXV/WBanZ2NmJjYxEWFoaoqCjI5XJER0cjPT0dhYWFOHDgAKKjo+Hp6Qmg9s4vERER+Omnn2BlZYXw8HB1HZMnT0Z0dPRjt1CcN28e+vbti6ioKEydOhXh4eFo06YNIiIiEBERgb179+L69etPbb+HDx/Cz88Pa9euRXR0NKqrq9V3ogGANm3aIDo6GhMnTsSKFSueeP38+fMxYcIE/PTTT/jhhx/qvQXfhAkTEB0djbt37z59hf5DRkYGli5ditjYWERFRSE3Nxfh4eEYP348QkJC1M/766+/sGfPHixatAgzZ87E1KlTcfDgQfz2229IT09HSUkJNmzYgO3bt+PHH39Ez549sX37dvXr27Vrhx9//JEOOgkhzYZGPAmREHd3d8TExGD48OE4ePAgli9fjkuXLqFfv37qj3Y9PDxw7tw5DB8+HAqFAq6ururX6+npYdiwYQCAnj174vTp0wBq75edlZWlft69e/dQVlYGoPaWcy1atAAAnD9/Ht999x0AYOjQoepbkp49exaXL1/G+PHjAQAPHjyAsbExhg0bhuvXr+Pzzz+Ho6MjBg8eDAD4/fffsW7dOty9exdlZWXqx1NSUvD999+rl2P16tV1tsPp06eRmZmJQ4cOAQDu3r2Lq1evolOnTvW2XU5ODjp27IiuXbsCAMaNG4fdu3dj6tSpAGq/yvCojVeuXPnYa+/du4fCwkI4OzsDAAwMDOqdj0qlwpgxY7Bz5051uz3Lyy+/rL6t5QsvvIBBgwYBAKytrdX3SAeAYcOGQSaTwcbGBiYmJrCxsQEAdO/eHTdu3EBBQQGysrLUXxGorKyEvb29+vVubm4NqocQQnihA09CJOT111/HypUrceXKFTx48AA9e/ZEQUFBvc83MDB47Lt8enp66vtLy+Vy9fc+a2pqsHfv3joPqFq2bPnMuhhjGDduHObNm/fE76KiovDLL7/ghx9+QGxsLFauXAk/Pz+sX78eL730Evbv34/ExMRnzuN/57d48eLHvu+qTaZMmQJPT0/1CC8AKBQK1NTUAKht78rKSvXv9PX11f+Xy+Xqn/+5jv75PJlM9sRrqqqqIJfLMWjQIHz99dd11tWQdUkIITzRR+2ESEjr1q3Rv39/+Pv7qz8u7dWrF86dO4eSkhJUV1fj4MGDcHBwaFTu4MGDH/tINz09vc7n9enTB7GxsQBqv0tYWloKABgwYAAOHTqE4uJiAMCdO3dw48YNlJSUgDEGV1dXzJ07F2lpaQCAsrIymJqaorKy8rGz1e3s7HD48GEAwMGDBx9b7kcjsI/qDQsLUx+85eTk4P79+09dxq5du+LGjRu4evUqgNoD4n+206PliomJeeLsf5VKBQsLC/VXGCoqKlBeXl7vvNq1a4cRI0aov0IAAJaWlrhy5QoA4OjRo48deIrF3t4eFy5cUC/j/fv3kZOTI/p8CCGkqWjEkxCJGTVqFGbNmqUe1TIzM8O8efMwZcoU9clFw4cPb1RmQEAAli1bBg8PD1RXV6Nv375YtmzZE8/78MMP4evri59++gn29vYwNTWFSqWCkZER5s6di2nTpqGmpgZ6enr49NNP0aJFCyxatEg90ufr6wsAmDNnDt58800YGRnBzs5OfVDp7++PBQsWYMOGDRgyZAhUKhWA2ssUyeVyjB49Gp6enpg8eTJu3LgBT09PMMbQvn17rF+//qnLaGBggJUrV2LOnDnqk4v+edZ6aWkpPDw8oK+vX+eI4erVq/Hpp5/im2++gZ6eHr755punfrQ/bdo07N69W/3zW2+9hQ8++ACjR4/GkCFD0KpVq6fW2xRGRkZYuXIlfH191ZeNmjt3rvrrBYQQ0txkjDHW3EUQQqShoqICcrkcSqUSycnJ+Oyzz9SXLRJDeXk5WrRoAZlMhoMHD+LAgQPYsGGDaPn1cXJyQnh4uFZfAokQQp4HNOJJCGmw/Px8zJ07Vz2q+fnnn4uaf+XKFSxbtgyMMbRt27bOs8sJIYRIF414EkIIIYQQjaCTiwghhBBCiEbQgSchhBBCCNEIOvAkhBBCCCEaQQeehBBCCCFEI+jAkxBCCCGEaAQdeBJCCCGEEI34fzWubn/21G3EAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(figsize=(12,10))\n",
    "g=plot_confusion_matrix(estimator=cv_results_rf_topic_no[\"250-350\"][\"model\"], X=cv_results_rf_topic_no[\"250-350\"][\"X_test\"], y_true=cv_results_rf_topic_no[\"250-350\"][\"y_test\"],cmap=\"magma\", ax=ax, xticks_rotation=\"horizontal\")\n",
    "g.ax_.set_xlabel('Vorhergesagte Topic Nummer')\n",
    "g.ax_.set_ylabel('Tatsächliche Topic Nummer')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [],
   "source": [
    "multi_dict = {}\n",
    "param_dict = {}\n",
    "for key in cv_results_rf_topic_no.keys():\n",
    "    multi_dict[key] = pd.DataFrame(cv_results_rf_topic_no[key]['report_dict'])\n",
    "    param_dict[key] = cv_results_rf_topic_no[key]['best_params']\n",
    "df = pd.concat(multi_dict.values(), keys=multi_dict.keys())\n",
    "param_df = pd.DataFrame(param_dict)\n",
    "df.round(2).to_excel(\"b_collection_extracted/classification_report_rf_topic_no.xlsx\")\n",
    "param_df.to_excel(\"b_collection_extracted/classification_params_rf_topic_no.xlsx\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explorative Analyse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Binäre logistische Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "cc0ff69e6e324deca132faed9b6dd626",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10 350\n",
      "10 350 4\n",
      "12140\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.296937\n",
      "         Iterations 9\n",
      "10 350 15\n",
      "2794\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.110225\n",
      "         Iterations 9\n",
      "10 350 17\n",
      "4405\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.169429\n",
      "         Iterations 9\n"
     ]
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "\n",
    "result_dict_stats_binary_no = {}\n",
    "for min_replies, max_replies in tqdm([(10, 350)]):\n",
    "    print(min_replies, max_replies)\n",
    "    graph_features = clean_features.loc[filter_on_stats(stat_df=stat_df, min_replies=min_replies)]\n",
    "    graph_features = graph_features.dropna()\n",
    "    X = graph_features.drop(columns=[\"dominant_topic\", \"replies\"]).astype(np.float64)\n",
    "    y = graph_features.dominant_topic \n",
    "    scaler.fit(X)\n",
    "    X_transformed = scaler.transform(X)\n",
    "    key = f\"{min_replies}-{max_replies}\"\n",
    "    result_dict_stats_binary_no[key] = {}\n",
    "    try:\n",
    "        temp_dict = {}\n",
    "        for dominant_topic in [4, 15, 17] if min_replies == 10 else [2, 4, 24]:\n",
    "            y_train_binary = y == dominant_topic\n",
    "            print(min_replies, max_replies, dominant_topic)\n",
    "            print(y_train_binary.sum())\n",
    "            model = sm.Logit(endog=y_train_binary, exog=sm.add_constant(X))\n",
    "            result = model.fit()\n",
    "            temp_dict[dominant_topic] = {'model': model,\n",
    "                                'result': result,\n",
    "                                }\n",
    "        result_dict_stats_binary_no[key]['predictors'] = temp_dict\n",
    "    except Exception as e:\n",
    "        print(f\"EXCEPTION!!! {e} !!!\")\n",
    "        print(traceback.format_exc())\n",
    "        continue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4\n",
      "15\n",
      "17\n"
     ]
    }
   ],
   "source": [
    "for min_replies in [10]:\n",
    "    for key in result_dict_stats_binary_no[f\"{min_replies}-350\"][\"predictors\"].keys():\n",
    "        print(key)\n",
    "        pd.DataFrame(result_dict_stats_binary_no[f\"{min_replies}-350\"][\"predictors\"][key][\"result\"].summary().tables[1]).transpose().to_excel(f\"b_collection_extracted/log_reg_result_topic_{min_replies}_top_{key}.xlsx\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Subgroups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "subgroup_dominant_topic = graph_features_clean.drop(columns=\"replies\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [],
   "source": [
    "def search_subgroups(data, min_replies, target_attribute, target, a):\n",
    "    result_dict = {}\n",
    "    ps_target = ps.BinaryTarget(target_attribute=target_attribute, target_value=target)\n",
    "    ps_search_space = ps.create_selectors(data, ignore=[target_attribute], nbins =30)\n",
    "    ps_task = ps.SubgroupDiscoveryTask(data, ps_target, ps_search_space, qf=ps.GeneralizationAware_StandardQF(a))\n",
    "    result_dict[f\"{min_replies}_{target}_GA_{a}_SDFS\"] = ps.SimpleDFS().execute(ps_task).to_dataframe()\n",
    "    ps_task = ps.SubgroupDiscoveryTask(data, ps_target, ps_search_space, qf=ps.StandardQF(a))\n",
    "    result_dict[f\"{min_replies}_{target}_SQF_{a}_BFS\"] = ps.BestFirstSearch().execute(ps_task).to_dataframe()\n",
    "    return result_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "faf634c7c2954e8aa52aa679c2173e20",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "adding search to mp pool:   0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3a62eb4020e649a28ad9668f40ce6a6e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "retrieving subgroups:   0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "subgroup_dominant_topic = graph_features_clean.drop(columns=\"replies\")\n",
    "subgroup_combs = []\n",
    "[subgroup_combs.append(prod) for prod in product([10], [\"dominant_topic\"], [4, 15, 17], [0, 0.5, 1])]\n",
    "[subgroup_combs.append(prod) for prod in product([250], [\"dominant_topic\"], [2, 4, 24], [0, 0.5, 1])]\n",
    "pool = mp.Pool(processes=18)\n",
    "subgroup_topic_results = [pool.apply_async(search_subgroups, (subgroup_dominant_topic.loc[filter_on_stats(stat_df, min_replies=item[0])], item[0], item[1], item[2], item[3],))\n",
    "           for item in tqdm(subgroup_combs,desc=\"adding search to mp pool\")]\n",
    "subgroup_topic_results = [result.get() for result in tqdm(subgroup_topic_results, desc=\"retrieving subgroups\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "250_2_SQF_0_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "1  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND 030...      2.0        9822.0   \n",
      "2  012: [0.01:0.01[ AND 021C<0.00 AND 030T: [0.00...      3.0        9822.0   \n",
      "3  012: [0.01:0.01[ AND 021C<0.00 AND 021D: [0.00...      1.0        9822.0   \n",
      "4  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "5  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "6  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "7  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "8             012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0        9822.0   \n",
      "9  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0              1.0              0.040623           1.0              399.0   \n",
      "1              1.0              0.040623           2.0              399.0   \n",
      "2              1.0              0.040623           3.0              399.0   \n",
      "3              1.0              0.040623           1.0              399.0   \n",
      "4              1.0              0.040623           1.0              399.0   \n",
      "5              1.0              0.040623           1.0              399.0   \n",
      "6              1.0              0.040623           1.0              399.0   \n",
      "7              1.0              0.040623           1.0              399.0   \n",
      "8              1.0              0.040623           1.0              399.0   \n",
      "9              1.0              0.040623           1.0              399.0   \n",
      "\n",
      "        lift  \n",
      "0  24.616541  \n",
      "1  24.616541  \n",
      "2  24.616541  \n",
      "3  24.616541  \n",
      "4  24.616541  \n",
      "5  24.616541  \n",
      "6  24.616541  \n",
      "7  24.616541  \n",
      "8  24.616541  \n",
      "9  24.616541  \n",
      "250_2_SQF_0.5_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                                   012: [0.01:0.01[    327.0        9822.0   \n",
      "1                                  021C: [0.00:0.00[    327.0        9822.0   \n",
      "2                 degree_centralization: [0.72:0.83[    327.0        9822.0   \n",
      "3             012: [0.01:0.01[ AND 021C: [0.00:0.00[    290.0        9822.0   \n",
      "4  012: [0.01:0.01[ AND degree_centralization: [0...    282.0        9822.0   \n",
      "5                                  021U: [0.01:0.01[    327.0        9822.0   \n",
      "6  021U: [0.01:0.01[ AND degree_centralization: [...    283.0        9822.0   \n",
      "7  012: [0.01:0.01[ AND 021U: [0.01:0.01[ AND deg...    242.0        9822.0   \n",
      "8  021C: [0.00:0.00[ AND degree_centralization: [...    251.0        9822.0   \n",
      "9             012: [0.01:0.01[ AND 021U: [0.01:0.01[    245.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.581040              0.040623         190.0              399.0   \n",
      "1         0.574924              0.040623         188.0              399.0   \n",
      "2         0.571865              0.040623         187.0              399.0   \n",
      "3         0.596552              0.040623         173.0              399.0   \n",
      "4         0.602837              0.040623         170.0              399.0   \n",
      "5         0.562691              0.040623         184.0              399.0   \n",
      "6         0.597173              0.040623         169.0              399.0   \n",
      "7         0.636364              0.040623         154.0              399.0   \n",
      "8         0.625498              0.040623         157.0              399.0   \n",
      "9         0.628571              0.040623         154.0              399.0   \n",
      "\n",
      "        lift  \n",
      "0  14.303189  \n",
      "1  14.152629  \n",
      "2  14.077349  \n",
      "3  14.685040  \n",
      "4  14.839759  \n",
      "5  13.851510  \n",
      "6  14.700337  \n",
      "7  15.665072  \n",
      "8  15.397598  \n",
      "9  15.473255  \n",
      "250_2_SQF_1_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                                   012: [0.01:0.01[    327.0        9822.0   \n",
      "1                                  021C: [0.00:0.00[    327.0        9822.0   \n",
      "2                 degree_centralization: [0.72:0.83[    327.0        9822.0   \n",
      "3                                  021U: [0.01:0.01[    327.0        9822.0   \n",
      "4             012: [0.01:0.01[ AND 021C: [0.00:0.00[    290.0        9822.0   \n",
      "5  012: [0.01:0.01[ AND degree_centralization: [0...    282.0        9822.0   \n",
      "6  021U: [0.01:0.01[ AND degree_centralization: [...    283.0        9822.0   \n",
      "7  021C: [0.00:0.00[ AND degree_centralization: [...    251.0        9822.0   \n",
      "8  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND deg...    249.0        9822.0   \n",
      "9  012: [0.01:0.01[ AND 021U: [0.01:0.01[ AND deg...    242.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.581040              0.040623         190.0              399.0   \n",
      "1         0.574924              0.040623         188.0              399.0   \n",
      "2         0.571865              0.040623         187.0              399.0   \n",
      "3         0.562691              0.040623         184.0              399.0   \n",
      "4         0.596552              0.040623         173.0              399.0   \n",
      "5         0.602837              0.040623         170.0              399.0   \n",
      "6         0.597173              0.040623         169.0              399.0   \n",
      "7         0.625498              0.040623         157.0              399.0   \n",
      "8         0.622490              0.040623         155.0              399.0   \n",
      "9         0.636364              0.040623         154.0              399.0   \n",
      "\n",
      "        lift  \n",
      "0  14.303189  \n",
      "1  14.152629  \n",
      "2  14.077349  \n",
      "3  13.851510  \n",
      "4  14.685040  \n",
      "5  14.839759  \n",
      "6  14.700337  \n",
      "7  15.397598  \n",
      "8  15.323550  \n",
      "9  15.665072  \n",
      "250_4_SQF_0_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "1  012: [0.01:0.02[ AND 021U: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "2             012: [0.02:0.02[ AND 021C: [0.00:0.00[      1.0        9822.0   \n",
      "3  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "4  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "5  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "6  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "7  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "8  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "9  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0              1.0              0.392588           1.0             3856.0   \n",
      "1              1.0              0.392588           1.0             3856.0   \n",
      "2              1.0              0.392588           1.0             3856.0   \n",
      "3              1.0              0.392588           1.0             3856.0   \n",
      "4              1.0              0.392588           1.0             3856.0   \n",
      "5              1.0              0.392588           1.0             3856.0   \n",
      "6              1.0              0.392588           1.0             3856.0   \n",
      "7              1.0              0.392588           1.0             3856.0   \n",
      "8              1.0              0.392588           1.0             3856.0   \n",
      "9              1.0              0.392588           1.0             3856.0   \n",
      "\n",
      "       lift  \n",
      "0  2.547199  \n",
      "1  2.547199  \n",
      "2  2.547199  \n",
      "3  2.547199  \n",
      "4  2.547199  \n",
      "5  2.547199  \n",
      "6  2.547199  \n",
      "7  2.547199  \n",
      "8  2.547199  \n",
      "9  2.547199  \n",
      "250_4_SQF_0.5_BFS\n",
      "                          subgroup  size_sg  size_dataset  target_share_sg  \\\n",
      "0  average_clustering: [0.07:0.08[    327.0        9822.0         0.917431   \n",
      "1         average_clustering>=0.08    328.0        9822.0         0.908537   \n",
      "2  average_clustering: [0.06:0.07[    328.0        9822.0         0.896341   \n",
      "3  average_clustering: [0.05:0.06[    327.0        9822.0         0.840979   \n",
      "4  average_clustering: [0.05:0.05[    327.0        9822.0         0.834862   \n",
      "5                030T: [0.00:0.00[    329.0        9822.0         0.832827   \n",
      "6                030T: [0.00:0.00[    327.0        9822.0         0.807339   \n",
      "7  average_clustering: [0.04:0.05[    328.0        9822.0         0.804878   \n",
      "8        transitivity: [0.28:0.30[    294.0        9822.0         0.812925   \n",
      "9                030T: [0.00:0.00[    327.0        9822.0         0.776758   \n",
      "\n",
      "   target_share_dataset  positives_sg  positives_dataset      lift  \n",
      "0              0.392588         300.0             3856.0  2.336880  \n",
      "1              0.392588         298.0             3856.0  2.314224  \n",
      "2              0.392588         294.0             3856.0  2.283160  \n",
      "3              0.392588         275.0             3856.0  2.142140  \n",
      "4              0.392588         273.0             3856.0  2.126561  \n",
      "5              0.392588         274.0             3856.0  2.121376  \n",
      "6              0.392588         264.0             3856.0  2.056454  \n",
      "7              0.392588         264.0             3856.0  2.050185  \n",
      "8              0.392588         239.0             3856.0  2.070682  \n",
      "9              0.392588         254.0             3856.0  1.978558  \n",
      "250_4_SQF_1_BFS\n",
      "                          subgroup  size_sg  size_dataset  target_share_sg  \\\n",
      "0  average_clustering: [0.07:0.08[    327.0        9822.0         0.917431   \n",
      "1         average_clustering>=0.08    328.0        9822.0         0.908537   \n",
      "2  average_clustering: [0.06:0.07[    328.0        9822.0         0.896341   \n",
      "3  average_clustering: [0.05:0.06[    327.0        9822.0         0.840979   \n",
      "4                030T: [0.00:0.00[    329.0        9822.0         0.832827   \n",
      "5  average_clustering: [0.05:0.05[    327.0        9822.0         0.834862   \n",
      "6                030T: [0.00:0.00[    327.0        9822.0         0.807339   \n",
      "7  average_clustering: [0.04:0.05[    328.0        9822.0         0.804878   \n",
      "8        transitivity: [0.30:0.33[    366.0        9822.0         0.743169   \n",
      "9                030T: [0.00:0.00[    327.0        9822.0         0.776758   \n",
      "\n",
      "   target_share_dataset  positives_sg  positives_dataset      lift  \n",
      "0              0.392588         300.0             3856.0  2.336880  \n",
      "1              0.392588         298.0             3856.0  2.314224  \n",
      "2              0.392588         294.0             3856.0  2.283160  \n",
      "3              0.392588         275.0             3856.0  2.142140  \n",
      "4              0.392588         274.0             3856.0  2.121376  \n",
      "5              0.392588         273.0             3856.0  2.126561  \n",
      "6              0.392588         264.0             3856.0  2.056454  \n",
      "7              0.392588         264.0             3856.0  2.050185  \n",
      "8              0.392588         272.0             3856.0  1.893000  \n",
      "9              0.392588         254.0             3856.0  1.978558  \n",
      "250_24_SQF_0_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "1  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "2  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "3  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "4  012: [0.02:0.02[ AND 021U: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "5  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "6  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "7  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "8  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "9  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0              1.0              0.034005           1.0              334.0   \n",
      "1              1.0              0.034005           1.0              334.0   \n",
      "2              1.0              0.034005           1.0              334.0   \n",
      "3              1.0              0.034005           1.0              334.0   \n",
      "4              1.0              0.034005           1.0              334.0   \n",
      "5              1.0              0.034005           1.0              334.0   \n",
      "6              1.0              0.034005           1.0              334.0   \n",
      "7              1.0              0.034005           1.0              334.0   \n",
      "8              1.0              0.034005           1.0              334.0   \n",
      "9              1.0              0.034005           1.0              334.0   \n",
      "\n",
      "        lift  \n",
      "0  29.407186  \n",
      "1  29.407186  \n",
      "2  29.407186  \n",
      "3  29.407186  \n",
      "4  29.407186  \n",
      "5  29.407186  \n",
      "6  29.407186  \n",
      "7  29.407186  \n",
      "8  29.407186  \n",
      "9  29.407186  \n",
      "250_24_SQF_0.5_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                        average_degree: [2.46:2.59[    327.0        9822.0   \n",
      "1                               average_degree>=2.59    328.0        9822.0   \n",
      "2  021D: [0.00:0.00[ AND average_degree: [2.46:2.59[     52.0        9822.0   \n",
      "3  average_degree: [2.46:2.59[ AND degree_central...     34.0        9822.0   \n",
      "4                021C>=0.00 AND average_degree>=2.59    162.0        9822.0   \n",
      "5  021U: [0.00:0.00[ AND average_degree: [2.46:2.59[     32.0        9822.0   \n",
      "6  average_degree>=2.59 AND degree_centralization...     32.0        9822.0   \n",
      "7  021C>=0.00 AND average_degree>=2.59 AND degree...     24.0        9822.0   \n",
      "8  021D: [0.00:0.00[ AND average_degree: [2.46:2.59[     69.0        9822.0   \n",
      "9  average_clustering: [0.02:0.03[ AND average_de...     23.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.290520              0.034005          95.0              334.0   \n",
      "1         0.256098              0.034005          84.0              334.0   \n",
      "2         0.557692              0.034005          29.0              334.0   \n",
      "3         0.647059              0.034005          22.0              334.0   \n",
      "4         0.314815              0.034005          51.0              334.0   \n",
      "5         0.656250              0.034005          21.0              334.0   \n",
      "6         0.656250              0.034005          21.0              334.0   \n",
      "7         0.750000              0.034005          18.0              334.0   \n",
      "8         0.434783              0.034005          30.0              334.0   \n",
      "9         0.695652              0.034005          16.0              334.0   \n",
      "\n",
      "        lift  \n",
      "0   8.543372  \n",
      "1   7.531109  \n",
      "2  16.400161  \n",
      "3  19.028179  \n",
      "4   9.257818  \n",
      "5  19.298466  \n",
      "6  19.298466  \n",
      "7  22.055389  \n",
      "8  12.785733  \n",
      "9  20.457173  \n",
      "250_24_SQF_1_BFS\n",
      "                              subgroup  size_sg  size_dataset  \\\n",
      "0          average_degree: [2.46:2.59[    327.0        9822.0   \n",
      "1                 average_degree>=2.59    328.0        9822.0   \n",
      "2                           021C>=0.00    328.0        9822.0   \n",
      "3                    021D: [0.00:0.00[    327.0        9822.0   \n",
      "4                           021D>=0.00    328.0        9822.0   \n",
      "5   degree_centralization: [0.18:0.19[    328.0        9822.0   \n",
      "6  021C>=0.00 AND average_degree>=2.59    162.0        9822.0   \n",
      "7   degree_centralization: [0.16:0.18[    327.0        9822.0   \n",
      "8                    021D: [0.00:0.00[    328.0        9822.0   \n",
      "9                    021D: [0.00:0.00[    327.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.290520              0.034005          95.0              334.0   \n",
      "1         0.256098              0.034005          84.0              334.0   \n",
      "2         0.195122              0.034005          64.0              334.0   \n",
      "3         0.180428              0.034005          59.0              334.0   \n",
      "4         0.176829              0.034005          58.0              334.0   \n",
      "5         0.173780              0.034005          57.0              334.0   \n",
      "6         0.314815              0.034005          51.0              334.0   \n",
      "7         0.171254              0.034005          56.0              334.0   \n",
