02_Erg_cluster_singlematrix-CombinedEndpoint.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "Bad key \"text.kerning_factor\" on line 4 in\n",
      "/scratch/ss2686/envs/my-rdkit-env/lib/python3.6/site-packages/matplotlib/mpl-data/stylelib/_classic_test_patch.mplstyle.\n",
      "You probably need to get an updated matplotlibrc file from\n",
      "https://github.com/matplotlib/matplotlib/blob/v3.1.3/matplotlibrc.template\n",
      "or from the matplotlib source distribution\n"
     ]
    }
   ],
   "source": [
    "from collections import defaultdict\n",
    "\n",
    "import matplotlib.pyplot as pyplot\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from scipy.stats import pearsonr\n",
    "from scipy.cluster import hierarchy\n",
    "\n",
    "from sklearn.datasets import load_breast_cancer\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.model_selection import StratifiedKFold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.io import arff\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import balanced_accuracy_score\n",
    "from sklearn.metrics import matthews_corrcoef\n",
    "from scipy.io import arff\n",
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve, auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>Cells_AreaShape_Area</th>\n",
       "      <th>Cells_AreaShape_Center_X</th>\n",
       "      <th>Cells_AreaShape_Center_Y</th>\n",
       "      <th>Cells_AreaShape_Compactness</th>\n",
       "      <th>Cells_AreaShape_Eccentricity</th>\n",
       "      <th>Cells_AreaShape_Extent</th>\n",
       "      <th>Cells_AreaShape_FormFactor</th>\n",
       "      <th>Cells_AreaShape_MajorAxisLength</th>\n",
       "      <th>Cells_AreaShape_MaxFeretDiameter</th>\n",
       "      <th>...</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_3_0</th>\n",
       "      <th>Nuclei_Texture_Variance_ER_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_Mito_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_Mito_3_0</th>\n",
       "      <th>Nuclei_Texture_Variance_Mito_5_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_10_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_3_0</th>\n",
       "      <th>Nuclei_Texture_Variance_RNA_5_0</th>\n",
       "      <th>smiles_r</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>-53.322266</td>\n",
       "      <td>15.138672</td>\n",
       "      <td>-7.832031</td>\n",
       "      <td>0.020546</td>\n",
       "      <td>0.011831</td>\n",
       "      <td>-0.005042</td>\n",
       "      <td>0.006361</td>\n",
       "      <td>1.567715</td>\n",
       "      <td>1.506277</td>\n",
       "      <td>...</td>\n",
       "      <td>0.007785</td>\n",
       "      <td>0.019312</td>\n",
       "      <td>0.020843</td>\n",
       "      <td>-0.064662</td>\n",
       "      <td>-0.072990</td>\n",
       "      <td>-0.084042</td>\n",
       "      <td>0.215716</td>\n",
       "      <td>0.155296</td>\n",
       "      <td>0.176099</td>\n",
       "      <td>Brc1c(NC2=NCCN2)ccc2nccnc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>41.678711</td>\n",
       "      <td>1.973633</td>\n",
       "      <td>-8.867188</td>\n",
       "      <td>0.008705</td>\n",
       "      <td>0.008857</td>\n",
       "      <td>-0.001514</td>\n",
       "      <td>-0.009787</td>\n",
       "      <td>0.946463</td>\n",
       "      <td>1.437743</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.018887</td>\n",
       "      <td>-0.003844</td>\n",
       "      <td>-0.019000</td>\n",
       "      <td>0.007360</td>\n",
       "      <td>0.025757</td>\n",
       "      <td>0.027188</td>\n",
       "      <td>0.088701</td>\n",
       "      <td>0.052283</td>\n",
       "      <td>0.062916</td>\n",
       "      <td>c1ccc(C(O[C@@H]2CCCN(CCc3ccc4c(c3)OCO4)C2)c2cc...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>418.492188</td>\n",
       "      <td>6.976562</td>\n",
       "      <td>7.766602</td>\n",
       "      <td>0.023997</td>\n",
       "      <td>0.005404</td>\n",
       "      <td>-0.005869</td>\n",
       "      <td>-0.024194</td>\n",
       "      <td>6.252890</td>\n",
       "      <td>6.685650</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.105881</td>\n",
       "      <td>0.011543</td>\n",
       "      <td>0.004370</td>\n",
       "      <td>-0.208441</td>\n",
       "      <td>-0.081369</td>\n",
       "      <td>-0.084740</td>\n",
       "      <td>-0.095445</td>\n",
       "      <td>-0.015152</td>\n",
       "      <td>-0.045636</td>\n",
       "      <td>c1cnc(N2CCN(Cc3ccc4c(c3)OCO4)CC2)nc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>-10.445312</td>\n",
       "      <td>0.101562</td>\n",
       "      <td>-0.045898</td>\n",
       "      <td>-0.003597</td>\n",
       "      <td>0.000294</td>\n",
       "      <td>0.002599</td>\n",
       "      <td>-0.001359</td>\n",
       "      <td>-0.473211</td>\n",
       "      <td>-0.001053</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.024427</td>\n",
       "      <td>-0.053814</td>\n",
       "      <td>-0.040165</td>\n",
       "      <td>-0.066490</td>\n",
       "      <td>-0.036203</td>\n",
       "      <td>-0.040287</td>\n",
       "      <td>0.040549</td>\n",
       "      <td>0.026975</td>\n",
       "      <td>0.027425</td>\n",
       "      <td>c1ccc(C2CC2)c(OCC2=NCCN2)c1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>48.655273</td>\n",
       "      <td>1.666992</td>\n",
       "      <td>2.486328</td>\n",
       "      <td>0.018529</td>\n",
       "      <td>0.011256</td>\n",
       "      <td>-0.002420</td>\n",
       "      <td>0.004862</td>\n",
       "      <td>1.776159</td>\n",
       "      <td>1.611745</td>\n",
       "      <td>...</td>\n",
       "      <td>0.116695</td>\n",
       "      <td>0.172022</td>\n",
       "      <td>0.164453</td>\n",
       "      <td>-0.012553</td>\n",
       "      <td>0.000800</td>\n",
       "      <td>-0.010659</td>\n",
       "      <td>0.088510</td>\n",
       "      <td>0.083851</td>\n",
       "      <td>0.076598</td>\n",
       "      <td>C(=C/c1ccccc1)\\CN1CCN(C(c2ccccc2)c2ccccc2)CC1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>641</th>\n",
       "      <td>641</td>\n",
       "      <td>-133.445312</td>\n",
       "      <td>-2.585938</td>\n",
       "      <td>-4.170898</td>\n",
       "      <td>-0.006187</td>\n",
       "      <td>-0.002578</td>\n",
       "      <td>-0.001801</td>\n",
       "      <td>-0.007760</td>\n",
       "      <td>-2.365016</td>\n",
       "      <td>-2.547309</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.069310</td>\n",
       "      <td>-0.098329</td>\n",
       "      <td>-0.084651</td>\n",
       "      <td>-0.048275</td>\n",
       "      <td>-0.071216</td>\n",
       "      <td>-0.062632</td>\n",
       "      <td>-0.079043</td>\n",
       "      <td>-0.088448</td>\n",
       "      <td>-0.085335</td>\n",
       "      <td>c1ccc2c(c1)[nH]c1cnccc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>642</th>\n",
       "      <td>642</td>\n",
       "      <td>-26.469727</td>\n",
       "      <td>-0.145508</td>\n",
       "      <td>-0.763672</td>\n",
       "      <td>-0.006640</td>\n",
       "      <td>-0.000166</td>\n",
       "      <td>0.001977</td>\n",
       "      <td>-0.003229</td>\n",
       "      <td>-0.770855</td>\n",
       "      <td>-0.478971</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.040332</td>\n",
       "      <td>-0.035610</td>\n",
       "      <td>-0.022689</td>\n",
       "      <td>-0.003744</td>\n",
       "      <td>-0.012261</td>\n",
       "      <td>-0.003150</td>\n",
       "      <td>-0.006708</td>\n",
       "      <td>-0.009239</td>\n",
       "      <td>-0.004748</td>\n",
       "      <td>c1ccc2c(c1)cc1ccc3cccc4ccc2c1c34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>643</th>\n",
       "      <td>643</td>\n",
       "      <td>215.939732</td>\n",
       "      <td>2.107143</td>\n",
       "      <td>-7.992188</td>\n",
       "      <td>0.014079</td>\n",
       "      <td>0.009538</td>\n",
       "      <td>-0.005964</td>\n",
       "      <td>0.005382</td>\n",
       "      <td>3.785917</td>\n",
       "      <td>4.371880</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.084987</td>\n",
       "      <td>-0.049924</td>\n",
       "      <td>-0.072038</td>\n",
       "      <td>-0.012332</td>\n",
       "      <td>0.017893</td>\n",
       "      <td>0.006418</td>\n",
       "      <td>0.005525</td>\n",
       "      <td>0.026849</td>\n",
       "      <td>0.029421</td>\n",
       "      <td>c1ccc(-c2ccc(C(c3ccccc3)n3ccnc3)cc2)cc1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>644</th>\n",
       "      <td>644</td>\n",
       "      <td>1810.023438</td>\n",
       "      <td>4.320312</td>\n",
       "      <td>-1.496094</td>\n",
       "      <td>0.085425</td>\n",
       "      <td>0.025699</td>\n",
       "      <td>-0.030275</td>\n",
       "      <td>-0.063700</td>\n",
       "      <td>27.234844</td>\n",
       "      <td>31.637251</td>\n",
       "      <td>...</td>\n",
       "      <td>0.243208</td>\n",
       "      <td>0.345452</td>\n",
       "      <td>0.292939</td>\n",
       "      <td>0.108634</td>\n",
       "      <td>0.165323</td>\n",
       "      <td>0.152386</td>\n",
       "      <td>-0.089590</td>\n",
       "      <td>-0.020275</td>\n",
       "      <td>-0.033736</td>\n",
       "      <td>c1cnc2c(c1)ccc1cccnc12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>645</th>\n",
       "      <td>645</td>\n",
       "      <td>338.552734</td>\n",
       "      <td>-3.798828</td>\n",
       "      <td>-6.207031</td>\n",
       "      <td>0.046928</td>\n",
       "      <td>0.026305</td>\n",
       "      <td>-0.007931</td>\n",
       "      <td>0.013154</td>\n",
       "      <td>8.679539</td>\n",
       "      <td>8.436147</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.136895</td>\n",
       "      <td>-0.130600</td>\n",
       "      <td>-0.136309</td>\n",
       "      <td>-0.079303</td>\n",
       "      <td>-0.119351</td>\n",
       "      <td>-0.125692</td>\n",
       "      <td>-0.012512</td>\n",
       "      <td>-0.029041</td>\n",
       "      <td>-0.022962</td>\n",
       "      <td>c1ccc2[nH]c(-c3cscn3)nc2c1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>646 rows × 1768 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Unnamed: 0  Cells_AreaShape_Area  Cells_AreaShape_Center_X  \\\n",
       "0             0            -53.322266                 15.138672   \n",