      "8         0.170732              0.034005          56.0              334.0   \n",
      "9         0.168196              0.034005          55.0              334.0   \n",
      "\n",
      "       lift  \n",
      "0  8.543372  \n",
      "1  7.531109  \n",
      "2  5.737987  \n",
      "3  5.305884  \n",
      "4  5.200051  \n",
      "5  5.110395  \n",
      "6  9.257818  \n",
      "7  5.036093  \n",
      "8  5.020739  \n",
      "9  4.946163  \n"
     ]
    }
   ],
   "source": [
    "\n",
    "for sg_res in subgroup_topic_results:\n",
    "    for res in sg_res.keys():\n",
    "        result_dict_sg_topic[res]=sg_res[res]\n",
    "        print(res)\n",
    "        pprint(sg_res[res][[\"subgroup\", \"size_sg\", \"size_dataset\", \"target_share_sg\", \"target_share_dataset\", \"positives_sg\", \"positives_dataset\", \"lift\"]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10_4_GA_0_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.884491  012: [0.09:0.10[ AND 021U: [0.15:0.19[ AND ave...      1.0   \n",
      "1  0.884491  012: [0.09:0.10[ AND average_degree: [1.82:1.8...      1.0   \n",
      "2  0.884491  021U: [0.07:0.08[ AND average_degree<1.82 AND ...      1.0   \n",
      "3  0.884491          021U: [0.01:0.01[ AND average_degree<1.82      1.0   \n",
      "4  0.884491          021D: [0.00:0.00[ AND average_degree<1.82      1.0   \n",
      "5  0.884491  021U: [0.15:0.19[ AND average_degree: [1.82:1....      1.0   \n",
      "6  0.884491  average_degree<1.82 AND betweenness_centraliza...      1.0   \n",
      "7  0.884491  average_degree<1.82 AND betweenness_centraliza...      1.0   \n",
      "8  0.884491  average_degree: [1.82:1.83[ AND degree_central...      1.0   \n",
      "9  0.884491  021U: [0.07:0.08[ AND average_degree: [1.82:1....      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0           1.0            12140.0         105099.0   \n",
      "1      105100.0           1.0            12140.0         105099.0   \n",
      "2      105100.0           1.0            12140.0         105099.0   \n",
      "3      105100.0           1.0            12140.0         105099.0   \n",
      "4      105100.0           1.0            12140.0         105099.0   \n",
      "5      105100.0           1.0            12140.0         105099.0   \n",
      "6      105100.0           1.0            12140.0         105099.0   \n",
      "7      105100.0           1.0            12140.0         105099.0   \n",
      "8      105100.0           1.0            12140.0         105099.0   \n",
      "9      105100.0           1.0            12140.0         105099.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0           0.00001                   0.99999     0.000082   \n",
      "1           0.00001                   0.99999     0.000082   \n",
      "2           0.00001                   0.99999     0.000082   \n",
      "3           0.00001                   0.99999     0.000082   \n",
      "4           0.00001                   0.99999     0.000082   \n",
      "5           0.00001                   0.99999     0.000082   \n",
      "6           0.00001                   0.99999     0.000082   \n",
      "7           0.00001                   0.99999     0.000082   \n",
      "8           0.00001                   0.99999     0.000082   \n",
      "9           0.00001                   0.99999     0.000082   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999918              1.0                 0.115501   \n",
      "1             0.999918              1.0                 0.115501   \n",
      "2             0.999918              1.0                 0.115501   \n",
      "3             0.999918              1.0                 0.115501   \n",
      "4             0.999918              1.0                 0.115501   \n",
      "5             0.999918              1.0                 0.115501   \n",
      "6             0.999918              1.0                 0.115501   \n",
      "7             0.999918              1.0                 0.115501   \n",
      "8             0.999918              1.0                 0.115501   \n",
      "9             0.999918              1.0                 0.115501   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.115509  8.657331  \n",
      "1              0.115509  8.657331  \n",
      "2              0.115509  8.657331  \n",
      "3              0.115509  8.657331  \n",
      "4              0.115509  8.657331  \n",
      "5              0.115509  8.657331  \n",
      "6              0.115509  8.657331  \n",
      "7              0.115509  8.657331  \n",
      "8              0.115509  8.657331  \n",
      "9              0.115509  8.657331  \n",
      "10_4_SQF_0_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.884491  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "1  0.884491  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0   \n",
      "2  0.884491  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0   \n",
      "3  0.884491             012: [0.01:0.02[ AND 021C: [0.00:0.00[      1.0   \n",
      "4  0.884491  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0   \n",
      "5  0.884491  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "6  0.884491  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "7  0.884491  012: [0.01:0.02[ AND 021U: [0.00:0.00[ AND 030...      1.0   \n",
      "8  0.884491  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "9  0.884491  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0           1.0            12140.0         105099.0   \n",
      "1      105100.0           1.0            12140.0         105099.0   \n",
      "2      105100.0           1.0            12140.0         105099.0   \n",
      "3      105100.0           1.0            12140.0         105099.0   \n",
      "4      105100.0           1.0            12140.0         105099.0   \n",
      "5      105100.0           1.0            12140.0         105099.0   \n",
      "6      105100.0           1.0            12140.0         105099.0   \n",
      "7      105100.0           1.0            12140.0         105099.0   \n",
      "8      105100.0           1.0            12140.0         105099.0   \n",
      "9      105100.0           1.0            12140.0         105099.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0           0.00001                   0.99999     0.000082   \n",
      "1           0.00001                   0.99999     0.000082   \n",
      "2           0.00001                   0.99999     0.000082   \n",
      "3           0.00001                   0.99999     0.000082   \n",
      "4           0.00001                   0.99999     0.000082   \n",
      "5           0.00001                   0.99999     0.000082   \n",
      "6           0.00001                   0.99999     0.000082   \n",
      "7           0.00001                   0.99999     0.000082   \n",
      "8           0.00001                   0.99999     0.000082   \n",
      "9           0.00001                   0.99999     0.000082   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999918              1.0                 0.115501   \n",
      "1             0.999918              1.0                 0.115501   \n",
      "2             0.999918              1.0                 0.115501   \n",
      "3             0.999918              1.0                 0.115501   \n",
      "4             0.999918              1.0                 0.115501   \n",
      "5             0.999918              1.0                 0.115501   \n",
      "6             0.999918              1.0                 0.115501   \n",
      "7             0.999918              1.0                 0.115501   \n",
      "8             0.999918              1.0                 0.115501   \n",
      "9             0.999918              1.0                 0.115501   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.115509  8.657331  \n",
      "1              0.115509  8.657331  \n",
      "2              0.115509  8.657331  \n",
      "3              0.115509  8.657331  \n",
      "4              0.115509  8.657331  \n",
      "5              0.115509  8.657331  \n",
      "6              0.115509  8.657331  \n",
      "7              0.115509  8.657331  \n",
      "8              0.115509  8.657331  \n",
      "9              0.115509  8.657331  \n",
      "10_4_GA_0.5_SDFS\n",
      "    quality                   subgroup  size_sg  size_dataset  positives_sg  \\\n",
      "0  0.091101       density: [0.00:0.00[   3504.0      105100.0        2153.0   \n",
      "1  0.088410          021C: [0.00:0.00[   3503.0      105100.0        2101.0   \n",
      "2  0.081895           012: [0.02:0.02[   3503.0      105100.0        1976.0   \n",
      "3  0.080000  transitivity: [0.25:0.33[   2820.0      105100.0        1703.0   \n",
      "4  0.077792           012: [0.02:0.02[   3490.0      105100.0        1893.0   \n",
      "5  0.067407       density: [0.00:0.00[   3503.0      105100.0        1698.0   \n",
      "6  0.063832          021D: [0.00:0.00[   3505.0      105100.0        1630.0   \n",
      "7  0.060214          021C: [0.00:0.00[   3503.0      105100.0        1560.0   \n",
      "8  0.058484          030T: [0.00:0.00[   3511.0      105100.0        1529.0   \n",
      "9  0.052673  transitivity: [0.20:0.25[   3179.0      105100.0        1330.0   \n",
      "\n",
      "   positives_dataset  size_complement  relative_size_sg  \\\n",
      "0            12140.0         101596.0          0.033340   \n",
      "1            12140.0         101597.0          0.033330   \n",
      "2            12140.0         101597.0          0.033330   \n",
      "3            12140.0         102280.0          0.026832   \n",
      "4            12140.0         101610.0          0.033206   \n",
      "5            12140.0         101597.0          0.033330   \n",
      "6            12140.0         101595.0          0.033349   \n",
      "7            12140.0         101597.0          0.033330   \n",
      "8            12140.0         101589.0          0.033406   \n",
      "9            12140.0         101921.0          0.030247   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966660     0.177348             0.822652   \n",
      "1                  0.966670     0.173064             0.826936   \n",
      "2                  0.966670     0.162768             0.837232   \n",
      "3                  0.973168     0.140280             0.859720   \n",
      "4                  0.966794     0.155931             0.844069   \n",
      "5                  0.966670     0.139868             0.860132   \n",
      "6                  0.966651     0.134267             0.865733   \n",
      "7                  0.966670     0.128501             0.871499   \n",
      "8                  0.966594     0.125947             0.874053   \n",
      "9                  0.969753     0.109555             0.890445   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.614441                 0.098301              0.115509  5.319416  \n",
      "1         0.599772                 0.098812              0.115509  5.192422  \n",
      "2         0.564088                 0.100042              0.115509  4.883496  \n",
      "3         0.603901                 0.102043              0.115509  5.228168  \n",
      "4         0.542407                 0.100846              0.115509  4.695796  \n",
      "5         0.484727                 0.102779              0.115509  4.196445  \n",
      "6         0.465050                 0.103450              0.115509  4.026091  \n",
      "7         0.445333                 0.104137              0.115509  3.855392  \n",
      "8         0.435488                 0.104450              0.115509  3.770168  \n",
      "9         0.418371                 0.106063              0.115509  3.621972  \n",
      "10_4_SQF_0.5_BFS\n",
      "    quality                                    subgroup  size_sg  \\\n",
      "0  0.091101                        density: [0.00:0.00[   3504.0   \n",
      "1  0.088410                           021C: [0.00:0.00[   3503.0   \n",
      "2  0.081895                            012: [0.02:0.02[   3503.0   \n",
      "3  0.080000                   transitivity: [0.25:0.33[   2820.0   \n",
      "4  0.077792                            012: [0.02:0.02[   3490.0   \n",
      "5  0.076249   012: [0.02:0.02[ AND density: [0.00:0.00[   1791.0   \n",
      "6  0.075332  021C: [0.00:0.00[ AND density: [0.00:0.00[   1516.0   \n",
      "7  0.071450      012: [0.02:0.02[ AND 021C: [0.00:0.00[   1955.0   \n",
      "8  0.067407                        density: [0.00:0.00[   3503.0   \n",
      "9  0.064729   012: [0.02:0.02[ AND density: [0.00:0.00[   1634.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0        2153.0            12140.0         101596.0   \n",
      "1      105100.0        2101.0            12140.0         101597.0   \n",
      "2      105100.0        1976.0            12140.0         101597.0   \n",
      "3      105100.0        1703.0            12140.0         102280.0   \n",
      "4      105100.0        1893.0            12140.0         101610.0   \n",
      "5      105100.0        1253.0            12140.0         103309.0   \n",
      "6      105100.0        1126.0            12140.0         103584.0   \n",
      "7      105100.0        1250.0            12140.0         103145.0   \n",
      "8      105100.0        1698.0            12140.0         101597.0   \n",
      "9      105100.0        1037.0            12140.0         103466.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033340                  0.966660     0.177348   \n",
      "1          0.033330                  0.966670     0.173064   \n",
      "2          0.033330                  0.966670     0.162768   \n",
      "3          0.026832                  0.973168     0.140280   \n",
      "4          0.033206                  0.966794     0.155931   \n",
      "5          0.017041                  0.982959     0.103213   \n",
      "6          0.014424                  0.985576     0.092751   \n",
      "7          0.018601                  0.981399     0.102965   \n",
      "8          0.033330                  0.966670     0.139868   \n",
      "9          0.015547                  0.984453     0.085420   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.822652         0.614441                 0.098301   \n",
      "1             0.826936         0.599772                 0.098812   \n",
      "2             0.837232         0.564088                 0.100042   \n",
      "3             0.859720         0.603901                 0.102043   \n",
      "4             0.844069         0.542407                 0.100846   \n",
      "5             0.896787         0.699609                 0.105383   \n",
      "6             0.907249         0.742744                 0.106329   \n",
      "7             0.897035         0.639386                 0.105580   \n",
      "8             0.860132         0.484727                 0.102779   \n",
      "9             0.914580         0.634639                 0.107311   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.115509  5.319416  \n",
      "1              0.115509  5.192422  \n",
      "2              0.115509  4.883496  \n",
      "3              0.115509  5.228168  \n",
      "4              0.115509  4.695796  \n",
      "5              0.115509  6.056748  \n",
      "6              0.115509  6.430181  \n",
      "7              0.115509  5.535378  \n",
      "8              0.115509  4.196445  \n",
      "9              0.115509  5.494279  \n",
      "10_4_GA_1_SDFS\n",
      "    quality                   subgroup  size_sg  size_dataset  positives_sg  \\\n",
      "0  0.016634       density: [0.00:0.00[   3504.0      105100.0        2153.0   \n",
      "1  0.016141          021C: [0.00:0.00[   3503.0      105100.0        2101.0   \n",
      "2  0.014951           012: [0.02:0.02[   3503.0      105100.0        1976.0   \n",
      "3  0.014176           012: [0.02:0.02[   3490.0      105100.0        1893.0   \n",
      "4  0.013104  transitivity: [0.25:0.33[   2820.0      105100.0        1703.0   \n",
      "5  0.012306       density: [0.00:0.00[   3503.0      105100.0        1698.0   \n",
      "6  0.011657          021D: [0.00:0.00[   3505.0      105100.0        1630.0   \n",
      "7  0.010993          021C: [0.00:0.00[   3503.0      105100.0        1560.0   \n",
      "8  0.010689          030T: [0.00:0.00[   3511.0      105100.0        1529.0   \n",
      "9  0.009520          030T: [0.00:0.00[   3501.0      105100.0        1405.0   \n",
      "\n",
      "   positives_dataset  size_complement  relative_size_sg  \\\n",
      "0            12140.0         101596.0          0.033340   \n",
      "1            12140.0         101597.0          0.033330   \n",
      "2            12140.0         101597.0          0.033330   \n",
      "3            12140.0         101610.0          0.033206   \n",
      "4            12140.0         102280.0          0.026832   \n",
      "5            12140.0         101597.0          0.033330   \n",
      "6            12140.0         101595.0          0.033349   \n",
      "7            12140.0         101597.0          0.033330   \n",
      "8            12140.0         101589.0          0.033406   \n",
      "9            12140.0         101599.0          0.033311   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966660     0.177348             0.822652   \n",
      "1                  0.966670     0.173064             0.826936   \n",
      "2                  0.966670     0.162768             0.837232   \n",
      "3                  0.966794     0.155931             0.844069   \n",
      "4                  0.973168     0.140280             0.859720   \n",
      "5                  0.966670     0.139868             0.860132   \n",
      "6                  0.966651     0.134267             0.865733   \n",
      "7                  0.966670     0.128501             0.871499   \n",
      "8                  0.966594     0.125947             0.874053   \n",
      "9                  0.966689     0.115733             0.884267   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.614441                 0.098301              0.115509  5.319416  \n",
      "1         0.599772                 0.098812              0.115509  5.192422  \n",
      "2         0.564088                 0.100042              0.115509  4.883496  \n",
      "3         0.542407                 0.100846              0.115509  4.695796  \n",
      "4         0.603901                 0.102043              0.115509  5.228168  \n",
      "5         0.484727                 0.102779              0.115509  4.196445  \n",
      "6         0.465050                 0.103450              0.115509  4.026091  \n",
      "7         0.445333                 0.104137              0.115509  3.855392  \n",
      "8         0.435488                 0.104450              0.115509  3.770168  \n",
      "9         0.401314                 0.105660              0.115509  3.474307  \n",
      "10_4_SQF_1_BFS\n",
      "    quality                                   subgroup  size_sg  size_dataset  \\\n",
      "0  0.016634                       density: [0.00:0.00[   3504.0      105100.0   \n",
      "1  0.016141                          021C: [0.00:0.00[   3503.0      105100.0   \n",
      "2  0.014951                           012: [0.02:0.02[   3503.0      105100.0   \n",
      "3  0.014176                           012: [0.02:0.02[   3490.0      105100.0   \n",
      "4  0.013104                  transitivity: [0.25:0.33[   2820.0      105100.0   \n",
      "5  0.012306                       density: [0.00:0.00[   3503.0      105100.0   \n",
      "6  0.011657                          021D: [0.00:0.00[   3505.0      105100.0   \n",
      "7  0.010993                          021C: [0.00:0.00[   3503.0      105100.0   \n",
      "8  0.010689                          030T: [0.00:0.00[   3511.0      105100.0   \n",
      "9  0.009954  012: [0.02:0.02[ AND density: [0.00:0.00[   1791.0      105100.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0        2153.0            12140.0         101596.0          0.033340   \n",
      "1        2101.0            12140.0         101597.0          0.033330   \n",
      "2        1976.0            12140.0         101597.0          0.033330   \n",
      "3        1893.0            12140.0         101610.0          0.033206   \n",
      "4        1703.0            12140.0         102280.0          0.026832   \n",
      "5        1698.0            12140.0         101597.0          0.033330   \n",
      "6        1630.0            12140.0         101595.0          0.033349   \n",
      "7        1560.0            12140.0         101597.0          0.033330   \n",
      "8        1529.0            12140.0         101589.0          0.033406   \n",
      "9        1253.0            12140.0         103309.0          0.017041   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966660     0.177348             0.822652   \n",
      "1                  0.966670     0.173064             0.826936   \n",
      "2                  0.966670     0.162768             0.837232   \n",
      "3                  0.966794     0.155931             0.844069   \n",
      "4                  0.973168     0.140280             0.859720   \n",
      "5                  0.966670     0.139868             0.860132   \n",
      "6                  0.966651     0.134267             0.865733   \n",
      "7                  0.966670     0.128501             0.871499   \n",
      "8                  0.966594     0.125947             0.874053   \n",
      "9                  0.982959     0.103213             0.896787   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.614441                 0.098301              0.115509  5.319416  \n",
      "1         0.599772                 0.098812              0.115509  5.192422  \n",
      "2         0.564088                 0.100042              0.115509  4.883496  \n",
      "3         0.542407                 0.100846              0.115509  4.695796  \n",
      "4         0.603901                 0.102043              0.115509  5.228168  \n",
      "5         0.484727                 0.102779              0.115509  4.196445  \n",
      "6         0.465050                 0.103450              0.115509  4.026091  \n",
      "7         0.445333                 0.104137              0.115509  3.855392  \n",
      "8         0.435488                 0.104450              0.115509  3.770168  \n",
      "9         0.699609                 0.105383              0.115509  6.056748  \n",
      "10_15_GA_0_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.973416  012: [0.13:0.15[ AND average_degree: [2.05:2.0...      1.0   \n",
      "1  0.973416  012: [0.18:0.20[ AND average_degree: [1.88:1.8...      1.0   \n",
      "2  0.973416  012: [0.18:0.20[ AND 021C: [0.01:0.01[ AND ave...      1.0   \n",
      "3  0.973416  012: [0.23:0.25[ AND average_degree: [1.85:1.8...      1.0   \n",
      "4  0.973416  021C: [0.00:0.00[ AND average_degree: [2.03:2....      1.0   \n",
      "5  0.973416  average_degree: [1.92:1.94[ AND betweenness_ce...      1.0   \n",
      "6  0.973416  average_degree: [1.94:1.96[ AND betweenness_ce...      1.0   \n",
      "7  0.973416  021C: [0.01:0.02[ AND average_degree: [1.88:1....      1.0   \n",
      "8  0.973416  average_degree: [2.05:2.06[ AND betweenness_ce...      1.0   \n",
      "9  0.973416  average_clustering: [0.06:0.08[ AND average_de...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0           1.0             2794.0         105099.0   \n",
      "1      105100.0           1.0             2794.0         105099.0   \n",
      "2      105100.0           1.0             2794.0         105099.0   \n",
      "3      105100.0           1.0             2794.0         105099.0   \n",
      "4      105100.0           1.0             2794.0         105099.0   \n",
      "5      105100.0           1.0             2794.0         105099.0   \n",
      "6      105100.0           1.0             2794.0         105099.0   \n",
      "7      105100.0           1.0             2794.0         105099.0   \n",
      "8      105100.0           1.0             2794.0         105099.0   \n",
      "9      105100.0           1.0             2794.0         105099.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0           0.00001                   0.99999     0.000358   \n",
      "1           0.00001                   0.99999     0.000358   \n",
      "2           0.00001                   0.99999     0.000358   \n",
      "3           0.00001                   0.99999     0.000358   \n",
      "4           0.00001                   0.99999     0.000358   \n",
      "5           0.00001                   0.99999     0.000358   \n",
      "6           0.00001                   0.99999     0.000358   \n",
      "7           0.00001                   0.99999     0.000358   \n",
      "8           0.00001                   0.99999     0.000358   \n",
      "9           0.00001                   0.99999     0.000358   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999642              1.0                 0.026575   \n",
      "1             0.999642              1.0                 0.026575   \n",
      "2             0.999642              1.0                 0.026575   \n",
      "3             0.999642              1.0                 0.026575   \n",
      "4             0.999642              1.0                 0.026575   \n",
      "5             0.999642              1.0                 0.026575   \n",
      "6             0.999642              1.0                 0.026575   \n",
      "7             0.999642              1.0                 0.026575   \n",
      "8             0.999642              1.0                 0.026575   \n",
      "9             0.999642              1.0                 0.026575   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.026584  37.616321  \n",
      "1              0.026584  37.616321  \n",
      "2              0.026584  37.616321  \n",
      "3              0.026584  37.616321  \n",
      "4              0.026584  37.616321  \n",
      "5              0.026584  37.616321  \n",
      "6              0.026584  37.616321  \n",
      "7              0.026584  37.616321  \n",
      "8              0.026584  37.616321  \n",
      "9              0.026584  37.616321  \n",
      "10_15_SQF_0_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.973416  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 021...      2.0   \n",
      "1  0.973416  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "2  0.973416  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "3  0.973416  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "4  0.973416  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "5  0.973416  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "6  0.973416                          transitivity: [0.00:0.00[      1.0   \n",
      "7  0.973416  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "8  0.973416                          transitivity: [0.00:0.00[      1.0   \n",
      "9  0.973416  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0           2.0             2794.0         105098.0   \n",
      "1      105100.0           1.0             2794.0         105099.0   \n",
      "2      105100.0           1.0             2794.0         105099.0   \n",
      "3      105100.0           1.0             2794.0         105099.0   \n",
      "4      105100.0           1.0             2794.0         105099.0   \n",
      "5      105100.0           1.0             2794.0         105099.0   \n",
      "6      105100.0           1.0             2794.0         105099.0   \n",
      "7      105100.0           1.0             2794.0         105099.0   \n",
      "8      105100.0           1.0             2794.0         105099.0   \n",
      "9      105100.0           1.0             2794.0         105099.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000019                  0.999981     0.000716   \n",
      "1          0.000010                  0.999990     0.000358   \n",
      "2          0.000010                  0.999990     0.000358   \n",
      "3          0.000010                  0.999990     0.000358   \n",
      "4          0.000010                  0.999990     0.000358   \n",
      "5          0.000010                  0.999990     0.000358   \n",
      "6          0.000010                  0.999990     0.000358   \n",
      "7          0.000010                  0.999990     0.000358   \n",
      "8          0.000010                  0.999990     0.000358   \n",
      "9          0.000010                  0.999990     0.000358   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999284              1.0                 0.026566   \n",
      "1             0.999642              1.0                 0.026575   \n",
      "2             0.999642              1.0                 0.026575   \n",
      "3             0.999642              1.0                 0.026575   \n",
      "4             0.999642              1.0                 0.026575   \n",
      "5             0.999642              1.0                 0.026575   \n",
      "6             0.999642              1.0                 0.026575   \n",
      "7             0.999642              1.0                 0.026575   \n",
      "8             0.999642              1.0                 0.026575   \n",
      "9             0.999642              1.0                 0.026575   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.026584  37.616321  \n",
      "1              0.026584  37.616321  \n",
      "2              0.026584  37.616321  \n",
      "3              0.026584  37.616321  \n",
      "4              0.026584  37.616321  \n",
      "5              0.026584  37.616321  \n",
      "6              0.026584  37.616321  \n",
      "7              0.026584  37.616321  \n",
      "8              0.026584  37.616321  \n",
      "9              0.026584  37.616321  \n",
      "10_15_GA_0.5_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.023598                               average_degree>=2.53   3504.0   \n",
      "1  0.022329                               density: [0.01:0.01[   3508.0   \n",
      "2  0.022082  betweenness_centralization>=0.07 AND density: ...     93.0   \n",
      "3  0.021169      average_degree>=2.53 AND density: [0.01:0.01[    374.0   \n",
      "4  0.020199         021C: [0.00:0.00[ AND average_degree>=2.53    235.0   \n",
      "5  0.019865         021U: [0.00:0.00[ AND average_degree>=2.53    259.0   \n",
      "6  0.019524      average_degree>=2.53 AND density: [0.01:0.01[    293.0   \n",
      "7  0.019462  average_degree>=2.53 AND degree_centralization...    265.0   \n",
      "8  0.018770  021C: [0.00:0.00[ AND betweenness_centralizati...     68.0   \n",
      "9  0.018631         021D: [0.00:0.00[ AND average_degree>=2.53    272.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0         546.0             2794.0         101596.0   \n",
      "1      105100.0         522.0             2794.0         101592.0   \n",
      "2      105100.0          77.0             2794.0         105007.0   \n",
      "3      105100.0         191.0             2794.0         104726.0   \n",
      "4      105100.0         137.0             2794.0         104865.0   \n",
      "5      105100.0         144.0             2794.0         104841.0   \n",
      "6      105100.0         154.0             2794.0         104807.0   \n",
      "7      105100.0         144.0             2794.0         104835.0   \n",
      "8      105100.0          56.0             2794.0         105032.0   \n",
      "9      105100.0         142.0             2794.0         104828.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033340                  0.966660     0.195419   \n",
      "1          0.033378                  0.966622     0.186829   \n",
      "2          0.000885                  0.999115     0.027559   \n",
      "3          0.003559                  0.996441     0.068361   \n",
      "4          0.002236                  0.997764     0.049034   \n",
      "5          0.002464                  0.997536     0.051539   \n",
      "6          0.002788                  0.997212     0.055118   \n",
      "7          0.002521                  0.997479     0.051539   \n",
      "8          0.000647                  0.999353     0.020043   \n",
      "9          0.002588                  0.997412     0.050823   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.804581         0.155822                 0.022127   \n",
      "1             0.813171         0.148803                 0.022364   \n",
      "2             0.972441         0.827957                 0.025874   \n",
      "3             0.931639         0.510695                 0.024855   \n",
      "4             0.950966         0.582979                 0.025337   \n",
      "5             0.948461         0.555985                 0.025276   \n",
      "6             0.944882         0.525597                 0.025189   \n",
      "7             0.948461         0.543396                 0.025278   \n",
      "8             0.979957         0.823529                 0.026068   \n",
      "9             0.949177         0.522059                 0.025299   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.026584   5.861447  \n",
      "1              0.026584   5.597411  \n",
      "2              0.026584  31.144696  \n",
      "3              0.026584  19.210474  \n",
      "4              0.026584  21.929515  \n",
      "5              0.026584  20.914093  \n",
      "6              0.026584  19.771035  \n",
      "7              0.026584  20.440567  \n",
      "8              0.026584  30.978146  \n",
      "9              0.026584  19.637932  \n",
      "10_15_SQF_0.5_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.028879      average_degree>=2.53 AND density: [0.01:0.01[    374.0   \n",
      "1  0.027751  021C: [0.00:0.00[ AND average_degree>=2.53 AND...    187.0   \n",
      "2  0.026348      average_degree>=2.53 AND density: [0.01:0.01[    293.0   \n",
      "3  0.026310         021C: [0.00:0.00[ AND average_degree>=2.53    235.0   \n",
      "4  0.026287  012: [0.04:0.05[ AND 021C: [0.00:0.00[ AND ave...    136.0   \n",
      "5  0.026280         021U: [0.00:0.00[ AND average_degree>=2.53    259.0   \n",
      "6  0.026227  average_degree>=2.53 AND betweenness_centraliz...    129.0   \n",
      "7  0.026198  021D: [0.00:0.00[ AND average_degree>=2.53 AND...    110.0   \n",
      "8  0.025951  average_degree>=2.53 AND degree_centralization...    265.0   \n",
      "9  0.025206         021D: [0.00:0.00[ AND average_degree>=2.53    272.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0         191.0             2794.0         104726.0   \n",
      "1      105100.0         128.0             2794.0         104913.0   \n",
      "2      105100.0         154.0             2794.0         104807.0   \n",
      "3      105100.0         137.0             2794.0         104865.0   \n",
      "4      105100.0         103.0             2794.0         104964.0   \n",
      "5      105100.0         144.0             2794.0         104841.0   \n",
      "6      105100.0         100.0             2794.0         104971.0   \n",
      "7      105100.0          92.0             2794.0         104990.0   \n",
      "8      105100.0         144.0             2794.0         104835.0   \n",
      "9      105100.0         142.0             2794.0         104828.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.003559                  0.996441     0.068361   \n",
      "1          0.001779                  0.998221     0.045812   \n",
      "2          0.002788                  0.997212     0.055118   \n",
      "3          0.002236                  0.997764     0.049034   \n",
      "4          0.001294                  0.998706     0.036865   \n",
      "5          0.002464                  0.997536     0.051539   \n",
      "6          0.001227                  0.998773     0.035791   \n",
      "7          0.001047                  0.998953     0.032928   \n",
      "8          0.002521                  0.997479     0.051539   \n",
      "9          0.002588                  0.997412     0.050823   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.931639         0.510695                 0.024855   \n",
      "1             0.954188         0.684492                 0.025412   \n",
      "2             0.944882         0.525597                 0.025189   \n",
      "3             0.950966         0.582979                 0.025337   \n",
      "4             0.963135         0.757353                 0.025637   \n",
      "5             0.948461         0.555985                 0.025276   \n",
      "6             0.964209         0.775194                 0.025664   \n",
      "7             0.967072         0.836364                 0.025736   \n",
      "8             0.948461         0.543396                 0.025278   \n",
      "9             0.949177         0.522059                 0.025299   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.026584  19.210474  \n",
      "1              0.026584  25.748070  \n",
      "2              0.026584  19.771035  \n",
      "3              0.026584  21.929515  \n",
      "4              0.026584  28.488831  \n",
      "5              0.026584  20.914093  \n",
      "6              0.026584  29.159939  \n",
      "7              0.026584  31.460923  \n",
      "8              0.026584  20.440567  \n",
      "9              0.026584  19.637932  \n",
      "10_15_GA_1_SDFS\n",
      "    quality                          subgroup  size_sg  size_dataset  \\\n",
      "0  0.004309              average_degree>=2.53   3504.0      105100.0   \n",
      "1  0.004079              density: [0.01:0.01[   3508.0      105100.0   \n",
      "2  0.002540              density: [0.01:0.01[   3499.0      105100.0   \n",
      "3  0.001968  betweenness_centralization>=0.07   3504.0      105100.0   \n",
      "4  0.001635                  012: [0.03:0.04[   3503.0      105100.0   \n",
      "5  0.001628         transitivity: [0.07:0.10[   3570.0      105100.0   \n",
      "6  0.001617         transitivity: [0.05:0.07[   2898.0      105100.0   \n",
      "7  0.001550         transitivity: [0.02:0.05[   3503.0      105100.0   \n",
      "8  0.001502                 021C: [0.00:0.00[   3503.0      105100.0   \n",
      "9  0.001492                  012: [0.03:0.03[   3504.0      105100.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         546.0             2794.0         101596.0          0.033340   \n",
      "1         522.0             2794.0         101592.0          0.033378   \n",
      "2         360.0             2794.0         101601.0          0.033292   \n",
      "3         300.0             2794.0         101596.0          0.033340   \n",
      "4         265.0             2794.0         101597.0          0.033330   \n",
      "5         266.0             2794.0         101530.0          0.033968   \n",
      "6         247.0             2794.0         102202.0          0.027574   \n",
      "7         256.0             2794.0         101597.0          0.033330   \n",
      "8         251.0             2794.0         101597.0          0.033330   \n",
      "9         250.0             2794.0         101596.0          0.033340   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966660     0.195419             0.804581   \n",
      "1                  0.966622     0.186829             0.813171   \n",
      "2                  0.966708     0.128848             0.871152   \n",
      "3                  0.966660     0.107373             0.892627   \n",
      "4                  0.966670     0.094846             0.905154   \n",
      "5                  0.966032     0.095204             0.904796   \n",
      "6                  0.972426     0.088404             0.911596   \n",
      "7                  0.966670     0.091625             0.908375   \n",
      "8                  0.966670     0.089835             0.910165   \n",
      "9                  0.966660     0.089477             0.910523   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.155822                 0.022127              0.026584  5.861447  \n",
      "1         0.148803                 0.022364              0.026584  5.597411  \n",
      "2         0.102887                 0.023956              0.026584  3.870213  \n",
      "3         0.085616                 0.024548              0.026584  3.220575  \n",
      "4         0.075649                 0.024892              0.026584  2.845654  \n",
      "5         0.074510                 0.024899              0.026584  2.802785  \n",
      "6         0.085231                 0.024921              0.026584  3.206084  \n",
      "7         0.073080                 0.024981              0.026584  2.749009  \n",
      "8         0.071653                 0.025030              0.026584  2.695317  \n",
      "9         0.071347                 0.025040              0.026584  2.683813  \n",
      "10_15_SQF_1_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.004309                               average_degree>=2.53   3504.0   \n",
      "1  0.004079                               density: [0.01:0.01[   3508.0   \n",
      "2  0.002540                               density: [0.01:0.01[   3499.0   \n",
      "3  0.001968                   betweenness_centralization>=0.07   3504.0   \n",
      "4  0.001789  average_degree>=2.53 AND betweenness_centraliz...    562.0   \n",
      "5  0.001723      average_degree>=2.53 AND density: [0.01:0.01[    374.0   \n",
      "6  0.001635                                   012: [0.03:0.04[   3503.0   \n",
      "7  0.001628                          transitivity: [0.07:0.10[   3570.0   \n",
      "8  0.001617                          transitivity: [0.05:0.07[   2898.0   \n",
      "9  0.001614         021C: [0.00:0.00[ AND density: [0.01:0.01[    765.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0         546.0             2794.0         101596.0   \n",
      "1      105100.0         522.0             2794.0         101592.0   \n",
      "2      105100.0         360.0             2794.0         101601.0   \n",
      "3      105100.0         300.0             2794.0         101596.0   \n",
      "4      105100.0         203.0             2794.0         104538.0   \n",
      "5      105100.0         191.0             2794.0         104726.0   \n",
      "6      105100.0         265.0             2794.0         101597.0   \n",
      "7      105100.0         266.0             2794.0         101530.0   \n",
      "8      105100.0         247.0             2794.0         102202.0   \n",
      "9      105100.0         190.0             2794.0         104335.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033340                  0.966660     0.195419   \n",
      "1          0.033378                  0.966622     0.186829   \n",
      "2          0.033292                  0.966708     0.128848   \n",
      "3          0.033340                  0.966660     0.107373   \n",
      "4          0.005347                  0.994653     0.072656   \n",
      "5          0.003559                  0.996441     0.068361   \n",
      "6          0.033330                  0.966670     0.094846   \n",
      "7          0.033968                  0.966032     0.095204   \n",
      "8          0.027574                  0.972426     0.088404   \n",
      "9          0.007279                  0.992721     0.068003   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.804581         0.155822                 0.022127   \n",
      "1             0.813171         0.148803                 0.022364   \n",
      "2             0.871152         0.102887                 0.023956   \n",
      "3             0.892627         0.085616                 0.024548   \n",
      "4             0.927344         0.361210                 0.024785   \n",
      "5             0.931639         0.510695                 0.024855   \n",
      "6             0.905154         0.075649                 0.024892   \n",
      "7             0.904796         0.074510                 0.024899   \n",
      "8             0.911596         0.085231                 0.024921   \n",
      "9             0.931997         0.248366                 0.024958   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.026584   5.861447  \n",
      "1              0.026584   5.597411  \n",
      "2              0.026584   3.870213  \n",
      "3              0.026584   3.220575  \n",
      "4              0.026584  13.587390  \n",
      "5              0.026584  19.210474  \n",
      "6              0.026584   2.845654  \n",
      "7              0.026584   2.802785  \n",
      "8              0.026584   3.206084  \n",
      "9              0.026584   9.342616  \n",
      "10_17_GA_0_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.958088           012: [0.06:0.07[ AND average_degree<1.82      1.0   \n",
      "1  0.958088  012: [0.16:0.17[ AND 021C: [0.00:0.00[ AND ave...      1.0   \n",
      "2  0.958088  021C: [0.01:0.01[ AND 021U: [0.04:0.05[ AND av...      1.0   \n",
      "3  0.958088  012: [0.25:0.27[ AND 021U: [0.03:0.03[ AND ave...      1.0   \n",
      "4  0.958088  012: [0.22:0.23[ AND 021C: [0.01:0.01[ AND ave...      1.0   \n",
      "5  0.958088  021D: [0.00:0.00[ AND average_degree: [1.88:1.88[      1.0   \n",
      "6  0.958088  average_degree: [1.82:1.83[ AND density: [0.03...      1.0   \n",
      "7  0.958088         021D>=0.01 AND average_degree: [1.86:1.88[      1.0   \n",
      "8  0.958088  021C: [0.01:0.01[ AND 021U: [0.04:0.05[ AND av...      1.0   \n",
      "9  0.958088  021C: [0.00:0.00[ AND average_degree: [1.89:1....      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0           1.0             4405.0         105099.0   \n",
      "1      105100.0           1.0             4405.0         105099.0   \n",
      "2      105100.0           1.0             4405.0         105099.0   \n",
      "3      105100.0           1.0             4405.0         105099.0   \n",
      "4      105100.0           1.0             4405.0         105099.0   \n",
      "5      105100.0           1.0             4405.0         105099.0   \n",
      "6      105100.0           1.0             4405.0         105099.0   \n",
      "7      105100.0           1.0             4405.0         105099.0   \n",
      "8      105100.0           1.0             4405.0         105099.0   \n",
      "9      105100.0           1.0             4405.0         105099.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0           0.00001                   0.99999     0.000227   \n",
      "1           0.00001                   0.99999     0.000227   \n",
      "2           0.00001                   0.99999     0.000227   \n",
      "3           0.00001                   0.99999     0.000227   \n",
      "4           0.00001                   0.99999     0.000227   \n",
      "5           0.00001                   0.99999     0.000227   \n",
      "6           0.00001                   0.99999     0.000227   \n",
      "7           0.00001                   0.99999     0.000227   \n",
      "8           0.00001                   0.99999     0.000227   \n",
      "9           0.00001                   0.99999     0.000227   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999773              1.0                 0.041903   \n",
      "1             0.999773              1.0                 0.041903   \n",
      "2             0.999773              1.0                 0.041903   \n",
      "3             0.999773              1.0                 0.041903   \n",
      "4             0.999773              1.0                 0.041903   \n",
      "5             0.999773              1.0                 0.041903   \n",
      "6             0.999773              1.0                 0.041903   \n",
      "7             0.999773              1.0                 0.041903   \n",
      "8             0.999773              1.0                 0.041903   \n",
      "9             0.999773              1.0                 0.041903   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.041912  23.859251  \n",
      "1              0.041912  23.859251  \n",
      "2              0.041912  23.859251  \n",
      "3              0.041912  23.859251  \n",
      "4              0.041912  23.859251  \n",
      "5              0.041912  23.859251  \n",
      "6              0.041912  23.859251  \n",
      "7              0.041912  23.859251  \n",
      "8              0.041912  23.859251  \n",
      "9              0.041912  23.859251  \n",
      "10_17_SQF_0_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.958088  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "1  0.958088  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 030...      2.0   \n",
      "2  0.958088  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "3  0.958088  012: [0.01:0.02[ AND 021U: [0.03:0.03[ AND 030...      1.0   \n",
      "4  0.958088  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0   \n",
      "5  0.958088  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "6  0.958088  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "7  0.958088  012: [0.02:0.02[ AND 021U: [0.02:0.02[ AND 030...      1.0   \n",
      "8  0.958088  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "9  0.958088  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0           1.0             4405.0         105099.0   \n",
      "1      105100.0           2.0             4405.0         105098.0   \n",
      "2      105100.0           1.0             4405.0         105099.0   \n",
      "3      105100.0           1.0             4405.0         105099.0   \n",
      "4      105100.0           1.0             4405.0         105099.0   \n",
      "5      105100.0           1.0             4405.0         105099.0   \n",
      "6      105100.0           1.0             4405.0         105099.0   \n",
      "7      105100.0           1.0             4405.0         105099.0   \n",
      "8      105100.0           1.0             4405.0         105099.0   \n",
      "9      105100.0           1.0             4405.0         105099.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000010                  0.999990     0.000227   \n",
      "1          0.000019                  0.999981     0.000454   \n",
      "2          0.000010                  0.999990     0.000227   \n",
      "3          0.000010                  0.999990     0.000227   \n",
      "4          0.000010                  0.999990     0.000227   \n",