       "1             1             41.678711                  1.973633   \n",
       "2             2            418.492188                  6.976562   \n",
       "3             3            -10.445312                  0.101562   \n",
       "4             4             48.655273                  1.666992   \n",
       "..          ...                   ...                       ...   \n",
       "641         641           -133.445312                 -2.585938   \n",
       "642         642            -26.469727                 -0.145508   \n",
       "643         643            215.939732                  2.107143   \n",
       "644         644           1810.023438                  4.320312   \n",
       "645         645            338.552734                 -3.798828   \n",
       "\n",
       "     Cells_AreaShape_Center_Y  Cells_AreaShape_Compactness  \\\n",
       "0                   -7.832031                     0.020546   \n",
       "1                   -8.867188                     0.008705   \n",
       "2                    7.766602                     0.023997   \n",
       "3                   -0.045898                    -0.003597   \n",
       "4                    2.486328                     0.018529   \n",
       "..                        ...                          ...   \n",
       "641                 -4.170898                    -0.006187   \n",
       "642                 -0.763672                    -0.006640   \n",
       "643                 -7.992188                     0.014079   \n",
       "644                 -1.496094                     0.085425   \n",
       "645                 -6.207031                     0.046928   \n",
       "\n",
       "     Cells_AreaShape_Eccentricity  Cells_AreaShape_Extent  \\\n",
       "0                        0.011831               -0.005042   \n",
       "1                        0.008857               -0.001514   \n",
       "2                        0.005404               -0.005869   \n",
       "3                        0.000294                0.002599   \n",
       "4                        0.011256               -0.002420   \n",
       "..                            ...                     ...   \n",
       "641                     -0.002578               -0.001801   \n",
       "642                     -0.000166                0.001977   \n",
       "643                      0.009538               -0.005964   \n",
       "644                      0.025699               -0.030275   \n",
       "645                      0.026305               -0.007931   \n",
       "\n",
       "     Cells_AreaShape_FormFactor  Cells_AreaShape_MajorAxisLength  \\\n",
       "0                      0.006361                         1.567715   \n",
       "1                     -0.009787                         0.946463   \n",
       "2                     -0.024194                         6.252890   \n",
       "3                     -0.001359                        -0.473211   \n",
       "4                      0.004862                         1.776159   \n",
       "..                          ...                              ...   \n",
       "641                   -0.007760                        -2.365016   \n",
       "642                   -0.003229                        -0.770855   \n",
       "643                    0.005382                         3.785917   \n",
       "644                   -0.063700                        27.234844   \n",
       "645                    0.013154                         8.679539   \n",
       "\n",
       "     Cells_AreaShape_MaxFeretDiameter  ...  Nuclei_Texture_Variance_ER_10_0  \\\n",
       "0                            1.506277  ...                         0.007785   \n",
       "1                            1.437743  ...                        -0.018887   \n",
       "2                            6.685650  ...                        -0.105881   \n",
       "3                           -0.001053  ...                        -0.024427   \n",
       "4                            1.611745  ...                         0.116695   \n",
       "..                                ...  ...                              ...   \n",
       "641                         -2.547309  ...                        -0.069310   \n",
       "642                         -0.478971  ...                        -0.040332   \n",
       "643                          4.371880  ...                        -0.084987   \n",
       "644                         31.637251  ...                         0.243208   \n",
       "645                          8.436147  ...                        -0.136895   \n",
       "\n",
       "     Nuclei_Texture_Variance_ER_3_0  Nuclei_Texture_Variance_ER_5_0  \\\n",
       "0                          0.019312                        0.020843   \n",
       "1                         -0.003844                       -0.019000   \n",
       "2                          0.011543                        0.004370   \n",
       "3                         -0.053814                       -0.040165   \n",
       "4                          0.172022                        0.164453   \n",
       "..                              ...                             ...   \n",
       "641                       -0.098329                       -0.084651   \n",
       "642                       -0.035610                       -0.022689   \n",
       "643                       -0.049924                       -0.072038   \n",
       "644                        0.345452                        0.292939   \n",
       "645                       -0.130600                       -0.136309   \n",
       "\n",
       "     Nuclei_Texture_Variance_Mito_10_0  Nuclei_Texture_Variance_Mito_3_0  \\\n",
       "0                            -0.064662                         -0.072990   \n",
       "1                             0.007360                          0.025757   \n",
       "2                            -0.208441                         -0.081369   \n",
       "3                            -0.066490                         -0.036203   \n",
       "4                            -0.012553                          0.000800   \n",
       "..                                 ...                               ...   \n",
       "641                          -0.048275                         -0.071216   \n",
       "642                          -0.003744                         -0.012261   \n",
       "643                          -0.012332                          0.017893   \n",
       "644                           0.108634                          0.165323   \n",
       "645                          -0.079303                         -0.119351   \n",
       "\n",
       "     Nuclei_Texture_Variance_Mito_5_0  Nuclei_Texture_Variance_RNA_10_0  \\\n",
       "0                           -0.084042                          0.215716   \n",
       "1                            0.027188                          0.088701   \n",
       "2                           -0.084740                         -0.095445   \n",
       "3                           -0.040287                          0.040549   \n",
       "4                           -0.010659                          0.088510   \n",
       "..                                ...                               ...   \n",
       "641                         -0.062632                         -0.079043   \n",
       "642                         -0.003150                         -0.006708   \n",
       "643                          0.006418                          0.005525   \n",
       "644                          0.152386                         -0.089590   \n",
       "645                         -0.125692                         -0.012512   \n",
       "\n",
       "     Nuclei_Texture_Variance_RNA_3_0  Nuclei_Texture_Variance_RNA_5_0  \\\n",
       "0                           0.155296                         0.176099   \n",
       "1                           0.052283                         0.062916   \n",
       "2                          -0.015152                        -0.045636   \n",
       "3                           0.026975                         0.027425   \n",
       "4                           0.083851                         0.076598   \n",
       "..                               ...                              ...   \n",
       "641                        -0.088448                        -0.085335   \n",
       "642                        -0.009239                        -0.004748   \n",
       "643                         0.026849                         0.029421   \n",
       "644                        -0.020275                        -0.033736   \n",
       "645                        -0.029041                        -0.022962   \n",
       "\n",
       "                                              smiles_r  \n",
       "0                          Brc1c(NC2=NCCN2)ccc2nccnc12  \n",
       "1    c1ccc(C(O[C@@H]2CCCN(CCc3ccc4c(c3)OCO4)C2)c2cc...  \n",
       "2                 c1cnc(N2CCN(Cc3ccc4c(c3)OCO4)CC2)nc1  \n",
       "3                          c1ccc(C2CC2)c(OCC2=NCCN2)c1  \n",
       "4        C(=C/c1ccccc1)\\CN1CCN(C(c2ccccc2)c2ccccc2)CC1  \n",
       "..                                                 ...  \n",
       "641                           c1ccc2c(c1)[nH]c1cnccc12  \n",
       "642                   c1ccc2c(c1)cc1ccc3cccc4ccc2c1c34  \n",
       "643            c1ccc(-c2ccc(C(c3ccccc3)n3ccnc3)cc2)cc1  \n",
       "644                             c1cnc2c(c1)ccc1cccnc12  \n",
       "645                         c1ccc2[nH]c(-c3cscn3)nc2c1  \n",
       "\n",
       "[646 rows x 1768 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cp = pd.read_csv(\"../../CenterMeanOVLP.csv\") \n",
    "cp = cp.loc[:,cp.apply(pd.Series.nunique) != 1]\n",
    "cp\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>combined_assay</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>629</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>633</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>637</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>640</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>641</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>135 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     combined_assay\n",
       "9               1.0\n",
       "12              0.0\n",
       "15              1.0\n",
       "16              1.0\n",