      "5          0.000010                  0.999990     0.000227   \n",
      "6          0.000010                  0.999990     0.000227   \n",
      "7          0.000010                  0.999990     0.000227   \n",
      "8          0.000010                  0.999990     0.000227   \n",
      "9          0.000010                  0.999990     0.000227   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999773              1.0                 0.041903   \n",
      "1             0.999546              1.0                 0.041894   \n",
      "2             0.999773              1.0                 0.041903   \n",
      "3             0.999773              1.0                 0.041903   \n",
      "4             0.999773              1.0                 0.041903   \n",
      "5             0.999773              1.0                 0.041903   \n",
      "6             0.999773              1.0                 0.041903   \n",
      "7             0.999773              1.0                 0.041903   \n",
      "8             0.999773              1.0                 0.041903   \n",
      "9             0.999773              1.0                 0.041903   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.041912  23.859251  \n",
      "1              0.041912  23.859251  \n",
      "2              0.041912  23.859251  \n",
      "3              0.041912  23.859251  \n",
      "4              0.041912  23.859251  \n",
      "5              0.041912  23.859251  \n",
      "6              0.041912  23.859251  \n",
      "7              0.041912  23.859251  \n",
      "8              0.041912  23.859251  \n",
      "9              0.041912  23.859251  \n",
      "10_17_GA_0.5_SDFS\n",
      "    quality                   subgroup  size_sg  size_dataset  positives_sg  \\\n",
      "0  0.027417          021C: [0.00:0.00[   3504.0      105100.0         673.0   \n",
      "1  0.026147           012: [0.02:0.03[   3516.0      105100.0         650.0   \n",
      "2  0.023775       density: [0.00:0.01[   3503.0      105100.0         603.0   \n",
      "3  0.022650       density: [0.01:0.01[   3499.0      105100.0         581.0   \n",
      "4  0.016161           012: [0.03:0.03[   3504.0      105100.0         457.0   \n",
      "5  0.014654          021D: [0.00:0.00[   3503.0      105100.0         428.0   \n",
      "6  0.013925          030T: [0.00:0.00[   3503.0      105100.0         414.0   \n",
      "7  0.013199          021D: [0.00:0.00[   3502.0      105100.0         400.0   \n",
      "8  0.012726          021C: [0.00:0.00[   3503.0      105100.0         391.0   \n",
      "9  0.012570  transitivity: [0.02:0.05[   3503.0      105100.0         388.0   \n",
      "\n",
      "   positives_dataset  size_complement  relative_size_sg  \\\n",
      "0             4405.0         101596.0          0.033340   \n",
      "1             4405.0         101584.0          0.033454   \n",
      "2             4405.0         101597.0          0.033330   \n",
      "3             4405.0         101601.0          0.033292   \n",
      "4             4405.0         101596.0          0.033340   \n",
      "5             4405.0         101597.0          0.033330   \n",
      "6             4405.0         101597.0          0.033330   \n",
      "7             4405.0         101598.0          0.033321   \n",
      "8             4405.0         101597.0          0.033330   \n",
      "9             4405.0         101597.0          0.033330   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966660     0.152781             0.847219   \n",
      "1                  0.966546     0.147560             0.852440   \n",
      "2                  0.966670     0.136890             0.863110   \n",
      "3                  0.966708     0.131896             0.868104   \n",
      "4                  0.966660     0.103746             0.896254   \n",
      "5                  0.966670     0.097162             0.902838   \n",
      "6                  0.966670     0.093984             0.906016   \n",
      "7                  0.966679     0.090806             0.909194   \n",
      "8                  0.966670     0.088763             0.911237   \n",
      "9                  0.966670     0.088082             0.911918   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.192066                 0.036734              0.041912  4.582556  \n",
      "1         0.184869                 0.036964              0.041912  4.410840  \n",
      "2         0.172138                 0.037422              0.041912  4.107088  \n",
      "3         0.166047                 0.037637              0.041912  3.961768  \n",
      "4         0.130422                 0.038860              0.041912  3.111780  \n",
      "5         0.122181                 0.039145              0.041912  2.915147  \n",
      "6         0.118184                 0.039283              0.041912  2.819792  \n",
      "7         0.114220                 0.039420              0.041912  2.725214  \n",
      "8         0.111619                 0.039509              0.041912  2.663136  \n",
      "9         0.110762                 0.039539              0.041912  2.642703  \n",
      "10_17_SQF_0.5_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.027417                                  021C: [0.00:0.00[   3504.0   \n",
      "1  0.026147                                   012: [0.02:0.03[   3516.0   \n",
      "2  0.026109             012: [0.02:0.03[ AND 021C: [0.00:0.00[   2012.0   \n",
      "3  0.025609          012: [0.02:0.03[ AND density: [0.00:0.01[   1558.0   \n",
      "4  0.025015         021C: [0.00:0.00[ AND density: [0.00:0.01[    992.0   \n",
      "5  0.023822  012: [0.02:0.03[ AND 021C: [0.00:0.00[ AND den...    857.0   \n",
      "6  0.023775                               density: [0.00:0.01[   3503.0   \n",
      "7  0.022939             012: [0.02:0.03[ AND 021D: [0.00:0.00[    694.0   \n",
      "8  0.022902            021C: [0.00:0.00[ AND 021D: [0.00:0.00[    626.0   \n",
      "9  0.022650                               density: [0.01:0.01[   3499.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0         673.0             4405.0         101596.0   \n",
      "1      105100.0         650.0             4405.0         101584.0   \n",
      "2      105100.0         464.0             4405.0         103088.0   \n",
      "3      105100.0         393.0             4405.0         103542.0   \n",
      "4      105100.0         297.0             4405.0         104108.0   \n",
      "5      105100.0         262.0             4405.0         104243.0   \n",
      "6      105100.0         603.0             4405.0         101597.0   \n",
      "7      105100.0         225.0             4405.0         104406.0   \n",
      "8      105100.0         212.0             4405.0         104474.0   \n",
      "9      105100.0         581.0             4405.0         101601.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033340                  0.966660     0.152781   \n",
      "1          0.033454                  0.966546     0.147560   \n",
      "2          0.019144                  0.980856     0.105335   \n",
      "3          0.014824                  0.985176     0.089217   \n",
      "4          0.009439                  0.990561     0.067423   \n",
      "5          0.008154                  0.991846     0.059478   \n",
      "6          0.033330                  0.966670     0.136890   \n",
      "7          0.006603                  0.993397     0.051078   \n",
      "8          0.005956                  0.994044     0.048127   \n",
      "9          0.033292                  0.966708     0.131896   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.847219         0.192066                 0.036734   \n",
      "1             0.852440         0.184869                 0.036964   \n",
      "2             0.894665         0.230616                 0.038229   \n",
      "3             0.910783         0.252246                 0.038748   \n",
      "4             0.932577         0.299395                 0.039459   \n",
      "5             0.940522         0.305718                 0.039744   \n",
      "6             0.863110         0.172138                 0.037422   \n",
      "7             0.948922         0.324207                 0.040036   \n",
      "8             0.951873         0.338658                 0.040134   \n",
      "9             0.868104         0.166047                 0.037637   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.041912  4.582556  \n",
      "1              0.041912  4.410840  \n",
      "2              0.041912  5.502332  \n",
      "3              0.041912  6.018412  \n",
      "4              0.041912  7.143344  \n",
      "5              0.041912  7.294193  \n",
      "6              0.041912  4.107088  \n",
      "7              0.041912  7.735348  \n",
      "8              0.041912  8.080130  \n",
      "9              0.041912  3.961768  \n",
      "10_17_GA_1_SDFS\n",
      "    quality                   subgroup  size_sg  size_dataset  positives_sg  \\\n",
      "0  0.005006          021C: [0.00:0.00[   3504.0      105100.0         673.0   \n",
      "1  0.004782           012: [0.02:0.03[   3516.0      105100.0         650.0   \n",
      "2  0.004340       density: [0.00:0.01[   3503.0      105100.0         603.0   \n",
      "3  0.004133       density: [0.01:0.01[   3499.0      105100.0         581.0   \n",
      "4  0.002951           012: [0.03:0.03[   3504.0      105100.0         457.0   \n",
      "5  0.002675          021D: [0.00:0.00[   3503.0      105100.0         428.0   \n",
      "6  0.002542          030T: [0.00:0.00[   3503.0      105100.0         414.0   \n",
      "7  0.002409          021D: [0.00:0.00[   3502.0      105100.0         400.0   \n",
      "8  0.002323          021C: [0.00:0.00[   3503.0      105100.0         391.0   \n",
      "9  0.002295  transitivity: [0.02:0.05[   3503.0      105100.0         388.0   \n",
      "\n",
      "   positives_dataset  size_complement  relative_size_sg  \\\n",
      "0             4405.0         101596.0          0.033340   \n",
      "1             4405.0         101584.0          0.033454   \n",
      "2             4405.0         101597.0          0.033330   \n",
      "3             4405.0         101601.0          0.033292   \n",
      "4             4405.0         101596.0          0.033340   \n",
      "5             4405.0         101597.0          0.033330   \n",
      "6             4405.0         101597.0          0.033330   \n",
      "7             4405.0         101598.0          0.033321   \n",
      "8             4405.0         101597.0          0.033330   \n",
      "9             4405.0         101597.0          0.033330   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966660     0.152781             0.847219   \n",
      "1                  0.966546     0.147560             0.852440   \n",
      "2                  0.966670     0.136890             0.863110   \n",
      "3                  0.966708     0.131896             0.868104   \n",
      "4                  0.966660     0.103746             0.896254   \n",
      "5                  0.966670     0.097162             0.902838   \n",
      "6                  0.966670     0.093984             0.906016   \n",
      "7                  0.966679     0.090806             0.909194   \n",
      "8                  0.966670     0.088763             0.911237   \n",
      "9                  0.966670     0.088082             0.911918   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.192066                 0.036734              0.041912  4.582556  \n",
      "1         0.184869                 0.036964              0.041912  4.410840  \n",
      "2         0.172138                 0.037422              0.041912  4.107088  \n",
      "3         0.166047                 0.037637              0.041912  3.961768  \n",
      "4         0.130422                 0.038860              0.041912  3.111780  \n",
      "5         0.122181                 0.039145              0.041912  2.915147  \n",
      "6         0.118184                 0.039283              0.041912  2.819792  \n",
      "7         0.114220                 0.039420              0.041912  2.725214  \n",
      "8         0.111619                 0.039509              0.041912  2.663136  \n",
      "9         0.110762                 0.039539              0.041912  2.642703  \n",
      "10_17_SQF_1_BFS\n",
      "    quality                                    subgroup  size_sg  \\\n",
      "0  0.005006                           021C: [0.00:0.00[   3504.0   \n",
      "1  0.004782                            012: [0.02:0.03[   3516.0   \n",
      "2  0.004340                        density: [0.00:0.01[   3503.0   \n",
      "3  0.004133                        density: [0.01:0.01[   3499.0   \n",
      "4  0.003612      012: [0.02:0.03[ AND 021C: [0.00:0.00[   2012.0   \n",
      "5  0.003118   012: [0.02:0.03[ AND density: [0.00:0.01[   1558.0   \n",
      "6  0.002951                            012: [0.03:0.03[   3504.0   \n",
      "7  0.002675                           021D: [0.00:0.00[   3503.0   \n",
      "8  0.002542                           030T: [0.00:0.00[   3503.0   \n",
      "9  0.002430  021C: [0.00:0.00[ AND density: [0.00:0.01[    992.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0      105100.0         673.0             4405.0         101596.0   \n",
      "1      105100.0         650.0             4405.0         101584.0   \n",
      "2      105100.0         603.0             4405.0         101597.0   \n",
      "3      105100.0         581.0             4405.0         101601.0   \n",
      "4      105100.0         464.0             4405.0         103088.0   \n",
      "5      105100.0         393.0             4405.0         103542.0   \n",
      "6      105100.0         457.0             4405.0         101596.0   \n",
      "7      105100.0         428.0             4405.0         101597.0   \n",
      "8      105100.0         414.0             4405.0         101597.0   \n",
      "9      105100.0         297.0             4405.0         104108.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033340                  0.966660     0.152781   \n",
      "1          0.033454                  0.966546     0.147560   \n",
      "2          0.033330                  0.966670     0.136890   \n",
      "3          0.033292                  0.966708     0.131896   \n",
      "4          0.019144                  0.980856     0.105335   \n",
      "5          0.014824                  0.985176     0.089217   \n",
      "6          0.033340                  0.966660     0.103746   \n",
      "7          0.033330                  0.966670     0.097162   \n",
      "8          0.033330                  0.966670     0.093984   \n",
      "9          0.009439                  0.990561     0.067423   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.847219         0.192066                 0.036734   \n",
      "1             0.852440         0.184869                 0.036964   \n",
      "2             0.863110         0.172138                 0.037422   \n",
      "3             0.868104         0.166047                 0.037637   \n",
      "4             0.894665         0.230616                 0.038229   \n",
      "5             0.910783         0.252246                 0.038748   \n",
      "6             0.896254         0.130422                 0.038860   \n",
      "7             0.902838         0.122181                 0.039145   \n",
      "8             0.906016         0.118184                 0.039283   \n",
      "9             0.932577         0.299395                 0.039459   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.041912  4.582556  \n",
      "1              0.041912  4.410840  \n",
      "2              0.041912  4.107088  \n",
      "3              0.041912  3.961768  \n",
      "4              0.041912  5.502332  \n",
      "5              0.041912  6.018412  \n",
      "6              0.041912  3.111780  \n",
      "7              0.041912  2.915147  \n",
      "8              0.041912  2.819792  \n",
      "9              0.041912  7.143344  \n",
      "250_2_GA_0_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.959377         021C: [0.00:0.00[ AND density: [0.00:0.00[      2.0   \n",
      "1  0.959377     012: [0.02:0.02[ AND transitivity: [0.16:0.17[      1.0   \n",
      "2  0.959377  030T: [0.00:0.00[ AND average_degree: [2.14:2.15[      1.0   \n",
      "3  0.959377   012: [0.02:0.02[ AND average_degree: [2.40:2.46[      1.0   \n",
      "4  0.959377  average_degree: [2.05:2.06[ AND transitivity: ...      1.0   \n",
      "5  0.959377         030T: [0.00:0.00[ AND density: [0.00:0.00[      1.0   \n",
      "6  0.959377    021U: [0.00:0.00[ AND longest_path: [0.10:0.11[      1.0   \n",
      "7  0.950000  012: [0.02:0.02[ AND 030T: [0.00:0.00[ AND ave...      1.0   \n",
      "8  0.947368  021C: [0.00:0.00[ AND 030T: [0.00:0.00[ AND lo...      1.0   \n",
      "9  0.947368  030T: [0.00:0.00[ AND average_degree: [2.20:2....      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0           2.0              399.0           9820.0   \n",
      "1        9822.0           1.0              399.0           9821.0   \n",
      "2        9822.0           1.0              399.0           9821.0   \n",
      "3        9822.0           1.0              399.0           9821.0   \n",
      "4        9822.0           1.0              399.0           9821.0   \n",
      "5        9822.0           1.0              399.0           9821.0   \n",
      "6        9822.0           1.0              399.0           9821.0   \n",
      "7        9822.0           1.0              399.0           9821.0   \n",
      "8        9822.0           1.0              399.0           9821.0   \n",
      "9        9822.0           1.0              399.0           9821.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000204                  0.999796     0.005013   \n",
      "1          0.000102                  0.999898     0.002506   \n",
      "2          0.000102                  0.999898     0.002506   \n",
      "3          0.000102                  0.999898     0.002506   \n",
      "4          0.000102                  0.999898     0.002506   \n",
      "5          0.000102                  0.999898     0.002506   \n",
      "6          0.000102                  0.999898     0.002506   \n",
      "7          0.000102                  0.999898     0.002506   \n",
      "8          0.000102                  0.999898     0.002506   \n",
      "9          0.000102                  0.999898     0.002506   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.994987              1.0                 0.040428   \n",
      "1             0.997494              1.0                 0.040525   \n",
      "2             0.997494              1.0                 0.040525   \n",
      "3             0.997494              1.0                 0.040525   \n",
      "4             0.997494              1.0                 0.040525   \n",
      "5             0.997494              1.0                 0.040525   \n",
      "6             0.997494              1.0                 0.040525   \n",
      "7             0.997494              1.0                 0.040525   \n",
      "8             0.997494              1.0                 0.040525   \n",
      "9             0.997494              1.0                 0.040525   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.040623  24.616541  \n",
      "1              0.040623  24.616541  \n",
      "2              0.040623  24.616541  \n",
      "3              0.040623  24.616541  \n",
      "4              0.040623  24.616541  \n",
      "5              0.040623  24.616541  \n",
      "6              0.040623  24.616541  \n",
      "7              0.040623  24.616541  \n",
      "8              0.040623  24.616541  \n",
      "9              0.040623  24.616541  \n",
      "250_2_SQF_0_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.959377  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "1  0.959377  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND 030...      2.0   \n",
      "2  0.959377  012: [0.01:0.01[ AND 021C<0.00 AND 030T: [0.00...      3.0   \n",
      "3  0.959377  012: [0.01:0.01[ AND 021C<0.00 AND 021D: [0.00...      1.0   \n",
      "4  0.959377  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "5  0.959377  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "6  0.959377  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "7  0.959377  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "8  0.959377             012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0   \n",
      "9  0.959377  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0           1.0              399.0           9821.0   \n",
      "1        9822.0           2.0              399.0           9820.0   \n",
      "2        9822.0           3.0              399.0           9819.0   \n",
      "3        9822.0           1.0              399.0           9821.0   \n",
      "4        9822.0           1.0              399.0           9821.0   \n",
      "5        9822.0           1.0              399.0           9821.0   \n",
      "6        9822.0           1.0              399.0           9821.0   \n",
      "7        9822.0           1.0              399.0           9821.0   \n",
      "8        9822.0           1.0              399.0           9821.0   \n",
      "9        9822.0           1.0              399.0           9821.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000102                  0.999898     0.002506   \n",
      "1          0.000204                  0.999796     0.005013   \n",
      "2          0.000305                  0.999695     0.007519   \n",
      "3          0.000102                  0.999898     0.002506   \n",
      "4          0.000102                  0.999898     0.002506   \n",
      "5          0.000102                  0.999898     0.002506   \n",
      "6          0.000102                  0.999898     0.002506   \n",
      "7          0.000102                  0.999898     0.002506   \n",
      "8          0.000102                  0.999898     0.002506   \n",
      "9          0.000102                  0.999898     0.002506   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.997494              1.0                 0.040525   \n",
      "1             0.994987              1.0                 0.040428   \n",
      "2             0.992481              1.0                 0.040330   \n",
      "3             0.997494              1.0                 0.040525   \n",
      "4             0.997494              1.0                 0.040525   \n",
      "5             0.997494              1.0                 0.040525   \n",
      "6             0.997494              1.0                 0.040525   \n",
      "7             0.997494              1.0                 0.040525   \n",
      "8             0.997494              1.0                 0.040525   \n",
      "9             0.997494              1.0                 0.040525   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.040623  24.616541  \n",
      "1              0.040623  24.616541  \n",
      "2              0.040623  24.616541  \n",
      "3              0.040623  24.616541  \n",
      "4              0.040623  24.616541  \n",
      "5              0.040623  24.616541  \n",
      "6              0.040623  24.616541  \n",
      "7              0.040623  24.616541  \n",
      "8              0.040623  24.616541  \n",
      "9              0.040623  24.616541  \n",
      "250_2_GA_0.5_SDFS\n",
      "    quality                                 subgroup  size_sg  size_dataset  \\\n",
      "0  0.098606                         012: [0.01:0.01[    327.0        9822.0   \n",
      "1  0.097490                        021C: [0.00:0.00[    327.0        9822.0   \n",
      "2  0.096932       degree_centralization: [0.72:0.83[    327.0        9822.0   \n",
      "3  0.095258                        021U: [0.01:0.01[    327.0        9822.0   \n",
      "4  0.074741                      average_degree<2.01    322.0        9822.0   \n",
      "5  0.073038                                021D<0.00    318.0        9822.0   \n",
      "6  0.056757  betweenness_centralization: [0.00:0.00[    327.0        9822.0   \n",
      "7  0.050619                                021C<0.00    327.0        9822.0   \n",
      "8  0.048387                longest_path: [0.02:0.02[    327.0        9822.0   \n",
      "9  0.047271          betweenness_centralization<0.00    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         190.0              399.0           9495.0          0.033293   \n",
      "1         188.0              399.0           9495.0          0.033293   \n",
      "2         187.0              399.0           9495.0          0.033293   \n",
      "3         184.0              399.0           9495.0          0.033293   \n",
      "4         146.0              399.0           9500.0          0.032784   \n",
      "5         142.0              399.0           9504.0          0.032376   \n",
      "6         115.0              399.0           9495.0          0.033293   \n",
      "7         104.0              399.0           9495.0          0.033293   \n",
      "8         100.0              399.0           9495.0          0.033293   \n",
      "9          98.0              399.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.476190             0.523810   \n",
      "1                  0.966707     0.471178             0.528822   \n",
      "2                  0.966707     0.468672             0.531328   \n",
      "3                  0.966707     0.461153             0.538847   \n",
      "4                  0.967216     0.365915             0.634085   \n",
      "5                  0.967624     0.355890             0.644110   \n",
      "6                  0.966707     0.288221             0.711779   \n",
      "7                  0.966707     0.260652             0.739348   \n",
      "8                  0.966707     0.250627             0.749373   \n",
      "9                  0.966707     0.245614             0.754386   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset       lift  \n",
      "0         0.581040                 0.022012              0.040623  14.303189  \n",
      "1         0.574924                 0.022222              0.040623  14.152629  \n",
      "2         0.571865                 0.022328              0.040623  14.077349  \n",
      "3         0.562691                 0.022643              0.040623  13.851510  \n",
      "4         0.453416                 0.026632              0.040623  11.161537  \n",
      "5         0.446541                 0.027041              0.040623  10.992292  \n",
      "6         0.351682                 0.029910              0.040623   8.657193  \n",
      "7         0.318043                 0.031069              0.040623   7.829114  \n",
      "8         0.305810                 0.031490              0.040623   7.527994  \n",
      "9         0.299694                 0.031701              0.040623   7.377434  \n",
      "250_2_SQF_0.5_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.098606                                   012: [0.01:0.01[    327.0   \n",
      "1  0.097490                                  021C: [0.00:0.00[    327.0   \n",
      "2  0.096932                 degree_centralization: [0.72:0.83[    327.0   \n",
      "3  0.095525             012: [0.01:0.01[ AND 021C: [0.00:0.00[    290.0   \n",
      "4  0.095263  012: [0.01:0.01[ AND degree_centralization: [0...    282.0   \n",
      "5  0.095258                                  021U: [0.01:0.01[    327.0   \n",
      "6  0.094471  021U: [0.01:0.01[ AND degree_centralization: [...    283.0   \n",
      "7  0.093511  012: [0.01:0.01[ AND 021U: [0.01:0.01[ AND deg...    242.0   \n",
      "8  0.093497  021C: [0.00:0.00[ AND degree_centralization: [...    251.0   \n",
      "9  0.092859             012: [0.01:0.01[ AND 021U: [0.01:0.01[    245.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0         190.0              399.0           9495.0   \n",
      "1        9822.0         188.0              399.0           9495.0   \n",
      "2        9822.0         187.0              399.0           9495.0   \n",
      "3        9822.0         173.0              399.0           9532.0   \n",
      "4        9822.0         170.0              399.0           9540.0   \n",
      "5        9822.0         184.0              399.0           9495.0   \n",
      "6        9822.0         169.0              399.0           9539.0   \n",
      "7        9822.0         154.0              399.0           9580.0   \n",
      "8        9822.0         157.0              399.0           9571.0   \n",
      "9        9822.0         154.0              399.0           9577.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033293                  0.966707     0.476190   \n",
      "1          0.033293                  0.966707     0.471178   \n",
      "2          0.033293                  0.966707     0.468672   \n",
      "3          0.029526                  0.970474     0.433584   \n",
      "4          0.028711                  0.971289     0.426065   \n",
      "5          0.033293                  0.966707     0.461153   \n",
      "6          0.028813                  0.971187     0.423559   \n",
      "7          0.024639                  0.975361     0.385965   \n",
      "8          0.025555                  0.974445     0.393484   \n",
      "9          0.024944                  0.975056     0.385965   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.523810         0.581040                 0.022012   \n",
      "1             0.528822         0.574924                 0.022222   \n",
      "2             0.531328         0.571865                 0.022328   \n",
      "3             0.566416         0.596552                 0.023710   \n",
      "4             0.573935         0.602837                 0.024004   \n",
      "5             0.538847         0.562691                 0.022643   \n",
      "6             0.576441         0.597173                 0.024112   \n",
      "7             0.614035         0.636364                 0.025574   \n",
      "8             0.606516         0.625498                 0.025285   \n",
      "9             0.614035         0.628571                 0.025582   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.040623  14.303189  \n",
      "1              0.040623  14.152629  \n",
      "2              0.040623  14.077349  \n",
      "3              0.040623  14.685040  \n",
      "4              0.040623  14.839759  \n",
      "5              0.040623  13.851510  \n",
      "6              0.040623  14.700337  \n",
      "7              0.040623  15.665072  \n",
      "8              0.040623  15.397598  \n",
      "9              0.040623  15.473255  \n",
      "250_2_GA_1_SDFS\n",
      "    quality                                 subgroup  size_sg  size_dataset  \\\n",
      "0  0.017992                         012: [0.01:0.01[    327.0        9822.0   \n",
      "1  0.017788                        021C: [0.00:0.00[    327.0        9822.0   \n",
      "2  0.017686       degree_centralization: [0.72:0.83[    327.0        9822.0   \n",
      "3  0.017381                        021U: [0.01:0.01[    327.0        9822.0   \n",
      "4  0.013533                      average_degree<2.01    322.0        9822.0   \n",
      "5  0.013142                                021D<0.00    318.0        9822.0   \n",
      "6  0.010356  betweenness_centralization: [0.00:0.00[    327.0        9822.0   \n",
      "7  0.009236                                021C<0.00    327.0        9822.0   \n",
      "8  0.009049                         030T: [0.0:0.00[    643.0        9822.0   \n",
      "9  0.008829                longest_path: [0.02:0.02[    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         190.0              399.0           9495.0          0.033293   \n",
      "1         188.0              399.0           9495.0          0.033293   \n",
      "2         187.0              399.0           9495.0          0.033293   \n",
      "3         184.0              399.0           9495.0          0.033293   \n",
      "4         146.0              399.0           9500.0          0.032784   \n",
      "5         142.0              399.0           9504.0          0.032376   \n",
      "6         115.0              399.0           9495.0          0.033293   \n",
      "7         104.0              399.0           9495.0          0.033293   \n",
      "8         115.0              399.0           9179.0          0.065465   \n",
      "9         100.0              399.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.476190             0.523810   \n",
      "1                  0.966707     0.471178             0.528822   \n",
      "2                  0.966707     0.468672             0.531328   \n",
      "3                  0.966707     0.461153             0.538847   \n",
      "4                  0.967216     0.365915             0.634085   \n",
      "5                  0.967624     0.355890             0.644110   \n",
      "6                  0.966707     0.288221             0.711779   \n",
      "7                  0.966707     0.260652             0.739348   \n",
      "8                  0.934535     0.288221             0.711779   \n",
      "9                  0.966707     0.250627             0.749373   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset       lift  \n",
      "0         0.581040                 0.022012              0.040623  14.303189  \n",
      "1         0.574924                 0.022222              0.040623  14.152629  \n",
      "2         0.571865                 0.022328              0.040623  14.077349  \n",
      "3         0.562691                 0.022643              0.040623  13.851510  \n",
      "4         0.453416                 0.026632              0.040623  11.161537  \n",
      "5         0.446541                 0.027041              0.040623  10.992292  \n",
      "6         0.351682                 0.029910              0.040623   8.657193  \n",
      "7         0.318043                 0.031069              0.040623   7.829114  \n",
      "8         0.178849                 0.030940              0.040623   4.402647  \n",
      "9         0.305810                 0.031490              0.040623   7.527994  \n",
      "250_2_SQF_1_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.017992                                   012: [0.01:0.01[    327.0   \n",
      "1  0.017788                                  021C: [0.00:0.00[    327.0   \n",
      "2  0.017686                 degree_centralization: [0.72:0.83[    327.0   \n",
      "3  0.017381                                  021U: [0.01:0.01[    327.0   \n",
      "4  0.016414             012: [0.01:0.01[ AND 021C: [0.00:0.00[    290.0   \n",
      "5  0.016142  012: [0.01:0.01[ AND degree_centralization: [0...    282.0   \n",
      "6  0.016036  021U: [0.01:0.01[ AND degree_centralization: [...    283.0   \n",
      "7  0.014946  021C: [0.00:0.00[ AND degree_centralization: [...    251.0   \n",
      "8  0.014751  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND deg...    249.0   \n",
      "9  0.014678  012: [0.01:0.01[ AND 021U: [0.01:0.01[ AND deg...    242.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0         190.0              399.0           9495.0   \n",
      "1        9822.0         188.0              399.0           9495.0   \n",
      "2        9822.0         187.0              399.0           9495.0   \n",
      "3        9822.0         184.0              399.0           9495.0   \n",
      "4        9822.0         173.0              399.0           9532.0   \n",
      "5        9822.0         170.0              399.0           9540.0   \n",
      "6        9822.0         169.0              399.0           9539.0   \n",
      "7        9822.0         157.0              399.0           9571.0   \n",
      "8        9822.0         155.0              399.0           9573.0   \n",
      "9        9822.0         154.0              399.0           9580.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033293                  0.966707     0.476190   \n",
      "1          0.033293                  0.966707     0.471178   \n",
      "2          0.033293                  0.966707     0.468672   \n",
      "3          0.033293                  0.966707     0.461153   \n",
      "4          0.029526                  0.970474     0.433584   \n",
      "5          0.028711                  0.971289     0.426065   \n",
      "6          0.028813                  0.971187     0.423559   \n",
      "7          0.025555                  0.974445     0.393484   \n",
      "8          0.025351                  0.974649     0.388471   \n",
      "9          0.024639                  0.975361     0.385965   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.523810         0.581040                 0.022012   \n",
      "1             0.528822         0.574924                 0.022222   \n",
      "2             0.531328         0.571865                 0.022328   \n",
      "3             0.538847         0.562691                 0.022643   \n",
      "4             0.566416         0.596552                 0.023710   \n",
      "5             0.573935         0.602837                 0.024004   \n",
      "6             0.576441         0.597173                 0.024112   \n",
      "7             0.606516         0.625498                 0.025285   \n",
      "8             0.611529         0.622490                 0.025488   \n",
      "9             0.614035         0.636364                 0.025574   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.040623  14.303189  \n",
      "1              0.040623  14.152629  \n",
      "2              0.040623  14.077349  \n",
      "3              0.040623  13.851510  \n",
      "4              0.040623  14.685040  \n",
      "5              0.040623  14.839759  \n",
      "6              0.040623  14.700337  \n",
      "7              0.040623  15.397598  \n",
      "8              0.040623  15.323550  \n",
      "9              0.040623  15.665072  \n",
      "250_4_GA_0_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.607412  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND ave...      1.0   \n",
      "1  0.607412  021D: [0.00:0.00[ AND average_degree: [2.01:2....      1.0   \n",
      "2  0.607412  021C: [0.00:0.00[ AND average_degree: [2.01:2....      1.0   \n",
      "3  0.607412  030T: [0.00:0.00[ AND average_degree: [2.01:2....      1.0   \n",
      "4  0.607412  average_clustering: [0.00:0.00[ AND average_de...      1.0   \n",
      "5  0.607412  021D: [0.00:0.00[ AND average_degree: [2.01:2....      1.0   \n",
      "6  0.607412  030T: [0.00:0.00[ AND average_degree: [2.01:2....      1.0   \n",
      "7  0.607412  average_clustering: [0.00:0.00[ AND average_de...      1.0   \n",
      "8  0.607412  average_clustering: [0.00:0.00[ AND average_de...      1.0   \n",
      "9  0.607412  average_degree: [2.02:2.04[ AND density: [0.00...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0           1.0             3856.0           9821.0   \n",
      "1        9822.0           1.0             3856.0           9821.0   \n",
      "2        9822.0           1.0             3856.0           9821.0   \n",
      "3        9822.0           1.0             3856.0           9821.0   \n",
      "4        9822.0           1.0             3856.0           9821.0   \n",
      "5        9822.0           1.0             3856.0           9821.0   \n",
      "6        9822.0           1.0             3856.0           9821.0   \n",
      "7        9822.0           1.0             3856.0           9821.0   \n",
      "8        9822.0           1.0             3856.0           9821.0   \n",
      "9        9822.0           1.0             3856.0           9821.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000102                  0.999898     0.000259   \n",
      "1          0.000102                  0.999898     0.000259   \n",
      "2          0.000102                  0.999898     0.000259   \n",
      "3          0.000102                  0.999898     0.000259   \n",
      "4          0.000102                  0.999898     0.000259   \n",
      "5          0.000102                  0.999898     0.000259   \n",
      "6          0.000102                  0.999898     0.000259   \n",
      "7          0.000102                  0.999898     0.000259   \n",
      "8          0.000102                  0.999898     0.000259   \n",
      "9          0.000102                  0.999898     0.000259   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999741              1.0                 0.392526   \n",
      "1             0.999741              1.0                 0.392526   \n",
      "2             0.999741              1.0                 0.392526   \n",
      "3             0.999741              1.0                 0.392526   \n",
      "4             0.999741              1.0                 0.392526   \n",
      "5             0.999741              1.0                 0.392526   \n",
      "6             0.999741              1.0                 0.392526   \n",
      "7             0.999741              1.0                 0.392526   \n",
      "8             0.999741              1.0                 0.392526   \n",
      "9             0.999741              1.0                 0.392526   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.392588  2.547199  \n",
      "1              0.392588  2.547199  \n",
      "2              0.392588  2.547199  \n",
      "3              0.392588  2.547199  \n",
      "4              0.392588  2.547199  \n",
      "5              0.392588  2.547199  \n",
      "6              0.392588  2.547199  \n",
      "7              0.392588  2.547199  \n",
      "8              0.392588  2.547199  \n",
      "9              0.392588  2.547199  \n",
      "250_4_SQF_0_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.607412  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "1  0.607412  012: [0.01:0.02[ AND 021U: [0.00:0.00[ AND 030...      1.0   \n",
      "2  0.607412             012: [0.02:0.02[ AND 021C: [0.00:0.00[      1.0   \n",
      "3  0.607412  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "4  0.607412  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "5  0.607412  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "6  0.607412  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "7  0.607412  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "8  0.607412  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "9  0.607412  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0           1.0             3856.0           9821.0   \n",
      "1        9822.0           1.0             3856.0           9821.0   \n",
      "2        9822.0           1.0             3856.0           9821.0   \n",
      "3        9822.0           1.0             3856.0           9821.0   \n",
      "4        9822.0           1.0             3856.0           9821.0   \n",
      "5        9822.0           1.0             3856.0           9821.0   \n",
      "6        9822.0           1.0             3856.0           9821.0   \n",
      "7        9822.0           1.0             3856.0           9821.0   \n",
      "8        9822.0           1.0             3856.0           9821.0   \n",
      "9        9822.0           1.0             3856.0           9821.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000102                  0.999898     0.000259   \n",
      "1          0.000102                  0.999898     0.000259   \n",
      "2          0.000102                  0.999898     0.000259   \n",
      "3          0.000102                  0.999898     0.000259   \n",
      "4          0.000102                  0.999898     0.000259   \n",
      "5          0.000102                  0.999898     0.000259   \n",
      "6          0.000102                  0.999898     0.000259   \n",
      "7          0.000102                  0.999898     0.000259   \n",
      "8          0.000102                  0.999898     0.000259   \n",
      "9          0.000102                  0.999898     0.000259   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.999741              1.0                 0.392526   \n",
      "1             0.999741              1.0                 0.392526   \n",
      "2             0.999741              1.0                 0.392526   \n",
      "3             0.999741              1.0                 0.392526   \n",
      "4             0.999741              1.0                 0.392526   \n",
      "5             0.999741              1.0                 0.392526   \n",
      "6             0.999741              1.0                 0.392526   \n",
      "7             0.999741              1.0                 0.392526   \n",
      "8             0.999741              1.0                 0.392526   \n",
      "9             0.999741              1.0                 0.392526   \n",
      "\n",
      "   target_share_dataset      lift  \n",
      "0              0.392588  2.547199  \n",
      "1              0.392588  2.547199  \n",
      "2              0.392588  2.547199  \n",
      "3              0.392588  2.547199  \n",
      "4              0.392588  2.547199  \n",
      "5              0.392588  2.547199  \n",
      "6              0.392588  2.547199  \n",
      "7              0.392588  2.547199  \n",
      "8              0.392588  2.547199  \n",
      "9              0.392588  2.547199  \n",
      "250_4_GA_0.5_SDFS\n",
      "    quality                         subgroup  size_sg  size_dataset  \\\n",
      "0  0.095764  average_clustering: [0.07:0.08[    327.0        9822.0   \n",
      "1  0.094285         average_clustering>=0.08    328.0        9822.0   \n",
      "2  0.092057  average_clustering: [0.06:0.07[    328.0        9822.0   \n",
      "3  0.081815  average_clustering: [0.05:0.06[    327.0        9822.0   \n",
      "4  0.080699  average_clustering: [0.05:0.05[    327.0        9822.0   \n",
      "5  0.080572                030T: [0.00:0.00[    329.0        9822.0   \n",
      "6  0.075677                030T: [0.00:0.00[    327.0        9822.0   \n",
      "7  0.075342  average_clustering: [0.04:0.05[    328.0        9822.0   \n",
      "8  0.072723        transitivity: [0.28:0.30[    294.0        9822.0   \n",
      "9  0.070097                030T: [0.00:0.00[    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         300.0             3856.0           9495.0          0.033293   \n",
      "1         298.0             3856.0           9494.0          0.033394   \n",
      "2         294.0             3856.0           9494.0          0.033394   \n",
      "3         275.0             3856.0           9495.0          0.033293   \n",
      "4         273.0             3856.0           9495.0          0.033293   \n",
      "5         274.0             3856.0           9493.0          0.033496   \n",
      "6         264.0             3856.0           9495.0          0.033293   \n",
      "7         264.0             3856.0           9494.0          0.033394   \n",
      "8         239.0             3856.0           9528.0          0.029933   \n",
      "9         254.0             3856.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.077801             0.922199   \n",
      "1                  0.966606     0.077282             0.922718   \n",
      "2                  0.966606     0.076245             0.923755   \n",
      "3                  0.966707     0.071317             0.928683   \n",
      "4                  0.966707     0.070799             0.929201   \n",
      "5                  0.966504     0.071058             0.928942   \n",
      "6                  0.966707     0.068465             0.931535   \n",
      "7                  0.966606     0.068465             0.931535   \n",
      "8                  0.970067     0.061981             0.938019   \n",
      "9                  0.966707     0.065871             0.934129   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.917431                 0.374513              0.392588  2.336880  \n",
      "1         0.908537                 0.374763              0.392588  2.314224  \n",
      "2         0.896341                 0.375184              0.392588  2.283160  \n",
      "3         0.840979                 0.377146              0.392588  2.142140  \n",
      "4         0.834862                 0.377357              0.392588  2.126561  \n",
      "5         0.832827                 0.377331              0.392588  2.121376  \n",
      "6         0.807339                 0.378304              0.392588  2.056454  \n",
      "7         0.804878                 0.378344              0.392588  2.050185  \n",
      "8         0.812925                 0.379618              0.392588  2.070682  \n",
      "9         0.776758                 0.379358              0.392588  1.978558  \n",
      "250_4_SQF_0.5_BFS\n",
      "    quality                         subgroup  size_sg  size_dataset  \\\n",
      "0  0.095764  average_clustering: [0.07:0.08[    327.0        9822.0   \n",
      "1  0.094285         average_clustering>=0.08    328.0        9822.0   \n",
      "2  0.092057  average_clustering: [0.06:0.07[    328.0        9822.0   \n",
      "3  0.081815  average_clustering: [0.05:0.06[    327.0        9822.0   \n",
      "4  0.080699  average_clustering: [0.05:0.05[    327.0        9822.0   \n",
      "5  0.080572                030T: [0.00:0.00[    329.0        9822.0   \n",
      "6  0.075677                030T: [0.00:0.00[    327.0        9822.0   \n",
      "7  0.075342  average_clustering: [0.04:0.05[    328.0        9822.0   \n",
      "8  0.072723        transitivity: [0.28:0.30[    294.0        9822.0   \n",
      "9  0.070097                030T: [0.00:0.00[    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         300.0             3856.0           9495.0          0.033293   \n",
      "1         298.0             3856.0           9494.0          0.033394   \n",
      "2         294.0             3856.0           9494.0          0.033394   \n",
      "3         275.0             3856.0           9495.0          0.033293   \n",
      "4         273.0             3856.0           9495.0          0.033293   \n",
      "5         274.0             3856.0           9493.0          0.033496   \n",
      "6         264.0             3856.0           9495.0          0.033293   \n",