       "19              1.0\n",
       "..              ...\n",
       "629             1.0\n",
       "633             0.0\n",
       "637             1.0\n",
       "640             1.0\n",
       "641             0.0\n",
       "\n",
       "[135 rows x 1 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "toxcast = pd.read_csv(\"../../toxcast_Overlap.csv\") \n",
    "\n",
    "toxcast=toxcast.reindex(columns={'BSK_hDFCGF_Proliferation_down',\n",
    "             'APR_HepG2_CellLoss_72h_dn',\n",
    "             'BSK_3C_Proliferation_down',\n",
    "             'BSK_SAg_Proliferation_down',\n",
    "             'BSK_3C_SRB_down',\n",
    "             'BSK_3C_Vis_down',\n",
    "             'APR_HepG2_CellLoss_24h_dn',\n",
    "             'BSK_LPS_SRB_down',\n",
    "             'BSK_CASM3C_Proliferation_down',\n",
    "             'BSK_SAg_SRB_down',\n",
    "             'BSK_4H_SRB_down',\n",
    "             'ACEA_T47D_80hr_Negative','smiles_r' })\n",
    "\n",
    "toxcast=toxcast.dropna(thresh=(len(toxcast.columns)-11))\n",
    "\n",
    "toxcast=toxcast.fillna(0)\n",
    "toxcast['combined_assay']=   toxcast.BSK_SAg_Proliferation_down + toxcast.BSK_LPS_SRB_down + toxcast.ACEA_T47D_80hr_Negative + toxcast.BSK_3C_Vis_down+toxcast.APR_HepG2_CellLoss_24h_dn+toxcast.BSK_CASM3C_Proliferation_down+toxcast.BSK_4H_SRB_down+toxcast.BSK_hDFCGF_Proliferation_down+toxcast.BSK_3C_Proliferation_down+toxcast.APR_HepG2_CellLoss_72h_dn+toxcast.BSK_SAg_SRB_down+toxcast.BSK_3C_SRB_down\n",
    "\n",
    "toxcast=toxcast[['combined_assay', 'smiles_r']]\n",
    "\n",
    "toxcast['combined_assay']=toxcast['combined_assay']/12\n",
    "\n",
    "toxcast['combined_assay']=toxcast['combined_assay'].apply(np.ceil)\n",
    "\n",
    "tox_list=['combined_assay']\n",
    "\n",
    "our_data=toxcast.reindex(columns=tox_list)\n",
    "our_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from rdkit import Chem\n",
    "from rdkit.Chem import AllChem\n",
    "from rdkit.DataStructs import cDataStructs\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "import rdkit.Chem.Descriptors\n",
    "from rdkit import Chem\n",
    "from rdkit.Chem.EState import Fingerprinter\n",
    "from rdkit.Chem import Descriptors\n",
    "from rdkit.Chem.rdmolops import RDKFingerprint\n",
    "from rdkit.Chem.rdMolDescriptors import *\n",
    "from rdkit.Chem.EState.Fingerprinter import FingerprintMol\n",
    "from rdkit.Avalon.pyAvalonTools import GetAvalonFP #GetAvalonCountFP  #int vector version\n",
    "from rdkit.Chem.AllChem import  GetMorganFingerprintAsBitVect, GetErGFingerprint\n",
    "from rdkit.DataStructs.cDataStructs import ConvertToNumpyArray\n",
    "import rdkit.DataStructs.cDataStructs\n",
    "from rdkit.Avalon.pyAvalonTools import GetAvalonFP\n",
    "\n",
    "def ErGFingerprint(s):\n",
    "    x = Chem.MolFromSmiles(s)\n",
    "    return (AllChem.GetErGFingerprint(x))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import VarianceThreshold\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#importing libraries\n",
    "from sklearn.datasets import load_boston\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "%matplotlib inline\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.feature_selection import RFE\n",
    "from sklearn.linear_model import RidgeCV, LassoCV, Ridge, Lasso"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_selection import VarianceThreshold\n",
    "from numpy  import array\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy import stats\n",
    "from scipy.stats import ks_2samp \n",
    "#Calculates a point biserial correlation coefficient and the associated p-value.\n",
    "\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import LabelBinarizer # one hot encoding\n",
    "from sklearn.preprocessing import PolynomialFeatures # add polynomial features\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "import seaborn as sns\n",
    " #Grid Search with Cross Validation RF\n",
    "from sklearn import metrics\n",
    "\n",
    "\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "# Create the parameter grid based on the results of random search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets, metrics, model_selection, svm\n",
    "from sklearn import svm, datasets\n",
    "from sklearn.metrics import auc\n",
    "from sklearn.metrics import plot_roc_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import datasets, metrics, model_selection, svm\n",
    "from sklearn import svm, datasets\n",
    "from sklearn.metrics import auc\n",
    "from sklearn.metrics import plot_roc_curve\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import gridspec\n",
    "\n",
    "from rdkit import Chem, DataStructs\n",
    "from rdkit.Chem import Descriptors,Crippen\n",
    "\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "import matplotlib.cm as cm\n",
    "from sklearn.metrics import silhouette_samples, silhouette_score\n",
    "from sklearn.cluster import KMeans\n",
    "from sklearn.model_selection import LeaveOneGroupOut\n",
    "from sklearn.model_selection import GroupKFold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def GetExactMolWt(s):\n",
    "    mol = Chem.MolFromSmiles(s)\n",
    "    return (Descriptors.ExactMolWt (mol))\n",
    "def GetCalcTPSA(s):\n",
    "    mol = Chem.MolFromSmiles(s)\n",
    "    return (Chem.rdMolDescriptors.CalcTPSA(mol))\n",
    "def GetNumRotatableBonds(s):\n",
    "    mol = Chem.MolFromSmiles(s)\n",
    "    return (Descriptors.NumRotatableBonds (mol))\n",
    "def GetNumHDonors(s):\n",
    "    mol = Chem.MolFromSmiles(s)\n",
    "    return (Descriptors.NumHDonors(mol))\n",
    "def GetNumHAcceptors(s):\n",
    "    mol = Chem.MolFromSmiles(s)\n",
    "    return (Descriptors.NumHAcceptors(mol) )\n",
    "def GetMolLogP(s):\n",
    "    mol = Chem.MolFromSmiles(s)\n",
    "    return (Descriptors.MolLogP(mol))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start\n",
      "1.0    86\n",
      "0.0    49\n",
      "Name: combined_assay, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Compounds in Assay:  135\n",
      "pca explained variance by first two components:  0.7280338488123717\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start\n",
      "1    41\n",
      "3    36\n",
      "0    36\n",
      "2    22\n",
      "Name: Cluster_PC1_PC2, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEZCAYAAABxbJkKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deZgdVbnv8e+PMcwIaTBkBKIIciVyQsQLIiIoo3KOjAIiIIGDIArnCKgIKpOKoM/VozdIIMwggyiDGoUYQQUTZA5jBBIISQcIEOYk7/ljrYadTXfv6u49dKd+n+epJ7tWVa317kr1fnetWrtKEYGZmZXTcq0OwMzMWsdJwMysxJwEzMxKzEnAzKzEnATMzErMScDMrMScBPoJSb+QdHKd6hohaaGk5fP8FElfqkfdub6bJR1cr/p60O5pkuZLerbg+iFpdKPj6qn8f7PRstJOUfU6Dut9PJedk0ATSHpC0muSXpa0QNJfJR0p6e39HxFHRsT3Cta1Y3frRMRTEbF6RCyuQ+ynSrqkqv5dImJSX+vuYRzDgeOBzSLivXWue1ROGCvUs96u5P+bmUXW7Usi60k7ZSPpi5Jua3Uc/YGTQPPsERFrACOBs4ATgPPr3UizPshaYCTwXETMa3UgZsvU31lEeGrwBDwB7FhVNg5YAmye5y8ETsuvBwM3AAuA54G/kBL2xXmb14CFwNeBUUAAhwFPAVMrylbI9U0BzgTuBF4ErgfWycu2B2Z3Fi+wM/Am8FZu756K+r6UXy8HfAt4EpgHXASslZd1xHFwjm0+8M1u9tNaefv2XN+3cv075ve8JMdxYRfb/zcwB3gGODS3PTov2w34J/ASMAs4tWK7p/K6C/P0UWBj4BbguRz3pcDaVfvoJOBB4AXgAmBQxfLDgcfy/99vgA0qllXGdSHwM+BG4GXgDmDjvGxqXveVHNe+dHFsdLE/CrXTxbZbA3/N7dwDbF+x7BBgRq5nJnBE1bafBe7O+/pxYOeK4+Z7wO152z8Ag7uJobt6Oo6/U4FLKrYZxdLH/hdzjC8D/wIOADYFXgcW5/26IK+7MnB2Ph7mAr8AVqn8OyF9eXsWuLjVnyt1+3xqdQBlmOgkCeTyp4D/zK8v5J0kcGY+AFfM08cAdVZXxUF/EbAasEonfwhTgKeBzfM613T84dBNEsivl/ojq6iv44/wUNKH3UbA6sC1HX8gFXGcl+PaAngD2LSL/XQRKUGtkbd9BDisqzirtt05/+F2vMfLWPpDcHvg/5CSyofyuntWxblCRX2jgZ3yB0Mb6QP5x1X76H5gOLAO6YOt4/9vB1Li2DJv//+AqRXbVn84P0/6UrACKdlc0dm6tY6NTvZJ4XaqthtKSn675v21U55vy8t3IyVJAR8HXgW2zMvGkb5o7JS3HQp8oOK4eRx4fz4epgBndRFDrXpqJoF8HLwEbJKXDQE+mF9/Ebitqs0fkxL2OqRj8LfAmRXHzyLg+/n/dJVWf67Ua3J3UGs9Qzrgqr1FOmBHRsRbEfGXyEdiN06NiFci4rUull8cEfdHxCvAycA+HReO++gA4JyImBkRC0nfjverOl3+TkS8FhH3kL5VblFdSY5lX+CkiHg5Ip4AfgQcVDCOfYALKt7jqZULI2JKRNwXEUsi4l7gctIHWKci4rGImBwRb0REO3BOJ+v/NCJmRcTzwOnA/rn8AGBiRNwVEW+Q9slHJY3qorlrI+LOiFhE+nAe08377M2x0dN2DgRuioib8v6aDEwjJQUi4saIeDySP5O+0X8sb3tYfu+T87ZPR8RDFXVfEBGP5OP0qm5iqFVPUUuAzSWtEhFzIuKBzlaSJNLZ29ci4vmIeBk4A9ivqq5T8jHR1d/ZgOMk0FpDSd/Oqv2Q9O36D5JmSjqxQF2zerD8SdK3yMGFouzeBrm+yrpXANavKKsczfMq6Yyh2mBgpU7qGtqDOKrf49skfUTSrZLaJb0IHEk371/SepKukPS0pJeASzpZv7q9DSpiebv9nByf6+a9FNk/HXpzbPS0nZHA3nkQwwJJC4BtSckHSbtI+ruk5/OyXXln3wwnfdvvawy16qkpfxnYl/R/PUfSjZI+0MXqbcCqwPSK9/y7XN6hPSJe70tM/ZGTQItI2or0ofCuEQr5m/DxEbERsAdwnKRPdizuospa3waHV7weQfpGOZ/U37xqRVzLs/SBX6veZ0gfGpV1LyJ1t/TE/BxTdV1PF9x+Du9+j5UuI53qD4+ItUhdKsrLOnuPZ+byD0XEmqRvx6pap7q9Z/LrpfaJpNWAdXvwXrpU49iol1mkM8e1K6bVIuIsSSuTuhPPBtaPiLWBm3hn38widRXVI4Yi9Sx1/AJLjRyLiN9HxE6kBPYQqWsS3v1/Pp903emDFe95rYioTFLL5C2XnQSaTNKaknYHriD1Zd7XyTq7SxqdT1FfIl3A6hjuOZfU/95TB0raTNKqwHeBqyMNIX0EGCRpN0krki7Grlyx3VxgVOVw1iqXA1+TtKGk1Umn0FfmLofCcixXAadLWkPSSOA40jfwIq4CvljxHk+pWr4G8HxEvC5pHPD5imXtpFP9jarWXwgskDSUdNG52pclDZO0DvAN4MpcfhlwiKQx+UPzDOCO3MXVU0v9f9c4NurlEmAPSZ+WtLykQZK2lzSMdLa2MmmfLZK0C/Cpim3PJ733T0paTtLQbr59d6doPXcD2+XfxqxF6noDQNL6kj6Tk/AbpP/Pyr+jYZJWAoiIJaQEca6k9fL2QyV9uhexDyhOAs3zW0kvk77hfJPUx3xIF+u+D/gj6aD9G/A/ETElLzsT+FY+Zf2vHrR/Meni4LPAIOArABHxInAU8EvSN9VXSKMgOvwq//ucpLs6qXdirnsqafTF68AxPYir0jG5/ZmkM6TLcv01RcTNpAt7t5C6S26pWuUo4Lv5/+DbpKTRse2rpD792/N+3Rr4DunC7oukETXXdtLsZaT+8Jl5Oi3X9yfSdZdrSGcoG7N033JPnApMynHtQ/fHRl1ExCzSyJxvkD7sZ5GS4HK5r/wrpP33AimZ/qZi2ztJx/W5pH33Z5Y+uysaQ6F68vWKK4F7gemkkVMdliP9tuQZUrfrx0nHAaTj4wHgWUnzc9kJpGPn77kL8I/AJj2NfaDpGHFiZj0g6QnSCJU/tjoWs77wmYCZWYk5CZiZlZi7g8zMSsxnAmZmJTbgboI0ePDgGDVqVKvDMDMbUKZPnz4/ItqqywdcEhg1ahTTpk1rdRhmZgOKpCc7K3d3kJlZiTkJmJmVmJOAmVmJOQmYmZWYk4CZWYk5CZiZlZiTgJlZiTkJmJmVmJOAmVmJDbhfDFtzPXvevq0Ood947+FX1l7JbIDxmYCZWYk1NQnk55X+U9INeX5DSXdIelTSlR3P+zQzs+Zo9pnAscCMivnvA+dGxPtIzys9rMnxmJmVWtOSgKRhwG6kB5ojScAOwNV5lUnAns2Kx8zMmnsm8GPg68CSPL8usCAiFuX52cDQzjaUNF7SNEnT2tvbGx+pmVlJNCUJSNodmBcR0yuLO1m102ddRsSEiBgbEWPb2t71TAQzM+ulZg0R3Qb4jKRdgUHAmqQzg7UlrZDPBoYBzzQpHjMzo0lnAhFxUkQMi4hRwH7ALRFxAHArsFde7WDg+mbEY2ZmSat/J3ACcJykx0jXCM5vcTxmZqXS9F8MR8QUYEp+PRMY1+wYzMwsafWZgJmZtZCTgJlZiTkJmJmVmJOAmVmJOQmYmZWYk4CZWYk5CZiZlZiTgJlZiTkJmJmVmJOAmVmJOQmYmZWYk4CZWYk5CZiZlZiTgJlZiTX9VtKNduBPbmx1CP3GJcfu1uoQzBpm+uHjWx1Cv/Fv503o9bY+EzAzK7FmPWh+kKQ7Jd0j6QFJ38nlF0r6l6S78zSmGfGYmVnSq+4gSasAiyPizYKbvAHsEBELJa0I3Cbp5rzsvyPi6t7EYWZmfVPoTEDS2ZLG5de7Ac8DCyTtUWT7SBbm2RXzFL2I18zM6qhod9ABwP359beBA4HPAGcUbUjS8pLuBuYBkyPijrzodEn3SjpX0spdbDte0jRJ09rb24s2aWZmNRRNAqtGxKuS1gU2iohrIuKPwMiiDUXE4ogYAwwDxknaHDgJ+ACwFbAOcEIX206IiLERMbatra1ok2ZmVkPRJPCIpAOAo4HJAJIGA6/1tMGIWABMAXaOiDm5q