      "7         264.0             3856.0           9494.0          0.033394   \n",
      "8         239.0             3856.0           9528.0          0.029933   \n",
      "9         254.0             3856.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.077801             0.922199   \n",
      "1                  0.966606     0.077282             0.922718   \n",
      "2                  0.966606     0.076245             0.923755   \n",
      "3                  0.966707     0.071317             0.928683   \n",
      "4                  0.966707     0.070799             0.929201   \n",
      "5                  0.966504     0.071058             0.928942   \n",
      "6                  0.966707     0.068465             0.931535   \n",
      "7                  0.966606     0.068465             0.931535   \n",
      "8                  0.970067     0.061981             0.938019   \n",
      "9                  0.966707     0.065871             0.934129   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.917431                 0.374513              0.392588  2.336880  \n",
      "1         0.908537                 0.374763              0.392588  2.314224  \n",
      "2         0.896341                 0.375184              0.392588  2.283160  \n",
      "3         0.840979                 0.377146              0.392588  2.142140  \n",
      "4         0.834862                 0.377357              0.392588  2.126561  \n",
      "5         0.832827                 0.377331              0.392588  2.121376  \n",
      "6         0.807339                 0.378304              0.392588  2.056454  \n",
      "7         0.804878                 0.378344              0.392588  2.050185  \n",
      "8         0.812925                 0.379618              0.392588  2.070682  \n",
      "9         0.776758                 0.379358              0.392588  1.978558  \n",
      "250_4_GA_1_SDFS\n",
      "    quality                         subgroup  size_sg  size_dataset  \\\n",
      "0  0.017473  average_clustering: [0.07:0.08[    327.0        9822.0   \n",
      "1  0.017230         average_clustering>=0.08    328.0        9822.0   \n",
      "2  0.016823  average_clustering: [0.06:0.07[    328.0        9822.0   \n",
      "3  0.014928  average_clustering: [0.05:0.06[    327.0        9822.0   \n",
      "4  0.014746                030T: [0.00:0.00[    329.0        9822.0   \n",
      "5  0.014724  average_clustering: [0.05:0.05[    327.0        9822.0   \n",
      "6  0.013808                030T: [0.00:0.00[    327.0        9822.0   \n",
      "7  0.013768  average_clustering: [0.04:0.05[    328.0        9822.0   \n",
      "8  0.013064        transitivity: [0.30:0.33[    366.0        9822.0   \n",
      "9  0.012790                030T: [0.00:0.00[    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         300.0             3856.0           9495.0          0.033293   \n",
      "1         298.0             3856.0           9494.0          0.033394   \n",
      "2         294.0             3856.0           9494.0          0.033394   \n",
      "3         275.0             3856.0           9495.0          0.033293   \n",
      "4         274.0             3856.0           9493.0          0.033496   \n",
      "5         273.0             3856.0           9495.0          0.033293   \n",
      "6         264.0             3856.0           9495.0          0.033293   \n",
      "7         264.0             3856.0           9494.0          0.033394   \n",
      "8         272.0             3856.0           9456.0          0.037263   \n",
      "9         254.0             3856.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.077801             0.922199   \n",
      "1                  0.966606     0.077282             0.922718   \n",
      "2                  0.966606     0.076245             0.923755   \n",
      "3                  0.966707     0.071317             0.928683   \n",
      "4                  0.966504     0.071058             0.928942   \n",
      "5                  0.966707     0.070799             0.929201   \n",
      "6                  0.966707     0.068465             0.931535   \n",
      "7                  0.966606     0.068465             0.931535   \n",
      "8                  0.962737     0.070539             0.929461   \n",
      "9                  0.966707     0.065871             0.934129   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.917431                 0.374513              0.392588  2.336880  \n",
      "1         0.908537                 0.374763              0.392588  2.314224  \n",
      "2         0.896341                 0.375184              0.392588  2.283160  \n",
      "3         0.840979                 0.377146              0.392588  2.142140  \n",
      "4         0.832827                 0.377331              0.392588  2.121376  \n",
      "5         0.834862                 0.377357              0.392588  2.126561  \n",
      "6         0.807339                 0.378304              0.392588  2.056454  \n",
      "7         0.804878                 0.378344              0.392588  2.050185  \n",
      "8         0.743169                 0.379019              0.392588  1.893000  \n",
      "9         0.776758                 0.379358              0.392588  1.978558  \n",
      "250_4_SQF_1_BFS\n",
      "    quality                         subgroup  size_sg  size_dataset  \\\n",
      "0  0.017473  average_clustering: [0.07:0.08[    327.0        9822.0   \n",
      "1  0.017230         average_clustering>=0.08    328.0        9822.0   \n",
      "2  0.016823  average_clustering: [0.06:0.07[    328.0        9822.0   \n",
      "3  0.014928  average_clustering: [0.05:0.06[    327.0        9822.0   \n",
      "4  0.014746                030T: [0.00:0.00[    329.0        9822.0   \n",
      "5  0.014724  average_clustering: [0.05:0.05[    327.0        9822.0   \n",
      "6  0.013808                030T: [0.00:0.00[    327.0        9822.0   \n",
      "7  0.013768  average_clustering: [0.04:0.05[    328.0        9822.0   \n",
      "8  0.013064        transitivity: [0.30:0.33[    366.0        9822.0   \n",
      "9  0.012790                030T: [0.00:0.00[    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0         300.0             3856.0           9495.0          0.033293   \n",
      "1         298.0             3856.0           9494.0          0.033394   \n",
      "2         294.0             3856.0           9494.0          0.033394   \n",
      "3         275.0             3856.0           9495.0          0.033293   \n",
      "4         274.0             3856.0           9493.0          0.033496   \n",
      "5         273.0             3856.0           9495.0          0.033293   \n",
      "6         264.0             3856.0           9495.0          0.033293   \n",
      "7         264.0             3856.0           9494.0          0.033394   \n",
      "8         272.0             3856.0           9456.0          0.037263   \n",
      "9         254.0             3856.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.077801             0.922199   \n",
      "1                  0.966606     0.077282             0.922718   \n",
      "2                  0.966606     0.076245             0.923755   \n",
      "3                  0.966707     0.071317             0.928683   \n",
      "4                  0.966504     0.071058             0.928942   \n",
      "5                  0.966707     0.070799             0.929201   \n",
      "6                  0.966707     0.068465             0.931535   \n",
      "7                  0.966606     0.068465             0.931535   \n",
      "8                  0.962737     0.070539             0.929461   \n",
      "9                  0.966707     0.065871             0.934129   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.917431                 0.374513              0.392588  2.336880  \n",
      "1         0.908537                 0.374763              0.392588  2.314224  \n",
      "2         0.896341                 0.375184              0.392588  2.283160  \n",
      "3         0.840979                 0.377146              0.392588  2.142140  \n",
      "4         0.832827                 0.377331              0.392588  2.121376  \n",
      "5         0.834862                 0.377357              0.392588  2.126561  \n",
      "6         0.807339                 0.378304              0.392588  2.056454  \n",
      "7         0.804878                 0.378344              0.392588  2.050185  \n",
      "8         0.743169                 0.379019              0.392588  1.893000  \n",
      "9         0.776758                 0.379358              0.392588  1.978558  \n",
      "250_24_GA_0_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.965995             012<0.01 AND transitivity: [0.16:0.17[      1.0   \n",
      "1  0.965995  021U: [0.00:0.00[ AND degree_centralization: [...      1.0   \n",
      "2  0.965995  021D: [0.00:0.00[ AND average_degree: [2.01:2.02[      1.0   \n",
      "3  0.965995           021U>=0.01 AND longest_path: [0.05:0.05[      1.0   \n",
      "4  0.965995  021U>=0.01 AND betweenness_centralization: [0....      1.0   \n",
      "5  0.965995  average_clustering: [0.00:0.00[ AND average_de...      1.0   \n",
      "6  0.965995           021U>=0.01 AND transitivity: [0.16:0.17[      1.0   \n",
      "7  0.965995  average_degree: [2.06:2.07[ AND transitivity: ...      1.0   \n",
      "8  0.965995  degree_centralization>=0.83 AND transitivity: ...      1.0   \n",
      "9  0.965995  betweenness_centralization: [0.00:0.00[ AND de...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0           1.0              334.0           9821.0   \n",
      "1        9822.0           1.0              334.0           9821.0   \n",
      "2        9822.0           1.0              334.0           9821.0   \n",
      "3        9822.0           1.0              334.0           9821.0   \n",
      "4        9822.0           1.0              334.0           9821.0   \n",
      "5        9822.0           1.0              334.0           9821.0   \n",
      "6        9822.0           1.0              334.0           9821.0   \n",
      "7        9822.0           1.0              334.0           9821.0   \n",
      "8        9822.0           1.0              334.0           9821.0   \n",
      "9        9822.0           1.0              334.0           9821.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000102                  0.999898     0.002994   \n",
      "1          0.000102                  0.999898     0.002994   \n",
      "2          0.000102                  0.999898     0.002994   \n",
      "3          0.000102                  0.999898     0.002994   \n",
      "4          0.000102                  0.999898     0.002994   \n",
      "5          0.000102                  0.999898     0.002994   \n",
      "6          0.000102                  0.999898     0.002994   \n",
      "7          0.000102                  0.999898     0.002994   \n",
      "8          0.000102                  0.999898     0.002994   \n",
      "9          0.000102                  0.999898     0.002994   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.997006              1.0                 0.033907   \n",
      "1             0.997006              1.0                 0.033907   \n",
      "2             0.997006              1.0                 0.033907   \n",
      "3             0.997006              1.0                 0.033907   \n",
      "4             0.997006              1.0                 0.033907   \n",
      "5             0.997006              1.0                 0.033907   \n",
      "6             0.997006              1.0                 0.033907   \n",
      "7             0.997006              1.0                 0.033907   \n",
      "8             0.997006              1.0                 0.033907   \n",
      "9             0.997006              1.0                 0.033907   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.034005  29.407186  \n",
      "1              0.034005  29.407186  \n",
      "2              0.034005  29.407186  \n",
      "3              0.034005  29.407186  \n",
      "4              0.034005  29.407186  \n",
      "5              0.034005  29.407186  \n",
      "6              0.034005  29.407186  \n",
      "7              0.034005  29.407186  \n",
      "8              0.034005  29.407186  \n",
      "9              0.034005  29.407186  \n",
      "250_24_SQF_0_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.965995  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "1  0.965995  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0   \n",
      "2  0.965995  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "3  0.965995  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "4  0.965995  012: [0.02:0.02[ AND 021U: [0.00:0.00[ AND 030...      1.0   \n",
      "5  0.965995  012: [0.01:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "6  0.965995  012: [0.01:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "7  0.965995  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0   \n",
      "8  0.965995  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0   \n",
      "9  0.965995  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0           1.0              334.0           9821.0   \n",
      "1        9822.0           1.0              334.0           9821.0   \n",
      "2        9822.0           1.0              334.0           9821.0   \n",
      "3        9822.0           1.0              334.0           9821.0   \n",
      "4        9822.0           1.0              334.0           9821.0   \n",
      "5        9822.0           1.0              334.0           9821.0   \n",
      "6        9822.0           1.0              334.0           9821.0   \n",
      "7        9822.0           1.0              334.0           9821.0   \n",
      "8        9822.0           1.0              334.0           9821.0   \n",
      "9        9822.0           1.0              334.0           9821.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.000102                  0.999898     0.002994   \n",
      "1          0.000102                  0.999898     0.002994   \n",
      "2          0.000102                  0.999898     0.002994   \n",
      "3          0.000102                  0.999898     0.002994   \n",
      "4          0.000102                  0.999898     0.002994   \n",
      "5          0.000102                  0.999898     0.002994   \n",
      "6          0.000102                  0.999898     0.002994   \n",
      "7          0.000102                  0.999898     0.002994   \n",
      "8          0.000102                  0.999898     0.002994   \n",
      "9          0.000102                  0.999898     0.002994   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.997006              1.0                 0.033907   \n",
      "1             0.997006              1.0                 0.033907   \n",
      "2             0.997006              1.0                 0.033907   \n",
      "3             0.997006              1.0                 0.033907   \n",
      "4             0.997006              1.0                 0.033907   \n",
      "5             0.997006              1.0                 0.033907   \n",
      "6             0.997006              1.0                 0.033907   \n",
      "7             0.997006              1.0                 0.033907   \n",
      "8             0.997006              1.0                 0.033907   \n",
      "9             0.997006              1.0                 0.033907   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.034005  29.407186  \n",
      "1              0.034005  29.407186  \n",
      "2              0.034005  29.407186  \n",
      "3              0.034005  29.407186  \n",
      "4              0.034005  29.407186  \n",
      "5              0.034005  29.407186  \n",
      "6              0.034005  29.407186  \n",
      "7              0.034005  29.407186  \n",
      "8              0.034005  29.407186  \n",
      "9              0.034005  29.407186  \n",
      "250_24_GA_0.5_SDFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.046804                        average_degree: [2.46:2.59[    327.0   \n",
      "1  0.040585                               average_degree>=2.59    328.0   \n",
      "2  0.029443                                         021C>=0.00    328.0   \n",
      "3  0.026717                                  021D: [0.00:0.00[    327.0   \n",
      "4  0.026100                                         021D>=0.00    328.0   \n",
      "5  0.025543                 degree_centralization: [0.18:0.19[    328.0   \n",
      "6  0.025109  021D: [0.00:0.00[ AND degree_centralization: [...     21.0   \n",
      "7  0.025043                 degree_centralization: [0.16:0.18[    327.0   \n",
      "8  0.024986                                  021D: [0.00:0.00[    328.0   \n",
      "9  0.024485                                  021D: [0.00:0.00[    327.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0          95.0              334.0           9495.0   \n",
      "1        9822.0          84.0              334.0           9494.0   \n",
      "2        9822.0          64.0              334.0           9494.0   \n",
      "3        9822.0          59.0              334.0           9495.0   \n",
      "4        9822.0          58.0              334.0           9494.0   \n",
      "5        9822.0          57.0              334.0           9494.0   \n",
      "6        9822.0          15.0              334.0           9801.0   \n",
      "7        9822.0          56.0              334.0           9495.0   \n",
      "8        9822.0          56.0              334.0           9494.0   \n",
      "9        9822.0          55.0              334.0           9495.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033293                  0.966707     0.284431   \n",
      "1          0.033394                  0.966606     0.251497   \n",
      "2          0.033394                  0.966606     0.191617   \n",
      "3          0.033293                  0.966707     0.176647   \n",
      "4          0.033394                  0.966606     0.173653   \n",
      "5          0.033394                  0.966606     0.170659   \n",
      "6          0.002138                  0.997862     0.044910   \n",
      "7          0.033293                  0.966707     0.167665   \n",
      "8          0.033394                  0.966606     0.167665   \n",
      "9          0.033293                  0.966707     0.164671   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.715569         0.290520                 0.025171   \n",
      "1             0.748503         0.256098                 0.026332   \n",
      "2             0.808383         0.195122                 0.028439   \n",
      "3             0.823353         0.180428                 0.028963   \n",
      "4             0.826347         0.176829                 0.029071   \n",
      "5             0.829341         0.173780                 0.029176   \n",
      "6             0.955090         0.714286                 0.032548   \n",
      "7             0.832335         0.171254                 0.029279   \n",
      "8             0.832335         0.170732                 0.029282   \n",
      "9             0.835329         0.168196                 0.029384   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.034005   8.543372  \n",
      "1              0.034005   7.531109  \n",
      "2              0.034005   5.737987  \n",
      "3              0.034005   5.305884  \n",
      "4              0.034005   5.200051  \n",
      "5              0.034005   5.110395  \n",
      "6              0.034005  21.005133  \n",
      "7              0.034005   5.036093  \n",
      "8              0.034005   5.020739  \n",
      "9              0.034005   4.946163  \n",
      "250_24_SQF_0.5_BFS\n",
      "    quality                                           subgroup  size_sg  \\\n",
      "0  0.046804                        average_degree: [2.46:2.59[    327.0   \n",
      "1  0.040585                               average_degree>=2.59    328.0   \n",
      "2  0.038104  021D: [0.00:0.00[ AND average_degree: [2.46:2.59[     52.0   \n",
      "3  0.036069  average_degree: [2.46:2.59[ AND degree_central...     34.0   \n",
      "4  0.036064                021C>=0.00 AND average_degree>=2.59    162.0   \n",
      "5  0.035517  021U: [0.00:0.00[ AND average_degree: [2.46:2.59[     32.0   \n",
      "6  0.035517  average_degree>=2.59 AND degree_centralization...     32.0   \n",
      "7  0.035393  021C>=0.00 AND average_degree>=2.59 AND degree...     24.0   \n",
      "8  0.033591  021D: [0.00:0.00[ AND average_degree: [2.46:2.59[     69.0   \n",
      "9  0.032018  average_clustering: [0.02:0.03[ AND average_de...     23.0   \n",
      "\n",
      "   size_dataset  positives_sg  positives_dataset  size_complement  \\\n",
      "0        9822.0          95.0              334.0           9495.0   \n",
      "1        9822.0          84.0              334.0           9494.0   \n",
      "2        9822.0          29.0              334.0           9770.0   \n",
      "3        9822.0          22.0              334.0           9788.0   \n",
      "4        9822.0          51.0              334.0           9660.0   \n",
      "5        9822.0          21.0              334.0           9790.0   \n",
      "6        9822.0          21.0              334.0           9790.0   \n",
      "7        9822.0          18.0              334.0           9798.0   \n",
      "8        9822.0          30.0              334.0           9753.0   \n",
      "9        9822.0          16.0              334.0           9799.0   \n",
      "\n",
      "   relative_size_sg  relative_size_complement  coverage_sg  \\\n",
      "0          0.033293                  0.966707     0.284431   \n",
      "1          0.033394                  0.966606     0.251497   \n",
      "2          0.005294                  0.994706     0.086826   \n",
      "3          0.003462                  0.996538     0.065868   \n",
      "4          0.016494                  0.983506     0.152695   \n",
      "5          0.003258                  0.996742     0.062874   \n",
      "6          0.003258                  0.996742     0.062874   \n",
      "7          0.002443                  0.997557     0.053892   \n",
      "8          0.007025                  0.992975     0.089820   \n",
      "9          0.002342                  0.997658     0.047904   \n",
      "\n",
      "   coverage_complement  target_share_sg  target_share_complement  \\\n",
      "0             0.715569         0.290520                 0.025171   \n",
      "1             0.748503         0.256098                 0.026332   \n",
      "2             0.913174         0.557692                 0.031218   \n",
      "3             0.934132         0.647059                 0.031876   \n",
      "4             0.847305         0.314815                 0.029296   \n",
      "5             0.937126         0.656250                 0.031971   \n",
      "6             0.937126         0.656250                 0.031971   \n",
      "7             0.946108         0.750000                 0.032251   \n",
      "8             0.910180         0.434783                 0.031170   \n",
      "9             0.952096         0.695652                 0.032452   \n",
      "\n",
      "   target_share_dataset       lift  \n",
      "0              0.034005   8.543372  \n",
      "1              0.034005   7.531109  \n",
      "2              0.034005  16.400161  \n",
      "3              0.034005  19.028179  \n",
      "4              0.034005   9.257818  \n",
      "5              0.034005  19.298466  \n",
      "6              0.034005  19.298466  \n",
      "7              0.034005  22.055389  \n",
      "8              0.034005  12.785733  \n",
      "9              0.034005  20.457173  \n",
      "250_24_GA_1_SDFS\n",
      "    quality                            subgroup  size_sg  size_dataset  \\\n",
      "0  0.008540         average_degree: [2.46:2.59[    327.0        9822.0   \n",
      "1  0.007417                average_degree>=2.59    328.0        9822.0   \n",
      "2  0.005380                          021C>=0.00    328.0        9822.0   \n",
      "3  0.004875                   021D: [0.00:0.00[    327.0        9822.0   \n",
      "4  0.004770                          021D>=0.00    328.0        9822.0   \n",
      "5  0.004668  degree_centralization: [0.18:0.19[    328.0        9822.0   \n",
      "6  0.004569  degree_centralization: [0.16:0.18[    327.0        9822.0   \n",
      "7  0.004566                   021D: [0.00:0.00[    328.0        9822.0   \n",
      "8  0.004468                   021D: [0.00:0.00[    327.0        9822.0   \n",
      "9  0.004260                   021U: [0.00:0.00[    328.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0          95.0              334.0           9495.0          0.033293   \n",
      "1          84.0              334.0           9494.0          0.033394   \n",
      "2          64.0              334.0           9494.0          0.033394   \n",
      "3          59.0              334.0           9495.0          0.033293   \n",