+gN4AJgXE/rMzOz3iuaBI4CvgzsAJycyz4N/KHIxpLaJK2dX68C7Ag8JGlILhOwJ+90OZmZWRMUGh0UEf8A/m9V2aXApQXbGQJMkrQ8KfFcFRE3SLpFUhsg4G7gyMKRm5lZnxUeIippJ2A/YL2I2EPSWGDNiLil1rYRcS/w4U7Kd+hJsGZmVl9Fh4geA/wceBTYLhe/BpzWoLjMzKwJil4T+CqwY0ScBSzJZQ8BmzQkKjMza4qiSWANYFZ+3fEjrxWBor8YNjOzfqhoEpgKnFhV9hXg1vqGY2ZmzVT0wvAxwG8lHQ6sIelh4CWg0G0jzMysfyo6RHSOpK1Iv+wdSeoaujMilnS/pZmZ9WeFkkC+xfNzEXEncGcuGy5pnYi4p5EBmplZ4xS9JnAJ6UJwpZWAi+sbjpmZNVPRJDAiImZWFkTE48CoukdkZmZNUzQJzJa0ZWVBnn+m/iGZmVmzFB0ddC5wvaQfAI8DGwP/BZzeqMDMzKzxio4OOk/SAuAwYDhpdNDxfiykmdnAVvgGchHxK+BXDYzFzMyarCd3Ef0UMAZYvbI8Ir5d76DMzKw5iv5O4KfAPqTbRLxascgPizczG8CKngnsD4yJiFk11zQzswGj6BDR54AFvW1E0iBJd0q6R9IDkr6TyzeUdIekRyVdKWml3rZhZmY9VzQJ/Ai4VNJHJW1UORXc/g1gh4jYgnRdYWdJWwPfB86NiPcBL5BGH5mZWZMU7Q76ef5396ryAJavtXFEBLAwz66YpyA9uP7zuXwScGpFW2Zm1mCFzgQiYrkuppoJoIOk5SXdDcwDJpN+dLYgIhblVWYDQ7vYdrykaZKmtbe3F23SzMxqKNodBLx959Cte9NQRCyOiDHAMGAcsGlnq3Wx7YSIGBsRY9va2nrTvJmZdaLog+ZHSLqd9FzhP+ayvST9sqcNRsQCYAqwNbC2pI4uqWH4XkRmZk1V9Ezg/wM3kp41/FYumwzsVGRjSW2S1s6vVwF2BGaQfnewV17tYOD6gvGYmVkdFL0wPA7YLSKWSAqAiHhR0loFtx8CTJK0PCnxXBURN0h6ELhC0mnAP4Hzexi/mZn1QdEkMBcYDTzSUSBpM+CpIhtHxL3Ahzspn0lKMGZm1gJFu4POBm6QdAiwgqT9gStJ4/zNzGyAKnor6YmSngfGk24j/QXg5Ij4dSODMzOzxqqZBHI//inA6f7QNzNbttTsDoqIxcCXeWdUkJmZLSOKXhOYBBzZyEDMzKz5ejJE9BhJXyddE3j7l70RsV0jAjMzs8YrmgTOy5OZmS1Dil4Y3ph0YfiNxodkZmbN4gvDZmYl5gvDZmYl5gvDZmYl5gvDZmYlVvS2EZMaHYiZmTVfoSQg6dCulkXExPqFY2ZmzVS0O+igqvn3koaN3g44CZiZDVBFu4M+UV2Wzw46e06wmZkNED160HyVC4HDiqyYH1B/q6QZkh6QdGwuP1XS05LuztOufYjHzMx6qOg1gepksSpwILCgYDuLgOMj4i5JawDTJU3Oy86NiLML1mNmZnVU9JrAIip+G5A9TXrITE0RMQeYk1+/LGkGMLRokGZm1hhFk8CGVfOvRMT83jQoaRTpecN3ANsAR0v6AjCNdLbwQm/qNTOznit6TWAR8FJEPJmn+ZLeI2mDnjQmaXXgGuCrEfES8HPSKKMxpDOFH3Wx3XhJ0yRNa29v70mTZmbWjaJJ4NfAsKqyYcB1RRuStCIpAVwaEdcCRMTciFgcEUtIv0ge19m2ETEhIsZGxNi2traiTZqZWQ1Fk8AmEXFfZUGe/0CRjSUJOB+YERHnVJQPqVjt34H7C8ZjZmZ1UPSawDxJoyPisY4CSaOB5wpuvw3pB2f3Sbo7l30D2F/SGNJF5yeAIwrWZ2ZmdVA0CUwErpH0TWAmqR//e8Avi2wcEbcB6mTRTQXbNzOzBiiaBM4iPVTmbGA48BSpe+ec7jYyM7P+rehtI5YAP8yTmZktIwpdGJZ0oqStqsrG5YfMmJnZAFV0dNCxwINVZQ8CX61vOGZm1kxFk8BKvPtB828Cg+objpmZNVPRJDAdOKqq7EjgrvqGY2ZmzVR0dNDXgMmSDgIeB0YD6wM7NSowMzNrvKKjgx6Q9H5gd9IQ0WuBGyJiYSODMzOzxip6JgAwBHgSmB4RjzYoHjMza6Ka1wQk/YekJ4CHSc8UfkjSE5L2anRwZmbWWN2eCUjaDbgAOB24inS75yHAvsAvJb0eETc0PEqzZcShVx7a6hD6jYn7Tmx1CEbt7qCTgSMi4oqKsieA70t6Ki93EjAzG6BqdQd9kK6fGXAtsFl9wzEzs2aqlQTeANbsYtnapB+MmZnZAFUrCfwOOLOLZWcAv69vOGZm1ky1rgmcANwm6V7SoyE7Lgx/jnSGsG1jwzMzs0bqNglExNOStgSOA3YGBgPzgeuBcyPi+caHaGZmjVLzx2IR8QJpFNDJvW1E0nDgIuC9wBJgQkT8RNI6wJXAKNKoo31ye2Zm1gRFbyDXV4uA4yNiU2Br4MuSNgNOBP4UEe8D/pTnzcysSZqSBCJiTkTclV+/DMwAhgKfBSbl1SYBezYjHjMzS5p1JvA2SaOADwN3AOtHxBxIiQJYr4ttxkuaJmlae3t7s0I1M1vmdZkEJP294vUp9WhM0uqkUUZfjYiXim4XERMiYmxEjG1ra6tHKGZmRvdnAu+X1PHksOP72pCkFUkJ4NKIuDYXz5U0JC8fAszraztmZlZcd6ODrgceyXcQXUXS1M5WiojtajUiScD5wIyIOKdi0W+Ag4Gz8r/XF4zbzMzqoMskEBGHSNqWNHxzK9KHeG9tAxwE3Cfp7lz2DdKH/1WSDgOeAvbuQxtmZtZDtX4sdhvpF8MrRcSk7tYtUI+6WPzJ3tZrZmZ9U/TxkhMlfYL0bX4o8DRwSUTc0sjgzMyssQoNEZX0JdIve58l3UJ6DnCZpMMbGJuZmTVY0WcMfx3YKSLu6SiQdCVptM95jQjMzMwar+iPxdYFHqwqexhYp77hmJlZMxVNArcB50haFUDSasAPgb82KjAzM2u8okngSOBDwIuS5gILgC2AIxoVmJmZNV7R0UFzgI9LGgZsADwTEbMbGpmZmTVc0QvDAOQPfn/4m5ktI5p+F1EzM+s/nATMzEqsZhKQtJykHSSt1IyAzMyseWomgYhYAlwfEW82IR4zM2uiot1BUyVt3dBIzMys6YqODnoSuFnS9cAsIDoWRMS3GxGYmZk1XtEksArw6/x6WINiMTOzJiv6Y7FDGh2ImZk1X+EhopI2lXSypJ/m+U0kfajgthMlzZN0f0XZqZKelnR3nnbtefhmZtYXRZ8nsDcwlfRAmS/k4jWAc7rcaGkXAjt3Un5uRIzJ000F6zIzszopeibwXdLzBI4EFueye0g3kaspIqYCz/c8PDMza6SiSWA90oc+vDMyKCpe99bRku7N3UXv6WolSeMlTZM0rb29vY9NmplZh6JJYDrp+cKV9gPu7EPbPwc2BsaQHlf5o65WjIgJETE2Isa2tbX1oUkzM6tUdIjoV4A/SDoMWE3S74H3A5/qbcMRMbfjtaTzgBt6W5eZmfVO0SGiD0n6ALA76cN6FnBDRCzsbcOShuTnFAD8O3B/d+ubmVn9FX6eQES8Kul24F+kh8oUTgCSLge2BwZLmg2cAmwvaQzpusIT+CllZmZNVygJSBoBXApsDbwAvEfSHcABEfFkre0jYv9Ois/vSaBmZlZ/RS8MTyJdHF47ItYD3gP8I5ebmdkAVbQ76N+AT0XEWwARsVDSCcBzDYvMzMwaruiZwN+BcVVlY4G/1TccMzNrpi7PBCR9t2L2ceAmSTeSRgYNB3YFLmtseGZm1kjddQcNr5q/Nv+7HvAGcB0wqBFBmZlZc3SZBHz7aDOzZV/h3wlIWhUYDaxeWR4Rf613UGZm1hxFfyfwBeCnwJvAaxWLAhjRgLjMzKwJip4J/AD4XERMbmQwZmbWXEWHiL4JTGlgHGZm1gJFk8DJwDmSBjcyGDMza66iSeAR4DPAXEmL87RE0uJaG5qZWf9V9JrAxcBFwJUsfWHYzMwGsKJJYF3g2xHR18dJmplZP1K0O+gC3v14STMzG+CKngmMIz0U/pvA3MoFEbFd3aMyM7OmKJoEzstTr0iaSHo05byI2DyXrUO6xjCK9GSxfSLihd62YWZmPVf0GcN9fXjMhaRfHF9UUXYi8KeIOEvSiXn+hD62Y2ZmPVD0thGHdrUsIibW2j4ipkoaVVX8WdJzhyE9oWwKTgJmZk1VtDuo+qLwe4GNgduBmkmgC+tHxByAiJgjab2uVpQ0HhgPMGKEb1VkZlYvRbuDPlFdls8ONq17RJ23PwGYADB27FgPUzUzq5OiQ0Q7cyFwWB+2nytpCED+d14f6jIzs14olAQkLVc1rU7qnlnQh7Z/AxycXx8MXN+HuszMrBeKXhNYRHp2QKWngcOLbCzpctJF4MGSZgOnAGcBV0k6DHgK2LtgLGZmVidFk8CGVfOvRMT8oo1ExP5dLPpk0TrMzKz+il4YfrLRgZiZWfN1mwQk3cq7u4EqRUT427yZ2QBV60zgki7KhwJfAVatbzhmZtZM3SaBiDi/cl7SusBJpAvCVwLfbVxoZmbWaEWHiK4p6XvAY8D6wJYRMT4iZjc0OjMza6huk4CkVSSdBMwk/Tp424g4KCIeb0p0ZmbWULWuCfwLWB74ATANWF/S+pUrRMQtDYrNzMwarFYSeJ00Oug/u1gewEZ1jcjMzJqm1oXhUU2Kw8zMWqAvN5AzM7MBzknAzKzEnATMzErMScDMrMScBMzMSsxJwMysxJwEzMxKrOhDZRpG0hPAy8BiYFFEjG1tRGZm5dHyJJB9oidPKjMzs/pwd5CZWYn1hyQQwB8kTZc0vtXBmJmVSX/oDtomIp6RtB4wWdJDETG1coWcHMYDjBgxohUxmpktk1p+JhARz+R/5wHXAeM6WWdCRIyNiLFtbW3NDtHMbJnV0iQgaTVJa3S8Bj4F3N/KmMzMyqTV3UHrA9dJ6ojlsoj4XWtDMjMrj5YmgYiYCWzRyhjMzMqs5dcEzMysdZwEzMxKzEnAzKzEnATMzErMScDMrMScBMzMSsxJwMysxJwEzMxKzEnAzKzEnATMzErMScDMrMScBMzMSsxJwMysxJwEzMxKzEnAzKzEnATMzErMScDMrMRangQk7SzpYUmPSTqx1fGYmZVJqx80vzzwM2AXYDNgf0mbtTImM7MyafWZwDjgsYiYGRFvAlcAn21xTGZmpaGIaF3j0l7AzhHxpTx/EPCRiDi6ar3xwPg8uwnwcFMD7Z3BwPxWB7EM8f6sH+/L+hoo+3NkRLRVF67QikgqqJOyd2WliJgATGh8OPUjaVpEjG11HMsK78/68b6sr4G+P1vdHTQbGF4xPwx4pkWxmJmVTquTwD+A90naUNJKwH7Ab1ock5lZabS0OygiFkk6Gvg9sDwwMSIeaGVMdTSguq8GAO/P+vG+rK8BvT9bemHYzMxaq9XdQWZm1kJOAmZmJeYkUGe+DUb9SJooaZ6k+1sdy7JA0nBJt0qaIekBSce2OqaBTNIgSXdKuifvz++0Oqbe8DWBOsq3wXgE2Ik0/PUfwP4R8WBLAxugJG0HLAQuiojNWx3PQCdpCDAkIu6StAYwHdjTx2fvSBKwWkQslLQicBtwbET8vcWh9YjPBOrLt8Goo4iYCjzf6jiWFRExJyLuyq9fBmYAQ1sb1cAVycI8u2KeBty3aieB+hoKzKqYn43/yKwfkjQK+DBwR2sjGdgkLS/pbmAeMDkiBtz+dBKor0K3wTBrJUmrA9cAX42Il1odz0AWEYsjYgzpbgfjJA24bksngfrybTCsX8t919cAl0bEta2OZ1kREQuAKcDOLQ6lx5wE6su3wbB+K1/IPB+YERHntDqegU5Sm6S18+tVgB2Bh1obVc85CdRRRCwCOm6DMQO4ahm6DUbTSboc+BuwiaTZkg5rdUwD3DbAQcAOku7O066tDmoAGwLcKule0hfAyRFxQ4tj6jEPETUzKzGfCZiZlZiTgJlZiTkJmJmVmJOAmVmJOQmYmZWYk4CZWYk5CVi/J+lUSZe0Og6zZZGTgPUbkj4vaZqkhZLmSLpZ0rZ1qnuUpJDUkOdq57pfybE/LemcfGvxjuVdvjdJm0v6vaT5kgr/cEfSE5Jey3XOlXRBvi9Qx/JPS5oq6WVJ7ZL+LOkzedlukm6TtEDSs5LOy7eXtpJxErB+QdJxwI+BM4D1gRHA/9BPbsVdMHlsERGrA58EPg8cnret9d7eAq4CevOL6D1ym1sCWwHfym3uBfwKuIh0D6v1gW8De+Tt1gJOAzYANs3r/LAX7dtAFxGePLV0In0gLQT27mL5qcAl+fX2wOyq5U8AO+bX44BpwEvAXOCcXP4U6Y6uC/P00Vx+KOkWHy+QbvcxsqLeAL4MPAr8q8Z7CGB0xfyvgJ/Wem9VdYxOf5KF99vb7zvP/xC4gXQ326eA/+5BXf8B3NfqY8FT8yefCVh/8FFgEHBdHer6CfCTiFgT2Jj0DRtgu/zv2hGxekT8TdKewDdIH4BtwF+Ay6vq2xP4CLBZ0QAkbQZ8DPgn9X1v3bU5HNg1t7kJ6W62V/egiu0A3+eqhJwErD9YF5gf6QZ8ffUWMFrS4IhYGN0/6u8I4MyImJHbPgMYI2lkxTpnRsTzEfFagbbvkvQC8Fvgl8AF1Pe9debXkhaQHm34Z9J7WDcvm1OkAkk7AQeTuousZJwErD94Dhhcp4u2hwHvBx6S9A9Ju3ez7kjgJ/ni6ALSoyzF0k+Dm9Xplp3bMiLeExEbR8S3ImIJ9X1vndkzItaOiJERcVROVs/lZUNqbSxpa+AyYK+IeKRBMVo/5iRg/cHfgNdJXS+1vAKs2nvbc6MAAAGCSURBVDGTR+C0dcxHxKMRsT+wHvB94GpJq9H5E95mAUfkD9GOaZWI+GvFOn29zW5P3lu9PEx6b5/rbiVJHyY97+LQiPhTMwKz/sdJwFouIl4kdUX8TNKeklaVtKKkXST9oGr1R4BBeYjjiqTRMCt3LJR0oKS2/C18QS5eDLQDS4CNKur6BXCSpA/mbdeStHez35uSQcBKeX6QpJW7qbZWmwEcB5ws6RBJa0paTtK2kibkNjYHfgccExG/7ePbtAHMScD6hUhPujqO9KHeTvomezTw66r1XgSOIvW5P006M5hdscrOwAOSFpIuEu8XEa9HxKvA6cDtuftn64i4jnS2cIWkl4D7gV1a8N5GAq/xzoXZ10jf5vvS5tXAvqTRT8+QRkqdBlyfVzmedAZ1fv6dwUJJvjBcQn6ojJlZiflMwMysxBo1YsFsmSLpY8DNnS2L9IvdRrQ5Aniwi8WbRcRTjWjXysXdQWZmJebuIDOzEnMSMDMrMScBM7MScxIwMyux/wVV7W0HdLsawQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total Selected train:  94\n",
      "Total Selected test:  41\n",
      "Fitting 5 folds for each of 768 candidates, totalling 3840 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=40)]: Using backend LokyBackend with 40 concurrent workers.\n",
      "[Parallel(n_jobs=40)]: Done  82 tasks      | elapsed:    4.5s\n",
      "[Parallel(n_jobs=40)]: Done 285 tasks      | elapsed:   13.5s\n",
      "[Parallel(n_jobs=40)]: Done 568 tasks      | elapsed:   33.7s\n",
      "[Parallel(n_jobs=40)]: Done 933 tasks      | elapsed:  1.1min\n",
      "[Parallel(n_jobs=40)]: Done 1378 tasks      | elapsed:  1.9min\n",
      "[Parallel(n_jobs=40)]: Done 1905 tasks      | elapsed:  2.7min\n",
      "[Parallel(n_jobs=40)]: Done 2512 tasks      | elapsed:  3.8min\n",
      "[Parallel(n_jobs=40)]: Done 3201 tasks      | elapsed:  5.1min\n",
      "[Parallel(n_jobs=40)]: Done 3840 out of 3840 | elapsed:  6.1min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuned Mode:  RandomForestClassifier(bootstrap=False, ccp_alpha=0.0, class_weight='balanced',\n",
      "                       criterion='gini', max_depth=15, max_features='auto',\n",
      "                       max_leaf_nodes=None, max_samples=None,\n",
      "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                       min_samples_leaf=15, min_samples_split=6,\n",
      "                       min_weight_fraction_leaf=0.0, n_estimators=300,\n",
      "                       n_jobs=-1, oob_score=False, random_state=0, verbose=0,\n",
      "                       warm_start=False)\n",
      "In this fold\n",
      "conf_matrix: \n",
      "  [[ 1  6]\n",
      " [ 7 27]]\n",
      "classification_report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.12      0.14      0.13         7\n",
      "         1.0       0.82      0.79      0.81        34\n",
      "\n",
      "    accuracy                           0.68        41\n",
      "   macro avg       0.47      0.47      0.47        41\n",
      "weighted avg       0.70      0.68      0.69        41\n",
      "\n",
      "Total Selected train:  99\n",
      "Total Selected test:  36\n",
      "Fitting 5 folds for each of 768 candidates, totalling 3840 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=40)]: Using backend LokyBackend with 40 concurrent workers.\n",
      "[Parallel(n_jobs=40)]: Done  82 tasks      | elapsed:    7.0s\n",
      "[Parallel(n_jobs=40)]: Done 285 tasks      | elapsed:   20.3s\n",
      "[Parallel(n_jobs=40)]: Done 568 tasks      | elapsed:   48.0s\n",
      "[Parallel(n_jobs=40)]: Done 933 tasks      | elapsed:  1.4min\n",
      "[Parallel(n_jobs=40)]: Done 1378 tasks      | elapsed:  2.2min\n",
      "[Parallel(n_jobs=40)]: Done 1905 tasks      | elapsed:  3.0min\n",
      "[Parallel(n_jobs=40)]: Done 2512 tasks      | elapsed:  4.1min\n",
      "[Parallel(n_jobs=40)]: Done 3201 tasks      | elapsed:  5.3min\n",
      "[Parallel(n_jobs=40)]: Done 3840 out of 3840 | elapsed:  6.1min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuned Mode:  RandomForestClassifier(bootstrap=False, ccp_alpha=0.0, class_weight=None,\n",
      "                       criterion='gini', max_depth=15, max_features='auto',\n",
      "                       max_leaf_nodes=None, max_samples=None,\n",
      "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                       min_samples_leaf=3, min_samples_split=6,\n",
      "                       min_weight_fraction_leaf=0.0, n_estimators=100,\n",
      "                       n_jobs=-1, oob_score=False, random_state=0, verbose=0,\n",
      "                       warm_start=False)\n",
      "In this fold\n",
      "conf_matrix: \n",
      "  [[ 4 21]\n",
      " [ 0 11]]\n",
      "classification_report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "         0.0       1.00      0.16      0.28        25\n",
      "         1.0       0.34      1.00      0.51        11\n",
      "\n",
      "    accuracy                           0.42        36\n",
      "   macro avg       0.67      0.58      0.39        36\n",
      "weighted avg       0.80      0.42      0.35        36\n",
      "\n",
      "Total Selected train:  99\n",
      "Total Selected test:  36\n",
      "Fitting 5 folds for each of 768 candidates, totalling 3840 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=40)]: Using backend LokyBackend with 40 concurrent workers.\n",
      "[Parallel(n_jobs=40)]: Done  82 tasks      | elapsed:    9.9s\n",
      "[Parallel(n_jobs=40)]: Done 285 tasks      | elapsed:   27.4s\n",
      "[Parallel(n_jobs=40)]: Done 568 tasks      | elapsed:   55.8s\n",
      "[Parallel(n_jobs=40)]: Done 933 tasks      | elapsed:  1.5min\n",
      "[Parallel(n_jobs=40)]: Done 1378 tasks      | elapsed:  2.3min\n",
      "[Parallel(n_jobs=40)]: Done 1905 tasks      | elapsed:  3.2min\n",
      "[Parallel(n_jobs=40)]: Done 2512 tasks      | elapsed:  4.3min\n",
      "[Parallel(n_jobs=40)]: Done 3201 tasks      | elapsed:  5.2min\n",
      "[Parallel(n_jobs=40)]: Done 3840 out of 3840 | elapsed:  6.2min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuned Mode:  RandomForestClassifier(bootstrap=False, ccp_alpha=0.0, class_weight='balanced',\n",
      "                       criterion='entropy', max_depth=15, max_features='auto',\n",
      "                       max_leaf_nodes=None, max_samples=None,\n",
      "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                       min_samples_leaf=3, min_samples_split=9,\n",
      "                       min_weight_fraction_leaf=0.0, n_estimators=700,\n",
      "                       n_jobs=-1, oob_score=False, random_state=0, verbose=0,\n",
      "                       warm_start=False)\n",
      "In this fold\n",
      "conf_matrix: \n",
      "  [[ 5 10]\n",
      " [ 7 14]]\n",
      "classification_report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.42      0.33      0.37        15\n",
      "         1.0       0.58      0.67      0.62        21\n",
      "\n",
      "    accuracy                           0.53        36\n",
      "   macro avg       0.50      0.50      0.50        36\n",
      "weighted avg       0.51      0.53      0.52        36\n",
      "\n",
      "Total Selected train:  113\n",
      "Total Selected test:  22\n",
      "Fitting 5 folds for each of 768 candidates, totalling 3840 fits\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=40)]: Using backend LokyBackend with 40 concurrent workers.\n",
      "[Parallel(n_jobs=40)]: Done  82 tasks      | elapsed:    9.6s\n",
      "[Parallel(n_jobs=40)]: Done 285 tasks      | elapsed:   27.6s\n",
      "[Parallel(n_jobs=40)]: Done 568 tasks      | elapsed:   56.3s\n",
      "[Parallel(n_jobs=40)]: Done 933 tasks      | elapsed:  1.6min\n",
      "[Parallel(n_jobs=40)]: Done 1378 tasks      | elapsed:  2.3min\n",
      "[Parallel(n_jobs=40)]: Done 1905 tasks      | elapsed:  3.2min\n",
      "[Parallel(n_jobs=40)]: Done 2512 tasks      | elapsed:  4.2min\n",
      "[Parallel(n_jobs=40)]: Done 3201 tasks      | elapsed:  4.9min\n",
      "[Parallel(n_jobs=40)]: Done 3840 out of 3840 | elapsed:  5.6min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tuned Mode:  RandomForestClassifier(bootstrap=False, ccp_alpha=0.0, class_weight=None,\n",
      "                       criterion='entropy', max_depth=15, max_features='auto',\n",
      "                       max_leaf_nodes=None, max_samples=None,\n",
      "                       min_impurity_decrease=0.0, min_impurity_split=None,\n",
      "                       min_samples_leaf=6, min_samples_split=6,\n",
      "                       min_weight_fraction_leaf=0.0, n_estimators=300,\n",
      "                       n_jobs=-1, oob_score=False, random_state=0, verbose=0,\n",
      "                       warm_start=False)\n",
      "In this fold\n",
      "conf_matrix: \n",
      "  [[ 2  0]\n",
      " [ 4 16]]\n",
      "classification_report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.33      1.00      0.50         2\n",
      "         1.0       1.00      0.80      0.89        20\n",
      "\n",
      "    accuracy                           0.82        22\n",
      "   macro avg       0.67      0.90      0.69        22\n",
      "weighted avg       0.94      0.82      0.85        22\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean of Each fold\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUC-ROC all:  [0.5630252100840337, 0.5290909090909091, 0.5904761904761905, 0.8625]\n",
      "AUC-ROC = 0.64 ± 0.13)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Matthews_corrcoef all:  [-0.05984127209279335, 0.23452078799117146, 0.0, 0.5163977794943222]\n",
      "Matthews_corrcoef = 0.173 ± 0.227)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "balanced_accuracy all:  [0.4684873949579832, 0.58, 0.5, 0.9]\n",
      "balanced_accuracy = 0.61 ± 0.17)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sensitivity all:  [0.14285714285714285, 0.16, 0.3333333333333333, 1.0]\n",
      "Sensitivity = 0.41 ± 0.35)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Specificity all:  [0.7941176470588235, 1.0, 0.6666666666666666, 0.8]\n",
      "Specificity = 0.82 ± 0.12)\n",
      "One Combined Confusion Matrix\n",
      "conf_matrix: \n",
      "  [[12 37]\n",
      " [18 68]]\n",
      "classification_report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "         0.0       0.40      0.24      0.30        49\n",
      "         1.0       0.65      0.79      0.71        86\n",
      "\n",
      "    accuracy                           0.59       135\n",
      "   macro avg       0.52      0.52      0.51       135\n",
      "weighted avg       0.56      0.59      0.56       135\n",
      "\n",
      "matthews_corrcoef\n",
      "0.04117074412173107\n",
      "balanced_accuracy_score\n",
      "0.517797816801139\n",
      "Sensitivity\n",
      "0.24489795918367346\n",
      "Specificity\n",
      "0.7906976744186046\n",
      "AUC-ROC for the combined results: \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYQAAAEJCAYAAACUk1DVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deVxU9f7H8dew74vI4l6Kggq4prgkae6KC7hUdrVcuqZGWVpetcjMJetqZWZit8yupmak4J563TWTLMVQ86fmzqoo6zAz5/cHOjUhDiAzw/J5Ph49HpzzPcv7CzafOdv3qBRFURBCCFHtWVk6gBBCiIpBCoIQQghACoIQQoi7pCAIIYQApCAIIYS4y8bSAcpCp9ORnZ2Nra0tKpXK0nGEEKJSUBSFgoICnJ2dsbIqejxQKQtCdnY2Z8+etXQMIYSolJo0aYKrq2uR+ZWyINja2gKFnbKzsyv1+omJiQQFBZV3rApN+lw9SJ+rh7L2Wa1Wc/bsWf1n6N9VyoJw7zSRnZ0d9vb2ZdpGWderzKTP1YP0uXp4mD4Xd6pdLioLIYQApCAIIYS4q1KeMnoQnU7HlStXyM7OLnYZGxsbkpKSzJjK8qTPFZOtrS0+Pj64ublZOooQpi8IWVlZPPXUU3z22WfUrVvXoC0pKYkZM2aQnZ1N27ZtmTVrFjY2DxcpLS0NlUpFQEDAfW+rgsK7lJydnR9qP5WN9LniURSF3Nxcrl69CiBFQVicSU8Z/frrrzz99NNcvHjxvu1Tp07lrbfeYvv27SiKwrp16x56n7du3cLX17fYYiBERaFSqXBycqJOnTqkpKRYOo4Qpi0I69atIzo6Gh8fnyJtV69eJS8vj5YtWwIQERHBtm3bHnqfWq222FuqhKiIHB0dKSgosHQMIUx7ymjOnDnFtqWkpODt7a2f9vb2Jjk5uVz2K08vi8pE/r2K0jh5Lo30OxqTbNtiF5V1Op3B/wiKopT6f4zExMQi82xsbB54QfmekixTHrKysli8eDE///wz1tbWuLm5MXnyZJo2bfrQ2+7Xrx/Lly+ndu3aBvOXLl1Ks2bNCAsLM5hf2j4fO3aMZcuWsXz5coP50dHRbNq0ia1btxoc/b366qucOXOGzZs3l3gf48aN45///Cdt27a9b/u1a9cYN25csdtcuHAhmzZtYtu2bfqHFP+6zl/73Lp1a37++WcAUlNT+fDDDzlz5gzW1tb4+vry+uuvU7duXY4dO8bLL79MvXr1gMJ/qzk5OTz33HMMGTIEgPT0dBYuXMjJkydxcHCgZs2aREVFERgYqN/fpk2bWLNmDRqNBp1Ox+DBg3n66afv2w+1Wk1CQkKJf28PUl7bqUyqU59nr7lCaKArXq7l32eLFQQ/Pz9SU1P102lpafc9tfQgQUFBRR7OSEpKMnoh0VwXG3U6HWPHjqV9+/bExcVhY2PDkSNHiIqKYvPmzXh6ej7U9q2srHB0dCzSlylTphRZtix9dnBwwNraush6NjY2+Pr6cuDAAf7xj38AhYXvzJkzWFlZlWo/1tbWODg4FLuOo6NjsdvUaDTs3LmT1q1bc+DAAcLDww3WAYqs5+zsTE5ODv/85z8ZPXo0ixYtQqVSERcXx8SJE9m6dSsODg4EBwfz9ddf69dLSkpiyJAhREZGYmNjw/jx44mMjNSvf/DgQSZOnMjq1atp2LAha9euZc2aNSxbtgwfHx9u377N6NGjcXd3Z+jQoUX6YmdnR4sWLUr8eytOQkICbdq0eejtVCaVuc/X0rL4Mv4UGm3hiysnDW2Bl7sjB09cY+fRS0WWf+2Z1kz9Ry1upfxRpj7n5+ff94v0PRYrCHXq1MHe3l7/x9y4cSNdunSxVByT+PHHH7l+/TpRUVH6D6jQ0FDmzZuHTqcD4LPPPiMuLg5ra2s6derE1KlTuX79OhMnTqRhw4acO3eOZs2a0apVK77//nsyMzNZsmQJjRo1AuCTTz7h9OnT2NvbM2vWLAIDA5k2bRrt2rWjXbt2TJo0icaNG3Pq1Cm8vb356KOP8PDwYN++fXz88cdoNBrq1q3L7Nmz8fT05MCBA8ybNw97e3seffTRYvvWs2dPtm/fri8IO3fu5IknnmDfvn0A5ObmMnPmTM6cOYNKpWLMmDEMGjQItVrNjBkzSExMpE6dOty8eVO/zZiYGLZu3YpWq6Vz585MnTr1gb/fPXv2UL9+fQYNGsTKlSv1BcGYzZs3U6NGDYYPH66fN2DAAOzs7FCr1fdd5+rVqzg6OmJnZ8emTZvw8vJizJgx+vZOnToRERHB559/zty5c1m6dClz587Vf8lxc3PjvffeIysrq0QZRfVw8lw6RxJv8EgtN2xsrLj7sUC+WsutrPwiy+sU6BRSm4SE6ybJY/aCMG7cOKKioggODuaDDz5g5syZZGVl0bx5c0aOHFnu+/vXpweKzGvXtCaDuwaSp9Yw6/MjRdqfbFuf7u3qk5mVz/yVPxVp79vhUR5vVcfovn/77TcCAwOL3PF071TO3r172b17N9999x22tra89NJLrFmzhrCwMM6cOcO8efMIDAykV69e+Pj4sHbtWj755BPWrl3L9OnTAWjQoAHz589n7969TJs2jQ0bNhjs6/Tp08ydO5cGDRowbdo04uPj6devH//+979ZuXIl7u7urFmzhg8++IDo6GimTZvGV199RaNGjZgxY0axfWvatCkHDhwgLS2NmjVrsnXrVkaPHq0vCIsXL8bT05NNmzaRkZHB0KFDCQwM5ODBgwBs3bqVixcvMmDAAAD27dtHYmIi69evR6VSMXXqVOLi4h74LSg2NpbevXsTFhbGv/71L86dO4e/v7/Rv0tSUhLNmzcvMr937976nxMTExk4cCC5ublkZmbSvn17vvjiC+zs7Dh58iTBwcFF1n/sscdYuHAhGRkZXL9+nWbNmhm03yviovpJvZnLsaQb3HuBfeN6HjSu50meuvBaQPTYUGp6OOqX79a2Ht3a1jN7TrMUhN27d+t//uv56MDAQNavX2+OCBZhZWX1wPFGjhw5Qr9+/XB0LPyHEBkZyYYNGwgLC6NmzZr6DxQ/Pz86dOgAQO3atbly5Yp+G/dOP4SFhTF16lRu375tsA8vLy+aNWtGdnY2jRs3JjMzk19//ZXr16/rC7BOp8Pd3Z0zZ87g4+Oj/+AaPHgwH330UbH5e/bsyY4dO+jXrx9ZWVnUqfNnkTxy5Ahz584FoEaNGjz55JMcPXqUo0eP6r+ZP/LII7Rq1QqAw4cPc+LECSIiIgDIy8ujdu3axRaE9PR0Dh48yLvvvouDgwNdu3ZlzZo1zJw58763HP/1GpWVlZXRQRGDgoL4+uuvUavVTJ06FRcXF0JCQoDCi8BarbbIOgUFBahUKv3+q+P4OuL+1u8+y5ZDF/XTI3oH0rieJy2beONgZ42TQ8V4RrhipDCheRM6F5l370Kjg53NfdvvcXexf2C7MUFBQaxevbrIBfOFCxfSsWNH/Wmjv9JoCr8x/P0Dy9ra+r77+Ot8RVGKPNj31w8llUqFoihotVpat27NZ599BhSeV8zOzubatWsoiqJfvrh93tOnTx/mzZuHnZ0dPXr0MGj763