      "4          58.0              334.0           9494.0          0.033394   \n",
      "5          57.0              334.0           9494.0          0.033394   \n",
      "6          56.0              334.0           9495.0          0.033293   \n",
      "7          56.0              334.0           9494.0          0.033394   \n",
      "8          55.0              334.0           9495.0          0.033293   \n",
      "9          53.0              334.0           9494.0          0.033394   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.284431             0.715569   \n",
      "1                  0.966606     0.251497             0.748503   \n",
      "2                  0.966606     0.191617             0.808383   \n",
      "3                  0.966707     0.176647             0.823353   \n",
      "4                  0.966606     0.173653             0.826347   \n",
      "5                  0.966606     0.170659             0.829341   \n",
      "6                  0.966707     0.167665             0.832335   \n",
      "7                  0.966606     0.167665             0.832335   \n",
      "8                  0.966707     0.164671             0.835329   \n",
      "9                  0.966606     0.158683             0.841317   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.290520                 0.025171              0.034005  8.543372  \n",
      "1         0.256098                 0.026332              0.034005  7.531109  \n",
      "2         0.195122                 0.028439              0.034005  5.737987  \n",
      "3         0.180428                 0.028963              0.034005  5.305884  \n",
      "4         0.176829                 0.029071              0.034005  5.200051  \n",
      "5         0.173780                 0.029176              0.034005  5.110395  \n",
      "6         0.171254                 0.029279              0.034005  5.036093  \n",
      "7         0.170732                 0.029282              0.034005  5.020739  \n",
      "8         0.168196                 0.029384              0.034005  4.946163  \n",
      "9         0.161585                 0.029598              0.034005  4.751771  \n",
      "250_24_SQF_1_BFS\n",
      "    quality                             subgroup  size_sg  size_dataset  \\\n",
      "0  0.008540          average_degree: [2.46:2.59[    327.0        9822.0   \n",
      "1  0.007417                 average_degree>=2.59    328.0        9822.0   \n",
      "2  0.005380                           021C>=0.00    328.0        9822.0   \n",
      "3  0.004875                    021D: [0.00:0.00[    327.0        9822.0   \n",
      "4  0.004770                           021D>=0.00    328.0        9822.0   \n",
      "5  0.004668   degree_centralization: [0.18:0.19[    328.0        9822.0   \n",
      "6  0.004632  021C>=0.00 AND average_degree>=2.59    162.0        9822.0   \n",
      "7  0.004569   degree_centralization: [0.16:0.18[    327.0        9822.0   \n",
      "8  0.004566                    021D: [0.00:0.00[    328.0        9822.0   \n",
      "9  0.004468                    021D: [0.00:0.00[    327.0        9822.0   \n",
      "\n",
      "   positives_sg  positives_dataset  size_complement  relative_size_sg  \\\n",
      "0          95.0              334.0           9495.0          0.033293   \n",
      "1          84.0              334.0           9494.0          0.033394   \n",
      "2          64.0              334.0           9494.0          0.033394   \n",
      "3          59.0              334.0           9495.0          0.033293   \n",
      "4          58.0              334.0           9494.0          0.033394   \n",
      "5          57.0              334.0           9494.0          0.033394   \n",
      "6          51.0              334.0           9660.0          0.016494   \n",
      "7          56.0              334.0           9495.0          0.033293   \n",
      "8          56.0              334.0           9494.0          0.033394   \n",
      "9          55.0              334.0           9495.0          0.033293   \n",
      "\n",
      "   relative_size_complement  coverage_sg  coverage_complement  \\\n",
      "0                  0.966707     0.284431             0.715569   \n",
      "1                  0.966606     0.251497             0.748503   \n",
      "2                  0.966606     0.191617             0.808383   \n",
      "3                  0.966707     0.176647             0.823353   \n",
      "4                  0.966606     0.173653             0.826347   \n",
      "5                  0.966606     0.170659             0.829341   \n",
      "6                  0.983506     0.152695             0.847305   \n",
      "7                  0.966707     0.167665             0.832335   \n",
      "8                  0.966606     0.167665             0.832335   \n",
      "9                  0.966707     0.164671             0.835329   \n",
      "\n",
      "   target_share_sg  target_share_complement  target_share_dataset      lift  \n",
      "0         0.290520                 0.025171              0.034005  8.543372  \n",
      "1         0.256098                 0.026332              0.034005  7.531109  \n",
      "2         0.195122                 0.028439              0.034005  5.737987  \n",
      "3         0.180428                 0.028963              0.034005  5.305884  \n",
      "4         0.176829                 0.029071              0.034005  5.200051  \n",
      "5         0.173780                 0.029176              0.034005  5.110395  \n",
      "6         0.314815                 0.029296              0.034005  9.257818  \n",
      "7         0.171254                 0.029279              0.034005  5.036093  \n",
      "8         0.170732                 0.029282              0.034005  5.020739  \n",
      "9         0.168196                 0.029384              0.034005  4.946163  \n"
     ]
    }
   ],
   "source": [
    "result_dict_sg_topic\n",
    "for res in result_dict_sg_topic.keys():\n",
    "    print(res)\n",
    "    pprint(result_dict_sg_topic[res])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Cluster als abhängige Variable"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "10fc4b5e9b0740b1918c7ea8d7510dfd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/14 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10 50\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 175, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.12      0.18      0.14      1004\n",
      "           1       0.06      0.09      0.07       639\n",
      "           2       0.04      0.07      0.05        68\n",
      "           3       0.08      0.31      0.12       556\n",
      "           4       0.81      0.48      0.60     10362\n",
      "           5       0.03      0.15      0.06       141\n",
      "           6       0.04      0.05      0.04       578\n",
      "           7       0.01      0.24      0.02        33\n",
      "\n",
      "    accuracy                           0.41     13381\n",
      "   macro avg       0.15      0.20      0.14     13381\n",
      "weighted avg       0.64      0.41      0.49     13381\n",
      "\n",
      "10 100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py:691: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 200, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.15      0.22      0.17      1297\n",
      "           1       0.07      0.08      0.08       830\n",
      "           2       0.11      0.31      0.17       175\n",
      "           3       0.11      0.34      0.16       742\n",
      "           4       0.81      0.50      0.62     11894\n",
      "           5       0.07      0.22      0.11       244\n",
      "           6       0.05      0.10      0.07       682\n",
      "           7       0.01      0.20      0.02        50\n",
      "\n",
      "    accuracy                           0.42     15914\n",
      "   macro avg       0.17      0.25      0.17     15914\n",
      "weighted avg       0.63      0.42      0.49     15914\n",
      "\n",
      "10 350\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/joblib/externals/loky/process_executor.py:691: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.\n",
      "  \"timeout or by a memory leak.\", UserWarning\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': {4: 1.743922430330299e-05, 0: 0.00015019525382997898, 1: 0.00023696682464454977, 6: 0.00033500837520938025, 3: 0.00023980815347721823, 2: 0.00016204829039053638, 5: 0.0005399568034557236, 7: 0.001466275659824047}, 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 175, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.16      0.19      0.18      1664\n",
      "           1       0.08      0.07      0.07      1055\n",
      "           2       0.67      0.77      0.72      1543\n",
      "           3       0.17      0.34      0.22      1043\n",
      "           4       0.78      0.60      0.68     14335\n",
      "           5       0.26      0.31      0.28       463\n",
      "           6       0.05      0.15      0.08       746\n",
      "           7       0.16      0.49      0.24       171\n",
      "\n",
      "    accuracy                           0.52     21020\n",
      "   macro avg       0.29      0.36      0.31     21020\n",
      "weighted avg       0.62      0.52      0.56     21020\n",
      "\n",
      "50 100\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 10, 'max_features': 'auto', 'n_estimators': 100, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.16      0.25      0.20       304\n",
      "           1       0.14      0.10      0.11       198\n",
      "           2       0.15      0.31      0.20       110\n",
      "           3       0.15      0.41      0.22       194\n",
      "           4       0.71      0.39      0.50      1587\n",
      "           5       0.20      0.29      0.24       105\n",
      "           6       0.04      0.09      0.06       107\n",
      "           7       0.17      0.28      0.21        18\n",
      "\n",
      "    accuracy                           0.33      2623\n",
      "   macro avg       0.22      0.26      0.22      2623\n",
      "weighted avg       0.49      0.33      0.37      2623\n",
      "\n",
      "50 150\n",
      "{'class_weight': {0: 0.0006097560975609756, 4: 0.00011396011396011396, 6: 0.00186219739292365, 5: 0.0015503875968992248, 2: 0.0011918951132300357, 3: 0.0008952551477170994, 7: 0.008849557522123894, 1: 0.0009124087591240876}, 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 175, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.20      0.20      0.20       410\n",
      "           1       0.11      0.06      0.08       274\n",
      "           2       0.31      0.34      0.33       210\n",
      "           3       0.18      0.44      0.25       279\n",
      "           4       0.68      0.59      0.63      2194\n",
      "           5       0.23      0.17      0.20       161\n",
      "           6       0.04      0.04      0.04       134\n",
      "           7       0.19      0.28      0.23        29\n",
      "\n",
      "    accuracy                           0.44      3691\n",
      "   macro avg       0.24      0.26      0.24      3691\n",
      "weighted avg       0.48      0.44      0.45      3691\n",
      "\n",
      "50 250\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 200, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.17      0.21      0.19       528\n",
      "           1       0.11      0.08      0.09       342\n",
      "           2       0.58      0.59      0.59       712\n",
      "           3       0.22      0.44      0.29       374\n",
      "           4       0.71      0.58      0.64      3385\n",
      "           5       0.27      0.28      0.27       213\n",
      "           6       0.07      0.10      0.08       162\n",
      "           7       0.16      0.24      0.19        70\n",
      "\n",
      "    accuracy                           0.48      5786\n",
      "   macro avg       0.28      0.32      0.29      5786\n",
      "weighted avg       0.54      0.48      0.50      5786\n",
      "\n",
      "50 350\n",
      "{'class_weight': {4: 6.205013651030033e-05, 0: 0.00037257824143070045, 2: 0.00016923337282112033, 6: 0.001455604075691412, 1: 0.0005913660555884093, 3: 0.0005055611729019212, 7: 0.0018115942028985507, 5: 0.0007710100231303007}, 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 175, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.20      0.15      0.17       671\n",
      "           1       0.13      0.06      0.09       423\n",
      "           2       0.74      0.74      0.74      1478\n",
      "           3       0.26      0.43      0.32       494\n",
      "           4       0.68      0.70      0.69      4029\n",
      "           5       0.52      0.31      0.39       324\n",
      "           6       0.05      0.06      0.06       172\n",
      "           7       0.44      0.52      0.48       138\n",
      "\n",
      "    accuracy                           0.58      7729\n",
      "   macro avg       0.38      0.37      0.37      7729\n",
      "weighted avg       0.57      0.58      0.57      7729\n",
      "\n",
      "100 150\n",
      "{'class_weight': {3: 0.002840909090909091, 0: 0.0023148148148148147, 5: 0.0043859649122807015, 4: 0.0004012841091492777, 2: 0.002457002457002457, 1: 0.003215434083601286, 6: 0.009174311926605505, 7: 0.023255813953488372}, 'criterion': 'gini', 'max_depth': 10, 'max_features': 'auto', 'n_estimators': 150, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.11      0.10      0.11       108\n",
      "           1       0.09      0.12      0.10        78\n",
      "           2       0.25      0.41      0.31       102\n",
      "           3       0.25      0.53      0.34        88\n",
      "           4       0.69      0.50      0.58       623\n",
      "           5       0.27      0.26      0.27        57\n",
      "           6       0.00      0.00      0.00        27\n",
      "           7       0.20      0.18      0.19        11\n",
      "\n",
      "    accuracy                           0.40      1094\n",
      "   macro avg       0.23      0.26      0.24      1094\n",
      "weighted avg       0.47      0.40      0.42      1094\n",
      "\n",
      "100 250\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 10, 'max_features': 'auto', 'n_estimators': 200, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.12      0.18      0.14       226\n",
      "           1       0.11      0.12      0.12       146\n",
      "           2       0.58      0.65      0.61       604\n",
      "           3       0.25      0.57      0.35       183\n",
      "           4       0.75      0.47      0.58      1814\n",
      "           5       0.23      0.45      0.31       109\n",
      "           6       0.03      0.02      0.02        55\n",
      "           7       0.09      0.35      0.14        52\n",
      "\n",
      "    accuracy                           0.46      3189\n",
      "   macro avg       0.27      0.35      0.28      3189\n",
      "weighted avg       0.57      0.46      0.49      3189\n",
      "\n",
      "100 350\n",
      "{'class_weight': {0: 0.0006775067750677507, 2: 0.00018258170531312764, 5: 0.0011363636363636363, 4: 0.00010169836265636123, 1: 0.0011037527593818985, 7: 0.002074688796680498, 3: 0.0008244023083264633, 6: 0.003861003861003861}, 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 125, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.24      0.11      0.15       369\n",
      "           1       0.18      0.07      0.11       227\n",
      "           2       0.77      0.75      0.76      1370\n",
      "           3       0.32      0.45      0.37       303\n",
      "           4       0.68      0.79      0.73      2458\n",
      "           5       0.67      0.48      0.56       220\n",
      "           6       0.00      0.00      0.00        65\n",
      "           7       0.56      0.52      0.54       120\n",
      "\n",
      "    accuracy                           0.65      5132\n",
      "   macro avg       0.43      0.40      0.40      5132\n",
      "weighted avg       0.62      0.65      0.63      5132\n",
      "\n",
      "150 250\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 13, 'max_features': 'auto', 'n_estimators': 100, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.12      0.04      0.06       120\n",
      "           1       0.11      0.03      0.05        69\n",
      "           2       0.73      0.68      0.71       504\n",
      "           3       0.33      0.49      0.39        95\n",
      "           4       0.71      0.79      0.75      1199\n",
      "           5       0.67      0.15      0.25        53\n",
      "           6       0.00      0.00      0.00        28\n",
      "           7       0.12      0.24      0.16        42\n",
      "\n",
      "    accuracy                           0.65      2110\n",
      "   macro avg       0.35      0.30      0.30      2110\n",
      "weighted avg       0.62      0.65      0.63      2110\n",
      "\n",
      "150 350\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 15, 'max_features': 'auto', 'n_estimators': 125, 'random_state': 42}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.27      0.14      0.18       263\n",
      "           1       0.11      0.04      0.06       150\n",
      "           2       0.81      0.76      0.78      1269\n",
      "           3       0.38      0.56      0.46       216\n",
      "           4       0.69      0.80      0.74      1843\n",
      "           5       0.73      0.47      0.57       164\n",
      "           6       0.00      0.00      0.00        38\n",
      "           7       0.60      0.58      0.59       110\n",
      "\n",
      "    accuracy                           0.68      4053\n",
      "   macro avg       0.45      0.42      0.42      4053\n",
      "weighted avg       0.65      0.68      0.66      4053\n",
      "\n",
      "200 350\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 10, 'max_features': 'auto', 'n_estimators': 100, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.19      0.29      0.23       188\n",
      "           1       0.11      0.14      0.12       106\n",
      "           2       0.82      0.81      0.81      1126\n",
      "           3       0.34      0.59      0.43       162\n",
      "           4       0.70      0.57      0.63      1139\n",
      "           5       0.65      0.63      0.64       137\n",
      "           6       0.00      0.00      0.00        19\n",
      "           7       0.66      0.68      0.67        80\n",
      "\n",
      "    accuracy                           0.63      2957\n",
      "   macro avg       0.43      0.46      0.44      2957\n",
      "weighted avg       0.67      0.63      0.64      2957\n",
      "\n",
      "250 350\n",
      "{'class_weight': 'balanced', 'criterion': 'gini', 'max_depth': 10, 'max_features': 'auto', 'n_estimators': 50, 'random_state': 42}\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.26      0.37      0.31       144\n",
      "           1       0.17      0.17      0.17        82\n",
      "           2       0.85      0.84      0.85       779\n",
      "           3       0.41      0.57      0.48       122\n",
      "           4       0.68      0.60      0.64       649\n",
      "           5       0.68      0.70      0.69       111\n",
      "           6       0.00      0.00      0.00        10\n",
      "           7       0.78      0.72      0.75        68\n",
      "\n",
      "    accuracy                           0.67      1965\n",
      "   macro avg       0.48      0.50      0.49      1965\n",
      "weighted avg       0.68      0.67      0.67      1965\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n",
      "/home/hengelhardt/masterarbeit/.venv/lib/python3.6/site-packages/sklearn/metrics/_classification.py:1245: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    }
   ],
   "source": [
    "graph_features_all = clean_features_cluster\n",
    "graph_features_all['replies'] = stat_df.set_index('thread_id').replies\n",
    "scaler = StandardScaler()\n",
    "\n",
    "\n",
    "clf = RandomForestClassifier(random_state=42)\n",
    "\n",
    "cv = StratifiedKFold(\n",
    "    n_splits=5\n",
    ")\n",
    "cv_results_rf_cluster = {}\n",
    "for min_replies, max_replies in tqdm([(10, 50), (10, 100), (10, 350), (50,100), \n",
    "                                      (50, 150), (50, 250), (50, 350), (100, 150), \n",
    "                                      (100, 250), (100, 350), (150,250), (150, 350), \n",
    "                                      (200, 350), (250, 350)]):\n",
    "    print(min_replies, max_replies)\n",
    "    filter_array = filter_on_stats(stat_df, min_replies=min_replies, max_replies=max_replies)\n",
    "    graph_features = graph_features_all.loc[filter_array]\n",
    "    graph_features = graph_features.dropna()\n",
    "    X = graph_features.drop(columns=[\"km\", \"replies\", \"dominant_topic\"]).to_numpy()\n",
    "    y = graph_features.km.to_numpy()\n",
    "    scaler.fit(X)\n",
    "    X_transformed = scaler.transform(X)\n",
    "    X_train, X_test, y_train, y_test = train_test_split(X_transformed, y, test_size=0.2, \n",
    "                                                        random_state=42, stratify=y)\n",
    "    count = Counter(y_train)\n",
    "    \n",
    "    param_grid = { \n",
    "    'n_estimators': [50, 100, 125, 150, 175, 200],\n",
    "    'max_features': ['auto'],\n",
    "    'max_depth' : [10, 13, 15, None],\n",
    "    'criterion' :['gini'],\n",
    "    'class_weight': ['balanced', {label:1/count[label] for label in count.keys()}],\n",
    "    'random_state': [42]\n",
    "    }\n",
    "\n",
    "    search = GridSearchCV(\n",
    "        estimator=clf, param_grid=param_grid,\n",
    "        scoring='f1_macro', cv=cv,\n",
    "        n_jobs=20, refit = True\n",
    "    )\n",
    "    search.fit(X_train, y_train)\n",
    "    print(search.best_params_)\n",
    "    key = f\"{min_replies}-{max_replies}\"\n",
    "    cv_results_rf_cluster[key] = {\"model\":search, \n",
    "                      'X_test': X_test, \n",
    "                      'X_train': X_train, \n",
    "                      'y_test':y_test, \n",
    "                      \"y_train\":y_train\n",
    "                     }\n",
    "    cv_results_rf_cluster[key]['report'] = classification_report(cv_results_rf_cluster[key]['y_test'], cv_results_rf_cluster[key]['model'].predict(cv_results_rf_cluster[key]['X_test']))\n",
    "    cv_results_rf_cluster[key]['report_dict'] = classification_report(cv_results_rf_cluster[key]['y_test'], cv_results_rf_cluster[key]['model'].predict(cv_results_rf_cluster[key]['X_test']), output_dict=True)\n",
    "    cv_results_rf_cluster[key]['best_params'] = search.best_params_\n",
    "    print(classification_report(cv_results_rf_cluster[key]['y_test'], cv_results_rf_cluster[key]['model'].predict(cv_results_rf_cluster[key]['X_test'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"b_collection_extracted/rf_cv_results_rf_cluster\", \"wb\") as f:\n",
    "    pickle.dump(cv_results_rf_cluster, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"b_collection_extracted/rf_cv_results_rf_cluster\", \"rb\") as f:\n",
    "    cv_results_rf_cluster = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [],
   "source": [
    "multi_dict = {}\n",
    "param_dict = {}\n",
    "for key in cv_results_rf_cluster.keys():\n",
    "    multi_dict[key] = pd.DataFrame(cv_results_rf_cluster[key]['report_dict'])\n",
    "    param_dict[key] = cv_results_rf_cluster[key]['best_params']\n",
    "df = pd.concat(multi_dict.values(), keys=multi_dict.keys())\n",
    "param_df = pd.DataFrame(param_dict)\n",
    "df.round(2).to_excel(\"b_collection_extracted/classification_report_rf_cluster.xlsx\")\n",
    "param_df.to_excel(\"b_collection_extracted/classification_params_rf_cluster.xlsx\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [],
   "source": [
    "viz_dict_rf_cluster = {\"precision\":{},\n",
    "                        \"precision_weighted\":{},\n",
    "                        \"recall\":{},\n",
    "                        \"recall_weighted\":{},\n",
    "                        \"f1\":{},\n",
    "                        \"f1_weigthed\":{}\n",
    "                       }\n",
    "for key in cv_results_rf_cluster.keys():\n",
    "    df =  pd.DataFrame(cv_results_rf_cluster[key]['report_dict']).transpose()\n",
    "    viz_dict_rf_cluster[\"precision\"][key] = df.precision.loc[\"macro avg\"]\n",
    "    viz_dict_rf_cluster[\"precision_weighted\"][key] = df.precision.loc[\"weighted avg\"]\n",
    "    viz_dict_rf_cluster[\"recall\"][key] = df.recall.loc[\"macro avg\"]\n",
    "    viz_dict_rf_cluster[\"recall_weighted\"][key] = df.recall.loc[\"weighted avg\"]\n",
    "    viz_dict_rf_cluster[\"f1\"][key] = df[\"f1-score\"].loc[\"macro avg\"]\n",
    "    viz_dict_rf_cluster[\"f1_weigthed\"][key] = df[\"f1-score\"].loc[\"weighted avg\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f332dfc4940>"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x504 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "plt.figure(figsize=(12,7))\n",
    "sns.lineplot(data=viz_dict_rf_cluster).set_xlabel(\"Spanne an Antworten\")\n",