buTWu1Wn2Ge+7l1Wq1jBo1iueffx6A27dvY21tbXBK6a/i4uJQFEV/kTcvL4+CggKmTJmCm5sbd+7cMVg+PT0dd3d3oPDvEhsbW2SbM2bM4LnnnjOYZ2dnx7vvvkuvXr3YsmULffv2JSQkhG+++abI+sePHycoKAgPDw/q1atHYmIijz32mL796NGj7Nu3777XeETVptUpeLjY8/GUJ4DCzx6Auj6ufP12bxzsK8ZHsTy9ZUJt27bFy8uLTz75RP+Ncv/+/cTGxuLv709oaCibN28mLy8PjUbDd999R2hoaKn2ER8fD8APP/xAo0aNcHJyMrpOixYt+OWXX7hw4QIAn376KQsWLCAgIIC0tDROnz4NYPRuocDAQNLS0li3bp3B6RYovFZy7+gvIyODXbt20a5dOzp06EB8fDw6nY6rV6/q7/oJDQ1l48aNZGdno9FomDhxItu3by9237GxscyfP5/du3eze/duDhw4gLu7O1u2bMHFxYUGDRqwa9cu/fJr167VH2X17t2bq1ev8u233+rbv/vuO44ePUqDBg2K7MvV1ZWXXnqJBQsWkJeXR9++fcnNzWXZsmX64nbgwAFiY2P11xXGjBnD/Pnz9TdOZGRkMH/+/PtuX1R9LZt407fTo3i6OuDp6oDj3QJgbaWqMMUAqsERgiWpVCo+/fRT5s2bR//+/bGxscHT05OYmBhq1qxJ165dSUpKIjIyEo1GQ+fOnXn22We5ceNGifdx8eJFBg4ciLOzM/Pnzy/ROt7e3sydO5dXXnkFnU6Hr68v77//Pra2tixcuJCpU6diY2NT5Bz4/fTo0YOjR4/i5+dncCpr4sSJvP3224SHh6PVahk/fjzNmzenSZMm/P777/Tp04c6derQpEkTALp168bp06cZNmwYWq2Wxx9/nMGDB+uHdfirkydPcvPmTYOjEisrK0aNGsWaNWuIiIjg/fff58033+Tzzz+noKCAgIAA3nrrLaDw7qkVK1Ywd+5cVqxYgUqlom7duvprBPczdOhQvv76a7788ktefPFFvvrqKxYsWEDv3r1RqVTUrl2bL7/8Un+67emnn0aj0TB69Gj9UdHw4cPve4eRqPo6tzB+zbEiUCl/P7avBO7dOlXcbafG7vGv6GPcmIL0uWIryb/bkqjMt2CWVWXos7pAi1an6I8MHlZZ+/ygz06QU0ZCCGFyMRtOMn7+TkvHMEpOGQkhRBnkF2hZt7Pou91bNvYm2L8m2bkFxO45B8Dvl26ZO16ZSEEQQogyKNDoWL/79yLz7WytCPavSU6exqC9hX9Nc8YrkypZEMoyLpIQlnK/249FxXI5+Q45eYUj0up0sO/4FR5r7sfG9wcUu463p+MD2yuiKlcQHBwcSE9Px8vLS4qCqNAURaGgoIDk5ORKc/G7uvo8LpGfTxu+s6KWtzOtA0o3/lpFV+UKQt26dbly5YrBwHl/p1arjT6pWtVInysmGxsb3N3dqVmz4p9OqG7OXb7Fwm8SiBrWihG9Agnv3FDfZm2lonlDLwumM40qVxBsbW0fOCgbFN6yVR4jS1Ym0mchSkZRFE5fvEnc/v/jcnIW+QVaWjT2Nr5iFVDlCoIQQpTVhWuZzP/qJ66lZWNvZ023tvVoXM/D0rHMRgqCEKJau5qaxaUbt+kQXBvfGk74eTkz9MkmdAyphZND9XodrxQEIUS1dvTUDdbtPEurJj44Odgy64UOlo5kMfKkshCi2svKLUBX+UbxKXdSEIQQQgBSEIQQQtwlBUEIUa1l3n13sZU8yCoXlYUQ1YdGq+PnMyns/ukyoUF+PNGmHoPC/PGp4VShXlRjKfIbEEJUC/uPXyVmw0luZeXj5mxHSOPCp8M9XO3p2/HBD7NWF1IQhBDVwqGT19DqdMx8vh1tmvpiYy1nzP9OCoIQolqwsbaipocj7YNqWTpKhSUFQQhRLbw2omK/ZrMikIIghKjQbt7JQ11Q+M4IF0dbnB1t0Wh1pGfmFbZnaUjOyAHA1ckWJwdbCjRaMm7nF9mWt4cjVlZyN1FxpCAIISqss5du8tpH+/TTo8ObM/gJf5Izchg/f9efC8bdAGDikBb07vAIf1y/w+QP9xpsy97OmkFhjXi2d1OzZK+MpCAIISqse88IDO/RBL8azvqRRz1c7Hl5eCsALv5xkUcaPAJA4COeQOHbyu61/1Wjuu5mSF15SUEQQlR47Zr50aS+p37a2dGW7u3qA5BgnUqbNvUNlnd3sde3i5KTgiCEqLCaN/Ti49eeoFZNecWoOUhBEEJUWE4OtjxaW07zmIs8mSGEqLCSM3LYdOA8N+/kWTpKtSBHCEIIiyvQ6MjKUePp5kCeWsP5q5kAnL6YwZebfqNJfU88XR0snLLqM2lBiI+PZ+nSpWg0GkaNGsWIESMM2k+dOsVbb71FQUEBtWrV4v3338fNzc2UkYQQFVD8/vOkZ+YyblAwqTdzeeOTAwbtjjLwnFmY7LecnJzMokWLiI2Nxc7Ojqeeeor27dvj7++vX2bOnDlERUURFhbG/Pnz+c9//sPkyZNNFUkIUUHl5BcQt/884wYF4+3hyDt/eY2ls6MtdX1cLJiu+jDZNYRDhw4RGhqKh4cHTk5O9OrVi23bthkso9PpyM7OBiA3NxcHBzkkFKK6uvc6Agd7G1oF+Oj/a1LfE5W8q8AsTHaEkJKSgre3t37ax8eHEydOGCwzbdo0Ro8ezdy5c3F0dGTdunWl2kdiYmKZ8yUkJJR53cpK+lw9VMY+X7+eiaKUPXtl7PPDMkWfTVYQdDqdQVVXFMVgOi8vjxkzZrBixQpCQkL48ssveeONN4iJiSnxPoKCgrC3ty91toSEBNq0qV4DXUmfq4eK3uejv93gwC9X9dMvDA7BxdGWb48UXjMoS/aK3mdTKGuf8/PzH/hF2mQFwc/Pj2PHjumnU1NT8fHx0U+fPXsWe3t7QkJCABg+fDgfffSRqeIIISqAzQcucOJcGl7uhaeHtdrCQevaN/fDyUEuHFuaya4hdOzYkcOHD5ORkUFubi47duygS5cu+vYGDRpw48YNzp8/D8CuXbsIDg42VRwhRAXRqI47n8/oweczeuDuUniEP/gJf94aE2rhZMJkJdnX15fJkyczcuRICgoKGDJkCCEhIYwbN46oqCiCg4OZN28er7zyCoqi4OXlxdy5c00VRwhRAbQO9CFfrbV0DFEMkx6jhYeHEx4ebjBv+fLl+p/DwsIICwszZQQhRAUysEsjS0cQDyBDVwghhACkIAghzCh6+WGmfLzP+ILCIqQgCCHMR7n7n6iQpCAIIYQApCAIIYS4q0QF4caNG+zduxetVsu1a9dMnUkIIYQFGC0Ie/bs4amnnmLWrFmkp6fTr18/du7caY5sQogqJjS4Fo+3qmPpGKIYRgvCkiVLWLduHW5ubvj4+LB69Wo+/vhjc2QTQlQxfTo8Is8iVGBGC4JWqzUYg6hp06YyFK0QokzUBVryC+RJ5YrK6JPKjo6OXLt2TV8Ejh07VqYRRoUQVUueWkN0zGFu3ckv8iXx49eewM7WmrU7z/C/Y1f086+mZhFQ35MPXu7y982JCsBoQXjttdcYPXo0qampDB8+nIsXL7J48WJzZBNCVGB2NtZMGdGWlVt/Q6s1fLjgXoHwcnOkYR13/fyGddxp19zPrDlFyRktCK1bt2bdunUcP34cnU5HixYtqFGjhjmyCSEqMCsrFd6ejrz2TPHj8ndvV5/u7eqbMZV4GEavIYwdOxY3NzfCwsLo2rUrNWrUYNiwYebIJoSowDKz8onZcJKzl25aOoooJ8UeIURFRXHhwgUuX75sMGKpRqPBzs7OLOGEEJan0ylcT8+mjnfhi+7P/JGBWqMj/VYu8fvP07ieB03qe1o4pSgPxRaE119/natXr/Lmm2/y5ptv6udbW1vj7+9vlnBCCMs7lpTMNz+cYfYLHXBxsmPBfxNIycjRtzvay5vOqopi/5J169albt26bNu2DSsrwzNLOTk5xawlhKhqcvI1nLt8i8xsNS5Odkwd0Qa1pvDWUTsbaxrL0UGVYbS07969m48//picnBwURUGn03Hr1i2OHz9ujnxCiAom8BG5qaSqMloQFixYwCuvvMI333zDuHHj2LlzJ87OzubIJoQQwoxK9GBa3759SUpKwt7enrfffpt+/frxxhtvmCOfEKKc7Tt+hR9P3aBb23q0CfQl9WYuKzafKrJc7w6PENyoJmm3ci2QUliC0YJgb2+PWq2mfv36JCUl0b59exm6QohKLH7/ec5fu02bwMIhadQaLecu3yqy3O0QNQAtm3jzY6InXm4OZs0pzM9oQejWrRsvvPAC7733HsOHDychIQFPT7mIJERFd+aPDG6k5+DqbEfrgMIP/2NJyWRmq2n2aA26tS18YKyOtwvL/tW92O341/Xg/SgZaqI6MFoQxo8fz4ABA/D19WXJkiUcO3bM4LkEIUTFNOOzQ+SrtQQ08NQXhBWbTnE9LZvG9TwsnE5URA8sCBcuXMDZ2ZnatWsD0Lx5c2rWrMmcOXP497//bZaAQoiyKSjQ0qfDIwzp1lg/b+bo9mi0OnxrOFkwmaioih264vPPPyciIoJevXrx008/AbBixQr69u1Lamqq2QIKIcrOzdkOn798+Pt5OVPXxxVbG2sLphIVVbFHCGvXrmXLli1cv36dL774gm+++YajR4/y9ttvyykjISoBJwdbbG3ltemi5IotCI6OjtSqVYtatWoxYcIEWrZsyZYtW3BzczNnPiFEGX3zbl9LRxCVTLEFwdr6z0NKFxcXPvzwQxwc5LYzIYSoqkp0POnq6irFQIhK5r2VP7H/+FVLxxCVSLFHCOnp6Xz55ZdFfr7n+eefN20yIcRDOXTiGnW8XXicOpaOIiqJYgtCp06dOHv2bJGfhRBCVE3FFoR58+Y99Mbj4+NZunQpGo2GUaNGMWLECIP28+fPEx0dTWZmJt7e3ixcuBB3d/ditiaEEMKUTHZPWnJyMosWLWL16tVs2LCBtWvXcu7cOX27oii8+OKLjBs3jri4OJo2bUpMTIyp4gghhDDCZK86OnToEKGhoXh4FD4i36tXL7Zt28akSZMAOHXqFE5OTnTpUjhGyvjx47l9+7ap4ghRqZ29dJMP1/yMRqvwdM8Aurapx+XkO8z+4keD5fLz8hlvd40OwbXx8nDEyUHeZiZKzmT/WlJSUvD29tZP+/j4cOLECf30pUuXqFmzJtOnTycpKYmGDRsavKpTCPGnc1ducTk5iw7BtfBwsQfA3s6aJvUMB5rMyEjHzbmw/YuZPc2eU1RuJSoIJ06c4LfffiMiIoJTp07RqlUro+vodDqDYbIVRTGY1mg0HD16lP/+978EBwfz4YcfMn/+fObPn1/i8ImJiSVe9u8SEhLKvG5lJX2uvC5dygKgkz/osi6TkHAZgK5N/76kF3k3L5KQcNGs+SytqvydS8MUfTZaEGJjY/nPf/5Dfn4+PXr0YMKECUyePJlhw4Y9cD0/Pz+OHTumn05NTcXHx0c/7e3tTYMGDQgODgagf//+REVFlSp8UFAQ9vb2pVoHCn+Rbdq0KfV6lZn0uXJLzr8AP90ipEUInq7FPxNUlfpcUtLnksvPz3/gF2mjF5W//vpr1q5di4uLC15eXsTGxvLVV18Z3XHHjh05fPgwGRkZ5ObmsmPHDv31AoBWrVqRkZHB6dOngcJ3Nzdv3rwkfRKi2gnxr8nkp1vj5GBr6SiiCjN6hGBlZYWLi4t+ulatWgbDWhTH19eXyZMnM3LkSAoKChgyZAghISGMGzeOqKgogoODWbJkCTNnziQ3Nxc/Pz8WLFjwcL0Rooqq6+NKXR9XS8cQVZzRguDh4UFSUpL+/H9cXFyJnxUIDw8vMjLq8uXL9T+3aNGC9evXlyavENVSxu08rqdl06S+J7Y2MoKpMA2jBWH69Om8/PLLXLp0ic6dO2Nvb8+nn35qjmxCiLuOJF5n6XcnWPl2rwdeQxDiYRgtCA0bNmTjxo1cvHgRrVbLo48+iq2tnMcUQoiqxuixZ1hYGEuWLMHBwYEmTZpIMRDChE6cS2X+yp9Yt/PPscM+Xf8r24/8YcFUorowWhBWrFiBWq3mmWeeYcyYMWzbtg2NRmOObEJUO7uPXebIyeuk3MzRz7uamkWBRkvLxt64OtlZMJ2o6kp0ymjKlCm8+uqr7N+/nyVLlvDOO+9w6NAhc+QTotqp4e7ApKEt9dNzXuxkwTSiOinRk8rp6enExcXx/fff6welE0IIUbUYLQjjx4/n+PHj9OjRg9mzZ9OiRQtz5BKiSijQaMlXa4HCsYdsbazRanXk5hc97WpvZ/z5HiFMyWhB6NatG//+979xdnY2Rx4hqgyNVsfo2T9wKysfgBcjQ+jb8VEuXr/NK4v2Fln+1Wda82JkC7RanbmjCgE8oCBs3LiRgQMHkpWVxbp164q0yys0hXgwjUbHrax82jXzo0XjmjR9pAYAXu6OjBsYVGR5/7oe2Ntag60cKQjLKLYg/PFH4W1uv//+u9nCCFEVNW9YgwFdGumnPVztDaaFqCiKLQj3Rh598skn6d69u0Hbhg0bTJtKiCrA2tqKyK7+NK7vaXxhISqAYgvC7t270Wg0LFiwAEVRUBQFKHyPweLFixk0aJDZQgpRGdnaWPFcfxnBV1QexRaEpKQkjhw5Qnp6OitXrvxzBRsbnnvuOXNkE6JSUxSFrNwC7GytC68NCFHBFVsQJk6cyMSJE1m1ahUjRowwZyYhqoR8tZZn3tzK8/2bEdG1saXjCGGU0buM8vPz+fLLL4u0y11GQghRtchdRkIIIYAS3GU0b948/Ty1Wk1aWhq1a9c2fTIhhBBmZXS00x9++IHZs2eTlZVF7969GThwYIneqSyEEKJyMTp0xbJly5gzZw47duygZcuWvPPOO4wcOZJRo0aZI58QFV6eWsO6nWfJzTMcn2j0gCCe7RNIs0e9LJRMiNIxWhAURSEgIIDly5fTpUsXXFxc9M8kCCHg90u3+HZX4bU2V6c/XyD1wuBghncPsFQsIUrNaEGwsrJiy5Yt7N+/nzfeeIO9e/eiUqnMkU0Ii7iTo+b3y7f00/V8XPH2dCQnr4DTf9wssryPpyPvRz1OfV9XnBzkjYKi8jJaEN544w0++eQTXnvtNby9vVm6dCkzZ840RzYhLCJmw0n2JFzRT08Y0oI+HR7hWmo20TGHiyz/2og2PNG6rjkjCmESRgtC27ZtWbFiBVevXuWPP/5gzZo15sglhMXkq7X4eTnx6tNtAPDzcgKgjo8LCyY9XmT52t4yNLyoGowWhIsXLzJx4kRSUlLQ6XR4enqybNkyGjWS0RpF1VKg0WJrY03jeh54uTnQ9NEaBu2O9jZF5glRlRgtCLNnz2bs2LEMHjwYgO+++45Zs2YZjG8kRGWlLtDyY+INdh27xI30HJa+0Y2hTzaxdCwhLMJoQUhPT9cXA4DIyEhWrFhhykxCmMXanWf4/n/nyM7TUEEVWK0AABtBSURBVNPdga5t66HW6GQgOlFtGS0IWq2WW7du4eHhAUBGRobJQwlhDlsPXcSnhhNjwoMI8q+JtZXcPSeqN6MF4dlnn2X48OH06dMHlUrFli1b5KE0USVMGtoSD1d7/Ot6WDqKEBWC0YIwfPhwGjRowP79+9HpdERHR9OxY0dzZBPCpNo29bV0BCEqlAcWhL1793L+/Hkee+wxpk6daq5MQpjFyXNpuDnb0aCWm6WjCFEhFDu4XUxMDLNnz+bXX39l/PjxxMfHmzOXECb379UJbNz3f5aOIUSFUWxBiI+PZ8OGDXz44YesXLmSVatWlXrj8fHx9O3bl549ez5w/T179tCtW7dSb18IIUT5KfaUkY2NDS4uLgA0bNiQ7OzsUm04OTmZRYsWERsbi52dHU899RTt27fH39/fYLm0tDTee++9MkQXQghRnoy+D+EeGxuj158NHDp0iNDQUDw8PHBycqJXr15s27atyHIzZ85k0qRJpdq2EEKI8lfsp7xWqyUzM1M/1PXfp+89l1CclJQUvL299dM+Pj6cOHHCYJmVK1fSrFkzWrRoUabwiYmJZVoPICEhoczrVlbSZ0NqdQFp6WlV7vdS1fpTEtLn8lFsQTh79iyhoaEG7z5o3749ACqViqSkpAduWKfTGQyTrSiKwfTZs2fZsWMHK1as4MaNG2UKHxQUhL29fanXS0hIoE2bNmXaZ2UlfS5qukcars52NPCrOncZyd+5eihrn/Pz8x/4RbrYgnD69OlS7+yv/Pz8OHbsmH46NTUVHx8f/fS2bdtITU0lMjKSgoICUlJSeOaZZ1i9evVD7VeIkgpqVNPSEYSoUEp8DaG0OnbsyOHDh8nIyCA3N5cdO3bQpUsXfXtUVBTbt29n48aNxMTE4OPjI8VAmNVPv93g98tFX3gjRHVlsoLg6+vL5MmTGTlyJIMGDaJ///6EhIQwbtw4Tp48aardClFiS9b/ytZDFy0dQ4gKo3S3DpVSeHg44eHhBvOWL19eZLm6deuye/duU0YR1VSBRssbnxxgxvPt8HJ3ZMPe/2Pj3nMA3LydJ6+DFeIvSnSEkJeXx5kzZ1AUhdzcXFNnEqLc3M4ufD/ywRPXAKjl5USrAB9aBfjw5GP16dG+voUTClFxGD1C+OWXX5g0aRI2NjasWbOGgQMHsnTpUlq3bm2OfEKUCzubwncctA+qRfugWhZOI0TFZPQIYcGCBaxYsQIPDw/8/PxYsGABc+bMMUc2IYQQZmS0IOTl5RkMNxEWFoZWqzVpKCGEEOZntCDY2NiQmZmpv/h2/vx5k4cSoryoVCrsbK2xkrehCWGU0WsIL774Is8++yxpaWm8+uqrHDx4kHfeeccc2YR4aDXcHPhufn9LxxCiUjBaELp27UrDhg05ePAgOp2OiRMn0qhRI3NkE0IIYUZGC8KtW7dwd3enb9++BvOMDW4nREWQlaNm2fcn6d6uPi0aextfQYhqzGhBCA0NLfLwjre3N/v27TNZKCEexrW0LL6IO8WtW7dwPHaMX86m0ryhlxQEIYwwWhD+OsidWq1m06ZNXLhwwaShhCitY0nJfLz2OLP/2RFUkHozl5xcLQWKmib1PQho4GnpiEJUeKUausLOzo6IiAgiIiJ47bXXTJVJiFIr0Oi4eScfrU6hYR13PnrtiWo5LLIQD6NE1xDuURSFxMREbt++bdJQQgghzK/E1xDuvSjHy8uLGTNmmDyYEPejKAqZWWr9v0d3F3usrFTk5hdYOJkQlZ/RgrB+/XqCgoLMkUUIozYfvMCy7/8cPv3buf1wsLdh3/GrANhYywNoQpSV0YIwdepUtm7dao4sQhiVcTsPKxWMjwgBwMam8GH74d0DeLJtfer5uloynhCVmtGCEBAQQHx8PG3atMHJyUk/X55DEOVhyfpfOXf5JgO7NOKJNvVIzshh/ldHiyw3rHsAHYJroSigU6BPx0cN2ps+WsNckYWosowWhF27drFt2zaDeSqViqSkJJOFEtXH3p8v4+Zsj71d4T9FK5UKD1eHIsvZ2xYOX929XX3knTZCmEaxBUGtVmNnZyevuxQm1yG4Fh2CC99R4O3pSPTY0GKXrePtwsi+zcwVTYhqpdjRTocPH27OHEIIISys2IJw77Y+IUzJt4Yzrk52lo4hhOABp4zy8/P57bffii0MzZs3N1koUX0sntLV0hGEEHcVWxAuX77MSy+9dN+CoFKp2LVrl0mDCSGEMK9iC4K/vz8bNmwwZxZRxV1NzeJfSw4YzLt5J58JQ1rQp8MjlgklhNAr1eB2QjwMBztr2jX3M5hnpVIRKCORClEhFFsQ2rZta84cogq7nHyHCQt28/qzbZk0tKWl4wghilHsXUYzZ840Zw4hhBAWJqeMRLkp0Og4eS6NAo2W1oE+2NpYczn5Did+T7V0NCFECUhBEOXmx1PXeW/lMQC+frs3Hq7W7Pn5Cut2ngXA2dHWkvGEEEZIQRDlJl+tBeDN0e1xcSr88O/b8RE6BNfC3taauj4ulownhDBCCoIod/X9XLGxLrw85eXuiJe7o4UTCSFKotiLyuUhPj6evn370rNnT1atWlWkfefOnQwcOJABAwYwYcIEMjMzTRlHCCHEA5isICQnJ7No0SJWr17Nhg0bWLt2LefOndO3Z2Vl8fbbbxMTE0NcXBwBAQEsXrzYVHGEGbRo7E302FA8XO0tHUUIUQYmKwiHDh0iNDQUDw8PnJyc6NWrl8F7FQoKCoiOjsbX1xcofBHP9evXTRVHmEFND0faNvXFwU7ORApRGZmsIKSkpODt7a2f9vHxITk5WT/t6elJjx49AMjLyyMmJobu3bubKo4wg5SbORxJvE5evsbSUYQQZWCyr3I6nQ7VX15tpSiKwfQ9d+7cYeLEiQQGBjJ48OBS7SMxMbHM+RISEsq8bmVl6j7/cj6bDUdu8vIAPzxdKsZRgvydqwfpc/kw2f+1fn5+HDt2TD+dmpqKj4+PwTIpKSmMGTOG0NBQpk+fXup9BAUFYW9f+vPVCQkJtGnTptTrVWbm6PMt3SU4cpOgoCD8vJxNuq+SkL9z9SB9Lrn8/PwHfpE22Smjjh07cvjwYTIyMsjNzWXHjh106dJF367Vahk/fjx9+vRhxowZ9z16EJVDVo6ab3edZe3dB9Du3XIqhKhcTHaE4Ovry+TJkxk5ciQFBQUMGTKEkJAQxo0bR1RUFDdu3OC3335Dq9Wyfft2oPAb/5w5c0wVSZSjvHwNt7Ly8fNypkCj47/bThNQ35Ph3Zvg5e5g6XhCiDIw6Yne8PBwwsPDDeYtX74cgODgYE6fPm3K3QsT2f/LVRavO06T+p68O74Tnm4OrHizJ55uUgiEqMzk2F6UWuL/paFT4KkeAfp5UgyEqPwqxq0gokI6cS6VE7+nATCgSyPcnO04dT6ds5duYm9rTVCjmhZOKIQoT1IQRLG+2vwbZy/dwkoF3R6rh5uzHWf+yOD81UwpBkJUQVIQRLF0CrRt6kv02FD9vIiujYno2tiCqYQQpiLXEIQQQgByhCAeIDTID0d7+SciRHUh/7dXE3dytbwwbyc5eQUA1K7pwoKXHgdg1udH+P3yTYPlG9XxYNYLHcyeUwhhOVIQqonMbA3X07Jp29QXb09HPF3/vE00xL8m3p6GL7Hx9XQyd0QhhIVJQahm+nV6lLZNfQ3mDX7C30JphBAViRSEaqJWDTv+O6s3Tg7yonshxP1JQagmrK1UuLvIm8yEEMWT206ricxsDSu3/MbV1CxLRxFCVFBSEKqJO7lavt31O9fTsi0dRQhRQUlBEEIIAUhBEEIIcZcUBCGEEIAUBCGEEHfJbafVRG0vO76b3x9red+xEKIYUhCqCSuVCjtba0vHEEJUYFIQHoJWqyMtM6/IfN8aheMAZWblk6fWGrSpVOBzd5ygW3fyyS8wbLdSqfTjCt28nYdaozNot7FW4eVe2J5xO4+Cv7Xb2lhR4+7rLNMzc9FoFQAup+Zz7PsT9OnwCPX93MrUXyFE1SYF4SEsWf8rPxy9ZDDPxlrF9wsGAPDlplPs+umyQburky2rZ/cFYGnsrxw6cd2g3dvTkS9m9gRg0Tc/c/xsqkF7fT9XlkztBsC8FUc5/YfhKKUBDTz5IKoLANExh/njxp2/tKYSGlSL+n5l6KwQosqTgvAQBnRphE8NJ2q6/zlSqNVfTtH3aNeAoIaGr5q0tflzgX6dHuWxpoafzg72f57WGfSEP11a1TVod3b8cyyiYd2bkJmlNmh3d7HT/zyid1OycwuHu774x0UCmzQiqKFXSbsnhKhmpCA8hEdqufFIreJPvzRv6EXzB3wAh/h7P3D7rQN8Htj+WLMHf9XvEFxL/3OCdSptWtR54PJCiOpNbjl5CCf/L42E08mWjiGEEOVCjhAewvd7zpFxO482gb7GFxZCiApOjhCEEEIAUhCEEELcJQVBCCEEIAVBCCHEXXJR+SGMHRiE9u6TwEIIUdlJQXgItWu6WDqCEEKUG5MWhPj4eJYuXYpGo2HUqFGMGDHCoD0pKYkZM2aQnZ1N27ZtmTVrFjY2lqlROp3Cis2/ceuO4dhEzRt60Sv0EQA+XPMzOt2fRwTpmXn07fQonUJqmzOqEEKYhMk+fZOTk1m0aBGxsbHY2dnx1FNP0b59e/z9/fXLTJ06lXfffZeWLVsyffp01q1bxzPPPGOqSA90J0fN75dvkvh/6frB6QA8XR30PyddyED7l4KgKAq/X7opBUEIUSWYrCAcOnSI0NBQPDw8AOjVqxfbtm1j0qRJAFy9epW8vDxatmwJQEREBB9//LFZCsK1DDW5v141mKdSqZg3ofMD11v2r+6mjCWEEBZlsoKQkpKCt/efY/X4+Phw4sSJYtu9vb1JTi7dMBCJiYllynb8/7L5adsxg3nWVuDwVN1i1qgaEhISLB3B7KTP1YP0uXyYrCDodDpUKpV+WlEUg2lj7SURFBSEvb19qbPdzjnKqEHtDOapoEq/JyAhIYE2bdpYOoZZSZ+rB+lzyeXn5z/wi7TJCoKfnx/Hjv35LTw1NRUfHx+D9tTUP8f6T0tLM2g3JTcnaxpU4Q9/IYQoC5M9mNaxY0cOHz5MRkYGubm57Nixgy5duujb69Spg729vf6wZ+PGjQbtQgghzMtkBcHX15fJkyczcuRIBg0aRP/+/QkJCWHcuHGcPHkSgA8++IB58+bRu3dvcnJyGDlypKniCCGEMMKkN/2Hh4cTHh5uMG/58uX6nwMDA1m/fr0pIwghhCghGctICCEEIAVBCCHEXVIQhBBCAJV0cDtFKRw+Qq1Wl3kb+fn55RWn0pA+Vw/S5+qhLH2+95l57zP071RKcS0V2J07dzh79qylYwghRKXUpEkTXF1di8yvlAVBp9ORnZ2Nra1tqZ9uFkKI6kpRFAoKCnB2dsbKqugVg0pZEIQQQpQ/uagshBACkIIghBDiLikIQgghACkIQggh7pKCIIQQApCCIIQQ4i4pCEIIIYAqXhDi4+Pp27cvPXv2ZNWqVUXak5KSiIiIoFevXsyYMQONRmOBlOXLWJ937tzJwIEDGTBgABMmTCAzM9MCKcuXsT7fs2fPHrp162bGZKZjrM/nz5/nH//4BwMGDGDMmDHV4u986tQpIiMjGTBgAP/85z+5ffu2BVKWr6ysLPr378+VK1eKtJnk80upom7cuKF07dpVuXnzppKdna2Eh4crv//+u8Ey/fr1U44fP64oiqL861//UlatWmWJqOXGWJ/v3LmjdOrUSblx44aiKIry4YcfKrNnz7ZU3HJRkr+zoihKamqq0rt3b6Vr164WSFm+jPVZp9MpPXv2VPbu3asoiqK8//77yoIFCywVt1yU5O/89NNPK3v27FEURVHmzZunLFy40BJRy80vv/yi9O/fX2nevLly+fLlIu2m+PyqskcIhw4dIjQ0FA8PD5ycnOjVqxfbtm3Tt1+9epW8vDxatmwJQEREhEF7ZWSszwUFBURHR+Pr6wtAQEAA169ft1TccmGsz/fMnDmTSZMmWSBh+TPW51OnTuHk5KR/Je348eMZMWKEpeKWi5L8ne8NaQOQm5uLg4ODJaKWm3Xr1hEdHX3fd82b6vOryhaElJQUvL299dM+Pj4kJycX2+7t7W3QXhkZ67Onpyc9evQAIC8vj5iYGLp37272nOXJWJ8BVq5cSbNmzWjRooW545mEsT5funSJmjVrMn36dAYPHkx0dDROTk6WiFpuSvJ3njZtGjNnzqRz584cOnSIp556ytwxy9WcOXNo27btfdtM9flVZQuCTqczGPhOURSDaWPtlVFJ+3Tnzh1eeOEFAgMDGTx4sDkjljtjfT579iw7duxgwoQJlohnEsb6rNFoOHr0KE8//TTff/899erVY/78+ZaIWm6M9TkvL48ZM2awYsUKDhw4wDPPPMMbb7xhiahmYarPrypbEPz8/EhNTdVPp6amGhx6/b09LS3tvodmlYmxPkPhN4tnnnmGgIAA5syZY+6I5c5Yn7dt20ZqaiqRkZG88MIL+v5XZsb67O3tTYMGDQgODgagf//+nDhxwuw5y5OxPp89exZ7e3tCQkIAGD58OEePHjV7TnMx1edXlS0IHTt25PDhw2RkZJCbm8uOHTv051QB6tSpg729PQkJCQBs3LjRoL0yMtZnrVbL+PHj6dOnDzNmzKj0R0RgvM9RUVFs376djRs3EhMTg4+PD6tXr7Zg4odnrM+tWrUiIyOD06dPA7B7926aN29uqbjlwlifGzRowI0bNzh//jwAu3bt0hfEqshkn18PfVm6AouLi1P69eun9OzZU4mJiVEURVHGjh2rnDhxQlEURUlKSlIiIyOVXr16Ka+++qqSn59vybjl4kF93rFjhxIQEKAMGDBA/9/06dMtnPjhGfs733P58uUqcZeRohjv8y+//KJERkYqffv2VUaPHq2kpaVZMm65MNbnPXv2KOHh4Ur//v2VUaNGKZcuXbJk3HLTtWtX/V1Gpv78kvchCCGEAKrwKSMhhBClIwVBCCEEIAVBCCHEXVIQhBBCAFIQhBBC3GVj6QBC3BMQEECTJk2wsvrze0pQUNADH6CLjY1l+/btLFu27KH3v3jxYlatWoWvry8qlQqtVouXlxfR0dE8+uijpd5ecnIyL7/8MmvWrOHy5cssWLCAxYsXG8x/WFeuXKFHjx40adJEPy8nJwc/Pz/mzp1LvXr1Hrj+J598QmBgYKUfwkSUDykIokL56quvqFGjhsX237dvX9566y399Ndff81rr71GbGxsqbfl6+ur/9C/du0aFy5cKDK/PDg4OLBx40b9tKIovPvuuyxatIiFCxc+cN0ff/wRf3//cssiKjc5ZSQqhfXr1zN06FAGDRpE165d7/u08Y4dOxg8eDAREREMHTqUn376CSgcu2natGlEREQQHh7O3LlzSzx2fIcOHfQf5Ddu3GD8+PGEh4fTv39/Pv/8c6Bw7KDo6GjCw8OJiIggKiqK7Oxsrly5QqtWrdBqtcycOZNLly4xZswYg/lhYWEkJibq9/fKK6/o+7Z06VIGDx7MwIEDmTBhQokHL8vPzyclJQV3d3cALly4wPPPP8+wYcPo2rUrL774Ivn5+axatYrExEQWLFjADz/8gFqtZu7cuQwePJgBAwYwbdo0srKySrRPUTVIQRAVyqhRoxg4cKD+v/T0dLKzs/n222+JiYlhw4YNLFq0iPfff7/IugsWLCA6OprY2FhefvllfvzxRwDmzp1L8+bNiY2NZcOGDdy8eZMvv/zSaBaNRsP69etp3749AFOmTKF9+/bEx8fzzTffEBcXx+bNm/nll184evQocXFxxMbGUq9ePc6cOaPfjrW1Ne+++y7169fnP//5j8H8yMhI/dFHZmYmhw8fJjw8nA0bNnD27Fm+/fZbNm7cSFhYGDNnzrxvzry8PAYOHEh4eDgdO3Zk8ODBNGzYkClTpgCFwygPGjSIdevWsWPHDq5cucKePXsYMWIEQUFBvP766/To0YOYmBisra2JjY0lLi4OHx8fPvjggxL+5URVIKeMRIVS3Cmjzz77jL1793Lx4kVOnz5NTk5OkWX69evHpEmTCAsLo1OnTowbNw4ofFPayZMnWb9+PVD4AVqcLVu26MeHKSgooHnz5syePZucnBx+/vlnvvjiCwBcXV2JiIhg3759zJgxA2tra4YOHUrnzp3p1asXISEh933L1d9FRkYyZMgQpk2bxqZNm+jWrRuurq7873//4+TJk0RGRgKFo1vm5ubedxt/PWW0f/9+pk6dSteuXXF2dgZg6tSpHDx4kOXLl3Px4kVSUlLu+/vbs2cPd+7c4dChQ/r+e3l5Ge2DqDqkIIgK78aNGwwfPpxhw4bRpk0bevfuzf/+978iy02ePJnIyEgOHjxIbGwsX3zxBevXr0en0/HRRx/RqFEjAG7fvl3swH5/v4ZwT1ZWFn8f5UWn06HRaHBzc2Pjxo38/PPPHDlyhFdeeYUxY8YQFhZmtG916tShWbNm7Nmzh9jYWKZPn67f9tixY/Ujs6rV6hK9BvPxxx/n+eef5+WXX2bz5s24uLjw6quvotVq6dOnD0888QTXr18v0pd7+5w+fbo+d3Z2Nvn5+Ub3KaoOOWUkKrzExERq1KjBhAkT6Ny5s74YaLVa/TIajYZu3bqRm5vL008/TXR0NGfOnEGtVtO5c2dWrFiBoiio1WpefPFF/vvf/5Yqg4uLCy1atNC/y/fOnTts2LCBjh078r///Y/nnnuOVq1a8dJLLzFo0CCD6wJQeHqooKDgvtseNmwYy5cvJzc3lzZt2gDQuXNn1q9frz+H/9FHH/H666+XKOvo0aNxdnbm448/BuDAgQNMnDiRvn37AvDrr7/qf3fW1tb66ymdO3dm1apVqNVqdDodb775ptGL0qJqkSMEUeF16tSJ9evX07t3b1QqFe3ataNGjRr88ccf+mVsbGyYPn06U6ZMwcbGBpVKxdy5c7Gzs2PGjBnMmTOH8PBwCgoK6NixI2PHji11jg8++IB33nmH2NhY1Gq1/iKyTqdj37599O/fHycnJ9zd3Zk9e7bBuv7+/tjb2zNkyBAWLVpk0NatWzdmzZqlP8UFMHToUJKTkxk2bBgqlYpatWqV+CU3tra2vPnmm4wdO5YhQ4YwefJkJk6ciJOTEy4uLjz22GNcunRJv++FCxdSUFDAhAkTeO+99xg8eDBarZamTZsybdq0Uv+eROUlo50KIYQA5JSREEKIu6QgCCGEAKQgCCGEuEsKghBCCEAKghBCiLukIAghhACkIAghhLhLCoIQQggA/h8wx+qvNMgY3AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Combined Model AUC ROC:  0.484812529663028\n"
     ]
    }
   ],
   "source": [
    "list_of_lists = []\n",
    "\n",
    "for col in tox_list:\n",
    "    \n",
    "    method = \"ErG\"\n",
    "    print(\"Start\")\n",
    "    toxcast_counts=our_data[col].value_counts()\n",
    "    print(our_data[col].value_counts())\n",
    "    pyplot.figure()\n",
    "    sns.barplot(toxcast_counts.index, toxcast_counts.values, alpha=0.8)\n",
    "    pyplot.title('Distribution of 0 and 1')\n",
    "    pyplot.ylabel('Number of Occurrences', fontsize=12)\n",
    "    pyplot.xlabel(col, fontsize=12)\n",
    "    pyplot.show()\n",
    "    \n",
    "    n_cluster=4 \n",
    "        \n",
    "    target=col\n",
    "    \n",
    "    \n",
    "    df_temp=pd.merge(cp, toxcast[[target, 'smiles_r']], left_on = 'smiles_r', right_on = 'smiles_r', how = 'inner')\n",
    "    df_temp['ErGFingerprint'] =df_temp['smiles_r'].apply(ErGFingerprint)\n",
    "    df_temp=df_temp.dropna(subset=[target])\n",
    "    df_temp=df_temp.drop('Unnamed: 0', axis = 1)\n",
    "\n",
    "    X = np.array([x for x in df_temp['ErGFingerprint']])\n",
    "    \n",
    "    df_temp=df_temp.drop('ErGFingerprint', axis = 1)\n",
    "    data=df_temp\n",
    "    \n",
    "    data=data.drop('smiles_r', axis = 1)\n",
    "    y = data[target]\n",
    "\n",
    "    y=y.to_numpy()\n",
    "    \n",
    "    print(\"Total Compounds in Assay: \", len(X))\n",
    "    #PCA to groups\n",
    "    \n",
    "    df_temp['MolWt']=df_temp['smiles_r'].apply(GetExactMolWt)\n",
    "    df_temp['TPSA']=df_temp['smiles_r'].apply(GetCalcTPSA) #Topological Polar Surface Area\n",
    "    df_temp['nRotB']=df_temp['smiles_r'].apply(GetNumRotatableBonds) #Number of rotable bonds\n",
    "    df_temp['HBD']=df_temp['smiles_r'].apply(GetNumHDonors) #Number of H bond donors\n",
    "    df_temp['HBA']=df_temp['smiles_r'].apply(GetNumHAcceptors) #Number of H bond acceptors\n",
    "    df_temp['LogP']=df_temp['smiles_r'].apply(GetMolLogP) #LogP\n",
    "\n",
    "    descriptors = df_temp.loc[:, ['MolWt', 'TPSA', 'nRotB', 'HBD','HBA', 'LogP']].values\n",
    "    descriptors_std = StandardScaler().fit_transform(descriptors)\n",
    "    \n",
    "    pca = PCA()\n",
    "    descriptors_2d = pca.fit_transform(descriptors_std)\n",
    "    descriptors_pca= pd.DataFrame(descriptors_2d)\n",
    "    descriptors_pca.index = df_temp.index\n",
    "    descriptors_pca.columns = ['PC{}'.format(i+1) for i in descriptors_pca.columns]\n",
    "\n",
    "    print(\"pca explained variance by first two components: \", (pca.explained_variance_ratio_[0]+ pca.explained_variance_ratio_[1]))\n",
    "\n",
    "    # This normalization will be performed just for PC1 and PC2, but can be done for all the components.\n",
    "    scale1 = 1.0/(max(descriptors_pca['PC1']) - min(descriptors_pca['PC1']))\n",
    "    scale2 = 1.0/(max(descriptors_pca['PC2']) - min(descriptors_pca['PC2']))\n",
    "\n",
    "    # And we add the new values to our PCA table\n",
    "    descriptors_pca['PC1_normalized']=[i*scale1 for i in descriptors_pca['PC1']]\n",
    "    descriptors_pca['PC2_normalized']=[i*scale2 for i in descriptors_pca['PC2']]\n",
    "\n",
    "    kmeans = KMeans(n_clusters=n_cluster, random_state=0)\n",
    "    cluster_labels = kmeans.fit_predict(descriptors_pca[['PC1_normalized','PC2_normalized']])\n",
    "    \n",
    "  \n",
    "    colors = cm.nipy_spectral(cluster_labels.astype(float) / n_cluster)\n",
    "    plt.scatter(descriptors_pca['PC1_normalized'], descriptors_pca['PC2_normalized'], marker='.', s=30, lw=0, alpha=0.7,c=colors, edgecolor='k')\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "    df_temp['Cluster_PC1_PC2'] = pd.Series(kmeans.labels_, index=df_temp.index)\n",
    "    print(\"Start\")\n",
    "    cluster_counts=df_temp['Cluster_PC1_PC2'].value_counts()\n",
    "    print(df_temp['Cluster_PC1_PC2'].value_counts())\n",
    "    plt.figure()\n",
    "    sns.barplot(cluster_counts.index, cluster_counts.values, alpha=0.8)\n",
    "    plt.title('Distribution of datapoints in each cluster')\n",
    "    plt.ylabel('Number of Occurrences', fontsize=12)\n",