    "plt.legend([\"Genauigkeit\", \"Genauigkeit gewichtet\", \"Sensitivität\", \"Sensitivität gewichtet\", \"F1\", \"F1 gewichtet\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame(cv_results_rf_cluster[\"10-350\"]['report_dict']).transpose().round(2).join(pd.DataFrame(cv_results_rf_cluster[\"250-350\"]['report_dict']).transpose().round(2), lsuffix=\" (10-350)\", rsuffix=\" (250-350)\", how=\"left\").to_excel(\"b_collection_extracted/classification_report_rf_cluster_10-350_250-350.xlsx\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Text(0, 0, 'Beziehung'),\n",
       " Text(1, 0, 'Rassismus'),\n",
       " Text(2, 0, 'Social'),\n",
       " Text(3, 0, 'Politik'),\n",
       " Text(4, 0, 'Bilder'),\n",
       " Text(5, 0, 'roll'),\n",
       " Text(6, 0, 'Konsum'),\n",
       " Text(7, 0, 'kik')]"
      ]
     },
     "execution_count": 330,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x360 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\", {'axes.grid' : False})\n",
    "fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14,5))\n",
    "g1=plot_confusion_matrix(estimator=cv_results_rf_cluster[\"10-350\"][\"model\"], X=cv_results_rf_cluster[\"10-350\"][\"X_test\"], y_true=cv_results_rf_cluster[\"10-350\"][\"y_test\"],cmap=\"magma\", ax=ax1, display_labels=[\"Beziehung\", \"Rassismus\", \"Social\", \"Politik\",\"Bilder\", \"roll\", \"Konsum\", \"kik\"])\n",
    "g1.ax_.set_xlabel('Vorhergesagtes Cluster')\n",
    "g1.ax_.set_ylabel('Tatsächliches Cluster')\n",
    "g1.ax_.set_title(\"Threads mit 10-350 Antworten\")\n",
    "g1.ax_.tick_params(axis='x', labelrotation=30)\n",
    "g1.ax_.set_xticklabels([\"Beziehung\", \"Rassismus\", \"Social\", \"Politik\",\"Bilder\", \"roll\", \"Konsum\", \"kik\"], horizontalalignment=\"right\")\n",
    "g2=plot_confusion_matrix(estimator=cv_results_rf_cluster[\"250-350\"][\"model\"], X=cv_results_rf_cluster[\"250-350\"][\"X_test\"], y_true=cv_results_rf_cluster[\"250-350\"][\"y_test\"],cmap=\"magma\", ax=ax2, display_labels=[\"Beziehung\", \"Rassismus\", \"Social\", \"Politik\",\"Bilder\", \"roll\", \"Konsum\", \"kik\"])\n",
    "g2.ax_.set_xlabel('Vorhergesagtes Cluster')\n",
    "g2.ax_.set_ylabel('Tatsächliches Cluster')\n",
    "g2.ax_.set_title(\"Threads mit 250-350 Antworten\")\n",
    "g2.ax_.tick_params(axis='x', labelrotation=30)\n",
    "g2.ax_.set_xticklabels([\"Beziehung\", \"Rassismus\", \"Social\", \"Politik\",\"Bilder\", \"roll\", \"Konsum\", \"kik\"], horizontalalignment=\"right\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Explorative Analyse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## binäre logistische Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e9ccdcb88f00408a94d9c845076dbb72",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10 350\n",
      "10 350\n",
      "7714\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.135757\n",
      "         Iterations 20\n",
      "10 350\n",
      "71677\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.567960\n",
      "         Iterations 7\n",
      "10 350\n",
      "2315\n",
      "Optimization terminated successfully.\n",
      "         Current function value: 0.079520\n",
      "         Iterations 13\n"
     ]
    }
   ],
   "source": [
    "scaler = StandardScaler()\n",
    "\n",
    "result_dict_stats_binary_cluster = {}\n",
    "for min_replies, max_replies in tqdm([(10, 350)]):\n",
    "    print(min_replies, max_replies)\n",
    "    filter_array = filter_on_stats(stat_df, min_replies=min_replies)\n",
    "    graph_features = clean_features_cluster.loc[filter_array]\n",
    "    graph_features = graph_features.dropna()\n",
    "    X = graph_features.drop(columns=[\"km\", \"replies\", \"dominant_topic\"]).astype(np.float64)\n",
    "    y = graph_features.km \n",
    "    scaler.fit(X)\n",
    "    X_transformed = scaler.transform(X)\n",
    "    key = f\"{min_replies}-{max_replies}\"\n",
    "    result_dict_stats_binary_cluster[key] = {}\n",
    "    try:\n",
    "        temp_dict = {}\n",
    "        for dominant_topic in [2, 4, 5] if min_replies == 10 else [2, 5, 7]:\n",
    "            y_train_binary = y == dominant_topic\n",
    "            print(min_replies, max_replies)\n",
    "            print(y_train_binary.sum())\n",
    "            model = sm.Logit(endog=y_train_binary, exog=sm.add_constant(X))\n",
    "            result = model.fit()\n",
    "            temp_dict[dominant_topic] = {'model': model,\n",
    "                                'result': result,\n",
    "                                }\n",
    "        result_dict_stats_binary_cluster[key]['predictors'] = temp_dict\n",
    "    except Exception as e:\n",
    "        print(f\"EXCEPTION!!! {e} !!!\")\n",
    "        continue"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2\n",
      "4\n",
      "5\n"
     ]
    }
   ],
   "source": [
    "for min_replies in [10]:\n",
    "    for key in result_dict_stats_binary_cluster[f\"{min_replies}-350\"][\"predictors\"].keys():\n",
    "        print(key)\n",
    "        pd.DataFrame(result_dict_stats_binary_cluster[f\"{min_replies}-350\"][\"predictors\"][key][\"result\"].summary().tables[1]).transpose().to_excel(f\"b_collection_extracted/log_reg_result_cluster_{min_replies}_top_{key}.xlsx\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Subgroups von Clustern"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "563a27549b3b4a4d9f70b3a1cc9bcc3b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "adding search to mp pool:   0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "31df89640c1c44139b020e2f324bdc04",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "retrieving subgroups:   0%|          | 0/9 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "subgroup_dominant_topic = clean_features_cluster.drop(columns=[\"replies\", \"dominant_topic\"])\n",
    "subgroup_combs = []\n",
    "[subgroup_combs.append(prod) for prod in product([10], [\"km\"], [2, 4, 5], [0, 0.5, 1])]\n",
    "[subgroup_combs.append(prod) for prod in product([250], [\"km\"], [2, 5, 7], [0, 0.5, 1])]\n",
    "pool = mp.Pool(processes=18)\n",
    "sg_cluster_results = [pool.apply_async(search_subgroups, (subgroup_dominant_topic.loc[filter_on_stats(stat_df, min_replies=item[0])], item[0], item[1], item[2], item[3],))\n",
    "           for item in tqdm(subgroup_combs,desc=\"adding search to mp pool\")]\n",
    "sg_cluster_results = [result.get() for result in tqdm(sg_cluster_results, desc=\"retrieving subgroups\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "250_2_SQF_0_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "1  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "2  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "3             012: [0.02:0.02[ AND 021C: [0.00:0.00[      1.0        9822.0   \n",
      "4  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "5  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "6  012: [0.02:0.02[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "7  012: [0.02:0.02[ AND 021U: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "8  012: [0.02:0.02[ AND 021D: [0.00:0.00[ AND 021...      1.0        9822.0   \n",
      "9             012: [0.02:0.02[ AND 030T: [0.00:0.00[      1.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0              1.0              0.396151           1.0             3891.0   \n",
      "1              1.0              0.396151           1.0             3891.0   \n",
      "2              1.0              0.396151           1.0             3891.0   \n",
      "3              1.0              0.396151           1.0             3891.0   \n",
      "4              1.0              0.396151           1.0             3891.0   \n",
      "5              1.0              0.396151           1.0             3891.0   \n",
      "6              1.0              0.396151           1.0             3891.0   \n",
      "7              1.0              0.396151           1.0             3891.0   \n",
      "8              1.0              0.396151           1.0             3891.0   \n",
      "9              1.0              0.396151           1.0             3891.0   \n",
      "\n",
      "       lift  \n",
      "0  2.524287  \n",
      "1  2.524287  \n",
      "2  2.524287  \n",
      "3  2.524287  \n",
      "4  2.524287  \n",
      "5  2.524287  \n",
      "6  2.524287  \n",
      "7  2.524287  \n",
      "8  2.524287  \n",
      "9  2.524287  \n",
      "250_2_SQF_0.5_BFS\n",
      "                          subgroup  size_sg  size_dataset  target_share_sg  \\\n",
      "0  average_clustering: [0.07:0.08[    327.0        9822.0         0.929664   \n",
      "1         average_clustering>=0.08    328.0        9822.0         0.905488   \n",
      "2  average_clustering: [0.06:0.07[    328.0        9822.0         0.899390   \n",
      "3  average_clustering: [0.05:0.06[    327.0        9822.0         0.880734   \n",
      "4  average_clustering: [0.05:0.05[    327.0        9822.0         0.840979   \n",
      "5                030T: [0.00:0.00[    329.0        9822.0         0.835866   \n",
      "6                030T: [0.00:0.00[    327.0        9822.0         0.819572   \n",
      "7  average_clustering: [0.04:0.05[    328.0        9822.0         0.807927   \n",
      "8        transitivity: [0.28:0.30[    294.0        9822.0         0.823129   \n",
      "9                030T: [0.00:0.00[    327.0        9822.0         0.792049   \n",
      "\n",
      "   target_share_dataset  positives_sg  positives_dataset      lift  \n",
      "0              0.396151         304.0             3891.0  2.346738  \n",
      "1              0.396151         297.0             3891.0  2.285711  \n",
      "2              0.396151         295.0             3891.0  2.270319  \n",
      "3              0.396151         288.0             3891.0  2.223225  \n",
      "4              0.396151         275.0             3891.0  2.122871  \n",
      "5              0.396151         275.0             3891.0  2.109966  \n",
      "6              0.396151         268.0             3891.0  2.068834  \n",
      "7              0.396151         265.0             3891.0  2.039439  \n",
      "8              0.396151         242.0             3891.0  2.077814  \n",
      "9              0.396151         259.0             3891.0  1.999359  \n",
      "250_2_SQF_1_BFS\n",
      "                          subgroup  size_sg  size_dataset  target_share_sg  \\\n",
      "0  average_clustering: [0.07:0.08[    327.0        9822.0         0.929664   \n",
      "1         average_clustering>=0.08    328.0        9822.0         0.905488   \n",
      "2  average_clustering: [0.06:0.07[    328.0        9822.0         0.899390   \n",
      "3  average_clustering: [0.05:0.06[    327.0        9822.0         0.880734   \n",
      "4  average_clustering: [0.05:0.05[    327.0        9822.0         0.840979   \n",
      "5                030T: [0.00:0.00[    329.0        9822.0         0.835866   \n",
      "6                030T: [0.00:0.00[    327.0        9822.0         0.819572   \n",
      "7  average_clustering: [0.04:0.05[    328.0        9822.0         0.807927   \n",
      "8        transitivity: [0.30:0.33[    366.0        9822.0         0.759563   \n",
      "9                030T: [0.00:0.00[    327.0        9822.0         0.792049   \n",
      "\n",
      "   target_share_dataset  positives_sg  positives_dataset      lift  \n",
      "0              0.396151         304.0             3891.0  2.346738  \n",
      "1              0.396151         297.0             3891.0  2.285711  \n",
      "2              0.396151         295.0             3891.0  2.270319  \n",
      "3              0.396151         288.0             3891.0  2.223225  \n",
      "4              0.396151         275.0             3891.0  2.122871  \n",
      "5              0.396151         275.0             3891.0  2.109966  \n",
      "6              0.396151         268.0             3891.0  2.068834  \n",
      "7              0.396151         265.0             3891.0  2.039439  \n",
      "8              0.396151         278.0             3891.0  1.917354  \n",
      "9              0.396151         259.0             3891.0  1.999359  \n",
      "250_5_SQF_0_BFS\n",
      "                                 subgroup  size_sg  size_dataset  \\\n",
      "0  012: [0.01:0.01[ AND 021D: [0.00:0.00[      2.0        9822.0   \n",
      "1  012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0        9822.0   \n",
      "2  012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0        9822.0   \n",
      "3  012: [0.01:0.01[ AND 021D: [0.00:0.00[      3.0        9822.0   \n",
      "4  012: [0.01:0.01[ AND 021D: [0.00:0.00[      4.0        9822.0   \n",
      "5  012: [0.01:0.01[ AND 021D: [0.00:0.00[      2.0        9822.0   \n",
      "6  012: [0.01:0.01[ AND 021D: [0.00:0.00[      2.0        9822.0   \n",
      "7  012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0        9822.0   \n",
      "8  012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0        9822.0   \n",
      "9  012: [0.01:0.01[ AND 021D: [0.00:0.00[      5.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0              1.0              0.056709           2.0              557.0   \n",
      "1              1.0              0.056709           1.0              557.0   \n",
      "2              1.0              0.056709           1.0              557.0   \n",
      "3              1.0              0.056709           3.0              557.0   \n",
      "4              1.0              0.056709           4.0              557.0   \n",
      "5              1.0              0.056709           2.0              557.0   \n",
      "6              1.0              0.056709           2.0              557.0   \n",
      "7              1.0              0.056709           1.0              557.0   \n",
      "8              1.0              0.056709           1.0              557.0   \n",
      "9              1.0              0.056709           5.0              557.0   \n",
      "\n",
      "        lift  \n",
      "0  17.633752  \n",
      "1  17.633752  \n",
      "2  17.633752  \n",
      "3  17.633752  \n",
      "4  17.633752  \n",
      "5  17.633752  \n",
      "6  17.633752  \n",
      "7  17.633752  \n",
      "8  17.633752  \n",
      "9  17.633752  \n",
      "250_5_SQF_0.5_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                                  longest_path<0.02    326.0        9822.0   \n",
      "1                                         021U>=0.01    328.0        9822.0   \n",
      "2                        degree_centralization>=0.83    328.0        9822.0   \n",
      "3         021U>=0.01 AND degree_centralization>=0.83    301.0        9822.0   \n",
      "4           012<0.01 AND degree_centralization>=0.83    312.0        9822.0   \n",
      "5  012<0.01 AND 021U>=0.01 AND degree_centralizat...    288.0        9822.0   \n",
      "6                            012<0.01 AND 021U>=0.01    289.0        9822.0   \n",
      "7                                           012<0.01    327.0        9822.0   \n",
      "8  021C<0.00 AND 021U>=0.01 AND degree_centraliza...    272.0        9822.0   \n",
      "9          021C<0.00 AND degree_centralization>=0.83    294.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.677914              0.056709         221.0              557.0   \n",
      "1         0.661585              0.056709         217.0              557.0   \n",
      "2         0.658537              0.056709         216.0              557.0   \n",
      "3         0.677741              0.056709         204.0              557.0   \n",
      "4         0.657051              0.056709         205.0              557.0   \n",
      "5         0.680556              0.056709         196.0              557.0   \n",
      "6         0.678201              0.056709         196.0              557.0   \n",
      "7         0.639144              0.056709         209.0              557.0   \n",
      "8         0.683824              0.056709         186.0              557.0   \n",
      "9         0.659864              0.056709         194.0              557.0   \n",
      "\n",
      "        lift  \n",
      "0  11.954169  \n",
      "1  11.666232  \n",
      "2  11.612471  \n",
      "3  11.951114  \n",
      "4  11.586280  \n",
      "5  12.000748  \n",
      "6  11.959223  \n",
      "7  11.270502  \n",
      "8  12.058375  \n",
      "9  11.635877  \n",
      "250_5_SQF_1_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                                  longest_path<0.02    326.0        9822.0   \n",
      "1                                         021U>=0.01    328.0        9822.0   \n",
      "2                        degree_centralization>=0.83    328.0        9822.0   \n",
      "3                                           012<0.01    327.0        9822.0   \n",
      "4           012<0.01 AND degree_centralization>=0.83    312.0        9822.0   \n",
      "5         021U>=0.01 AND degree_centralization>=0.83    301.0        9822.0   \n",
      "6                    betweenness_centralization<0.00    327.0        9822.0   \n",
      "7                                          021C<0.00    327.0        9822.0   \n",
      "8  012<0.01 AND 021U>=0.01 AND degree_centralizat...    288.0        9822.0   \n",
      "9                            012<0.01 AND 021U>=0.01    289.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.677914              0.056709         221.0              557.0   \n",
      "1         0.661585              0.056709         217.0              557.0   \n",
      "2         0.658537              0.056709         216.0              557.0   \n",
      "3         0.639144              0.056709         209.0              557.0   \n",
      "4         0.657051              0.056709         205.0              557.0   \n",
      "5         0.677741              0.056709         204.0              557.0   \n",
      "6         0.626911              0.056709         205.0              557.0   \n",
      "7         0.623853              0.056709         204.0              557.0   \n",
      "8         0.680556              0.056709         196.0              557.0   \n",
      "9         0.678201              0.056709         196.0              557.0   \n",
      "\n",
      "        lift  \n",
      "0  11.954169  \n",
      "1  11.666232  \n",
      "2  11.612471  \n",
      "3  11.270502  \n",
      "4  11.586280  \n",
      "5  11.951114  \n",
      "6  11.054799  \n",
      "7  11.000873  \n",
      "8  12.000748  \n",
      "9  11.959223  \n",
      "250_7_SQF_0_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "1  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND 030...      2.0        9822.0   \n",
      "2  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "3  012: [0.01:0.01[ AND 021C<0.00 AND 030T: [0.00...      3.0        9822.0   \n",
      "4  012: [0.01:0.01[ AND 021C<0.00 AND 021D: [0.00...      1.0        9822.0   \n",
      "5  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "6  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "7  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "8             012: [0.01:0.01[ AND 021D: [0.00:0.00[      1.0        9822.0   \n",
      "9  012: [0.01:0.01[ AND 021D: [0.00:0.00[ AND 030...      1.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0              1.0              0.034616           1.0              340.0   \n",
      "1              1.0              0.034616           2.0              340.0   \n",
      "2              1.0              0.034616           1.0              340.0   \n",
      "3              1.0              0.034616           3.0              340.0   \n",
      "4              1.0              0.034616           1.0              340.0   \n",
      "5              1.0              0.034616           1.0              340.0   \n",
      "6              1.0              0.034616           1.0              340.0   \n",
      "7              1.0              0.034616           1.0              340.0   \n",
      "8              1.0              0.034616           1.0              340.0   \n",
      "9              1.0              0.034616           1.0              340.0   \n",
      "\n",
      "        lift  \n",
      "0  28.888235  \n",
      "1  28.888235  \n",
      "2  28.888235  \n",
      "3  28.888235  \n",
      "4  28.888235  \n",
      "5  28.888235  \n",
      "6  28.888235  \n",
      "7  28.888235  \n",
      "8  28.888235  \n",
      "9  28.888235  \n",
      "250_7_SQF_0.5_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                                   012: [0.01:0.01[    327.0        9822.0   \n",
      "1                                  021C: [0.00:0.00[    327.0        9822.0   \n",
      "2                 degree_centralization: [0.72:0.83[    327.0        9822.0   \n",
      "3             012: [0.01:0.01[ AND 021C: [0.00:0.00[    290.0        9822.0   \n",
      "4                                  021U: [0.01:0.01[    327.0        9822.0   \n",
      "5  012: [0.01:0.01[ AND degree_centralization: [0...    282.0        9822.0   \n",
      "6  021U: [0.01:0.01[ AND degree_centralization: [...    283.0        9822.0   \n",
      "7  021C: [0.00:0.00[ AND degree_centralization: [...    251.0        9822.0   \n",
      "8  012: [0.01:0.01[ AND 021U: [0.01:0.01[ AND deg...    242.0        9822.0   \n",
      "9             012: [0.01:0.01[ AND 021U: [0.01:0.01[    245.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.581040              0.034616         190.0              340.0   \n",
      "1         0.574924              0.034616         188.0              340.0   \n",
      "2         0.571865              0.034616         187.0              340.0   \n",
      "3         0.596552              0.034616         173.0              340.0   \n",
      "4         0.562691              0.034616         184.0              340.0   \n",
      "5         0.602837              0.034616         170.0              340.0   \n",
      "6         0.597173              0.034616         169.0              340.0   \n",
      "7         0.625498              0.034616         157.0              340.0   \n",
      "8         0.636364              0.034616         154.0              340.0   \n",
      "9         0.628571              0.034616         154.0              340.0   \n",
      "\n",
      "        lift  \n",
      "0  16.785213  \n",
      "1  16.608527  \n",
      "2  16.520183  \n",
      "3  17.233327  \n",
      "4  16.255154  \n",
      "5  17.414894  \n",
      "6  17.251278  \n",
      "7  18.069534  \n",
      "8  18.383422  \n",
      "9  18.158319  \n",
      "250_7_SQF_1_BFS\n",
      "                                            subgroup  size_sg  size_dataset  \\\n",
      "0                                   012: [0.01:0.01[    327.0        9822.0   \n",
      "1                                  021C: [0.00:0.00[    327.0        9822.0   \n",
      "2                 degree_centralization: [0.72:0.83[    327.0        9822.0   \n",
      "3                                  021U: [0.01:0.01[    327.0        9822.0   \n",
      "4             012: [0.01:0.01[ AND 021C: [0.00:0.00[    290.0        9822.0   \n",
      "5  012: [0.01:0.01[ AND degree_centralization: [0...    282.0        9822.0   \n",
      "6  021U: [0.01:0.01[ AND degree_centralization: [...    283.0        9822.0   \n",
      "7  021C: [0.00:0.00[ AND degree_centralization: [...    251.0        9822.0   \n",
      "8  012: [0.01:0.01[ AND 021C: [0.00:0.00[ AND deg...    249.0        9822.0   \n",
      "9  012: [0.01:0.01[ AND 021U: [0.01:0.01[ AND deg...    242.0        9822.0   \n",
      "\n",
      "   target_share_sg  target_share_dataset  positives_sg  positives_dataset  \\\n",
      "0         0.581040              0.034616         190.0              340.0   \n",
      "1         0.574924              0.034616         188.0              340.0   \n",
      "2         0.571865              0.034616         187.0              340.0   \n",
      "3         0.562691              0.034616         184.0              340.0   \n",
      "4         0.596552              0.034616         173.0              340.0   \n",
      "5         0.602837              0.034616         170.0              340.0   \n",
      "6         0.597173              0.034616         169.0              340.0   \n",
      "7         0.625498              0.034616         157.0              340.0   \n",
      "8         0.622490              0.034616         155.0              340.0   \n",
      "9         0.636364              0.034616         154.0              340.0   \n",
      "\n",
      "        lift  \n",
      "0  16.785213  \n",
      "1  16.608527  \n",
      "2  16.520183  \n",
      "3  16.255154  \n",
      "4  17.233327  \n",
      "5  17.414894  \n",
      "6  17.251278  \n",
      "7  18.069534  \n",
      "8  17.982636  \n",
      "9  18.383422  \n"
     ]
    }
   ],
   "source": [
    "\n",
    "for sg_res in sg_cluster_results:\n",
    "    for res in sg_res.keys():\n",
    "        print(res)\n",
    "        pprint(sg_res[res][[\"subgroup\", \"size_sg\", \"size_dataset\", \"target_share_sg\", \"target_share_dataset\", \"positives_sg\", \"positives_dataset\", \"lift\"]])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}