    "    plt.xlabel('Cluster_PC1_PC2', fontsize=12)\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "    outercv = GroupKFold(n_splits=n_cluster)\n",
    "    outercv.get_n_splits(X, y, kmeans.labels_)\n",
    "    \n",
    "    tprs = []\n",
    "    aucs = []\n",
    "    mean_fpr = np.linspace(0, 1, 100)\n",
    "\n",
    "    fig, ax = plt.subplots()\n",
    "    \n",
    "    mat_cof=[]\n",
    "    ba=[]\n",
    "    sen=[]\n",
    "    spe=[]\n",
    "    combined_y_test=[]\n",
    "    combined_y_pred=[]\n",
    "    \n",
    "    for i, (train_index, test_index) in enumerate(outercv.split(X, y, kmeans.labels_)):\n",
    "        \n",
    "        \n",
    "        X_train, X_test = X[train_index], X[test_index]\n",
    "        y_train, y_test = y[train_index], y[test_index]\n",
    "        \n",
    "        print(\"Total Selected train: \",  len(X_train))\n",
    "        print(\"Total Selected test: \",  len(X_test))\n",
    "        \n",
    "        inner_cv = StratifiedKFold(n_splits=5, shuffle=True, random_state=0)\n",
    "           \n",
    "        param_grid = { \n",
    "        'max_depth': [15, 10, 20],\n",
    "        'max_features': ['auto'],\n",
    "        'min_samples_leaf': [3, 6, 12, 15],\n",
    "        'min_samples_split': [6, 9, 12, 15],\n",
    "        'n_estimators':[100, 200, 300, 700],\n",
    "        'bootstrap': [False],\n",
    "        'oob_score': [False],\n",
    "        'random_state': [0],\n",
    "        'criterion': ['gini', 'entropy'],\n",
    "        'n_jobs': [-1],\n",
    "        'class_weight' : [None, 'balanced']\n",
    "        }\n",
    "\n",
    "       # Create a based model\n",
    "        rf = RandomForestClassifier()\n",
    "        # Instantiate the grid search model\n",
    "        grid_search = GridSearchCV(estimator = rf, param_grid = param_grid, \n",
    "                                  cv = inner_cv, n_jobs = 40, verbose = 2)\n",
    "\n",
    "        # Fit the grid search to the data\n",
    "        grid_search.fit(X_train, y_train)\n",
    "        print(\"Tuned Mode: \", grid_search.best_estimator_)\n",
    "\n",
    "        classifier = grid_search.best_estimator_\n",
    "        \n",
    "        classifier.fit(X_train, y_train)\n",
    "        \n",
    "        y_pred = classifier.predict(X_test)\n",
    "        y_proba = classifier.predict_proba(X_test)\n",
    "       \n",
    "        combined_y_test=np.concatenate((combined_y_test, y_test), axis=None)\n",
    "        combined_y_pred=np.concatenate((combined_y_pred, y_pred), axis=None)\n",
    "        \n",
    "        if i==0:\n",
    "            combined_y_proba=y_proba\n",
    "        else:\n",
    "            combined_y_proba=np.concatenate((combined_y_proba, y_proba), axis=0)\n",
    "    \n",
    "    \n",
    "        conf_matrix = confusion_matrix(y_test, y_pred)\n",
    "        print(\"In this fold\")\n",
    "        print(\"conf_matrix: \\n \" , conf_matrix)\n",
    "        print(\"classification_report: \\n\",classification_report(y_test, y_pred))\n",
    "    \n",
    "        mat_cof.append(matthews_corrcoef(y_test, y_pred))\n",
    "        ba.append(balanced_accuracy_score(y_test, y_pred))\n",
    "        \n",
    "        sen.append(conf_matrix[0,0]/(conf_matrix[0,0]+conf_matrix[0,1]))\n",
    "        spe.append(conf_matrix[1,1]/(conf_matrix[1,0]+conf_matrix[1,1]))\n",
    "\n",
    "        \n",
    "        viz = plot_roc_curve(classifier, X_test, y_test,\n",
    "                             name='ROC fold {}'.format(1+i),\n",
    "                             alpha=0.3, lw=1, ax=ax)\n",
    "        \n",
    "        interp_tpr = np.interp(mean_fpr, viz.fpr, viz.tpr)\n",
    "        interp_tpr[0] = 0.0\n",
    "        tprs.append(interp_tpr)\n",
    "        aucs.append(viz.roc_auc)\n",
    "\n",
    "    ax.plot([0, 1], [0, 1], linestyle='--', lw=2, color='r',\n",
    "            label='Chance', alpha=.8)\n",
    "\n",
    "    mean_tpr = np.mean(tprs, axis=0)\n",
    "    mean_tpr[-1] = 1.0\n",
    "    mean_auc = auc(mean_fpr, mean_tpr)\n",
    "    std_auc = np.std(aucs)\n",
    "    ax.plot(mean_fpr, mean_tpr, color='b',\n",
    "            label=r'Mean ROC (AUC = %0.2f $\\pm$ %0.2f)' % (mean_auc, std_auc),\n",
    "            lw=2, alpha=.8)\n",
    "\n",
    "    std_tpr = np.std(tprs, axis=0)\n",
    "    tprs_upper = np.minimum(mean_tpr + std_tpr, 1)\n",
    "    tprs_lower = np.maximum(mean_tpr - std_tpr, 0)\n",
    "    ax.fill_between(mean_fpr, tprs_lower, tprs_upper, color='grey', alpha=.2,\n",
    "                    label=r'$\\pm$ 1 std. dev.')\n",
    "\n",
    "    ax.set(xlim=[-0.05, 1.05], ylim=[-0.05, 1.05],\n",
    "           title=\"Receiver operating characteristic example\")\n",
    "    ax.legend(loc=\"lower right\")\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "    \n",
    "    print(\"Mean of Each fold\")\n",
    "    \n",
    "    sns.boxplot(x=aucs).set(title = 'AUC-ROC Score', xlabel = 'AUC-ROC Score' )\n",
    "    plt.show()\n",
    "    print(\"AUC-ROC all: \",(aucs))\n",
    "    print(r\"AUC-ROC = %0.2f ± %0.2f)\" % (np.mean(aucs), np.std(aucs)))\n",
    "    row=[col, method, 'AUC-ROC'] + aucs\n",
    "    list_of_lists.append(row)\n",
    "    sns.set(style=\"whitegrid\")    \n",
    "    \n",
    "    sns.set(style=\"whitegrid\")    \n",
    "    sns.boxplot(x=mat_cof).set(title = 'MCC Score', xlabel = 'MCC Score' )\n",
    "    plt.show()\n",
    "    print(\"Matthews_corrcoef all: \",(mat_cof))\n",
    "    print(r\"Matthews_corrcoef = %0.3f ± %0.3f)\" % (np.mean(mat_cof), np.std(mat_cof)))\n",
    "    row=[col, method, 'MCC'] + mat_cof\n",
    "    list_of_lists.append(row)\n",
    "    \n",
    "    sns.boxplot(x=ba).set(title = 'balanced_accuracy', xlabel = 'balanced_accuracy Score' )\n",
    "    plt.show() \n",
    "    print(\"balanced_accuracy all: \",(ba))\n",
    "    print(r\"balanced_accuracy = %0.2f ± %0.2f)\" % (np.mean(ba), np.std(ba)))\n",
    "    row=[col, method, 'BA'] + ba\n",
    "    list_of_lists.append(row)\n",
    "    \n",
    "    sns.boxplot(x=sen).set(title = 'Sensitivity', xlabel = 'Sensitivity' )\n",
    "    plt.show()\n",
    "    print(\"Sensitivity all: \",(sen))\n",
    "    print(r\"Sensitivity = %0.2f ± %0.2f)\" % (np.mean(sen), np.std(sen)))\n",
    "    row=[col, method, 'SEN'] + sen\n",
    "    list_of_lists.append(row)\n",
    "    \n",
    "    sns.boxplot(x=spe).set(title = 'Specificity', xlabel = 'Specificity' )\n",
    "    plt.show()\n",
    "    print(\"Specificity all: \",(spe))\n",
    "    print(r\"Specificity = %0.2f ± %0.2f)\" % (np.mean(spe), np.std(spe)))\n",
    "    row=[col, method, 'SPE'] + spe\n",
    "    list_of_lists.append(row)\n",
    "    \n",
    "    \n",
    "    \n",
    "    \n",
    "    \n",
    "    print(\"One Combined Confusion Matrix\")\n",
    "    combined_conf_matrix = confusion_matrix(combined_y_test, combined_y_pred)\n",
    "    print(\"conf_matrix: \\n \" , combined_conf_matrix)\n",
    "    print(\"classification_report: \\n\",classification_report(combined_y_test, combined_y_pred))\n",
    "    \n",
    "    print('matthews_corrcoef')\n",
    "    print(matthews_corrcoef(combined_y_test, combined_y_pred))\n",
    "    \n",
    "    print('balanced_accuracy_score')\n",
    "    print(balanced_accuracy_score(combined_y_test, combined_y_pred))\n",
    "    \n",
    "    print('Sensitivity')   \n",
    "    print(combined_conf_matrix[0,0]/(combined_conf_matrix[0,0]+combined_conf_matrix[0,1]))\n",
    "    \n",
    "    print('Specificity')  \n",
    "    print(combined_conf_matrix[1,1]/(combined_conf_matrix[1,0]+combined_conf_matrix[1,1]))\n",
    "       \n",
    "    print(\"AUC-ROC for the combined results: \")\n",
    "    \n",
    "    probs_combined_model= combined_y_proba[:,1]\n",
    "    roc_value_combined_model= roc_auc_score(combined_y_test, probs_combined_model) \n",
    "    base_model_fpr, base_model_tpr, _ = roc_curve(combined_y_test, probs_combined_model)\n",
    "    plt.plot(base_model_fpr, base_model_tpr, linestyle='--', label='Combined Model AUCROC')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    # show the legend\n",
    "    plt.legend()\n",
    "    # show the plot\n",
    "    plt.show()\n",
    "    print(\"Combined Model AUC ROC: \", roc_value_combined_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "df=pd.DataFrame(list_of_lists,columns=[\"target\", \"method\", \"metric\", \"1\", \"2\", \"3\", \"4\"])\n",
    "df['mean']=df.drop(['target', 'method', 'metric'], axis=1).apply(lambda x: x.mean(), axis=1)\n",
    "df['std']=df.drop(['target', 'method','metric'], axis=1).apply(lambda x: x.std(), axis=1)\n",
    "df['value']= df['mean'].round(2).astype(str) + \" ± \" + df['std'].round(2).astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "      <th>method</th>\n",
       "      <th>metric</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>combined_assay</td>\n",
       "      <td>ErG</td>\n",
       "      <td>AUC-ROC</td>\n",
       "      <td>0.563025</td>\n",
       "      <td>0.529091</td>\n",
       "      <td>0.590476</td>\n",
       "      <td>0.862500</td>\n",
       "      <td>0.636273</td>\n",
       "      <td>0.132410</td>\n",
       "      <td>0.64 ± 0.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>combined_assay</td>\n",
       "      <td>ErG</td>\n",
       "      <td>MCC</td>\n",
       "      <td>-0.059841</td>\n",
       "      <td>0.234521</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.516398</td>\n",
       "      <td>0.172769</td>\n",
       "      <td>0.226854</td>\n",
       "      <td>0.17 ± 0.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>combined_assay</td>\n",
       "      <td>ErG</td>\n",
       "      <td>BA</td>\n",
       "      <td>0.468487</td>\n",
       "      <td>0.580000</td>\n",
       "      <td>0.500000</td>\n",
       "      <td>0.900000</td>\n",
       "      <td>0.612122</td>\n",
       "      <td>0.171105</td>\n",
       "      <td>0.61 ± 0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>combined_assay</td>\n",
       "      <td>ErG</td>\n",
       "      <td>SEN</td>\n",
       "      <td>0.142857</td>\n",
       "      <td>0.160000</td>\n",
       "      <td>0.333333</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.409048</td>\n",
       "      <td>0.349228</td>\n",
       "      <td>0.41 ± 0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>combined_assay</td>\n",
       "      <td>ErG</td>\n",
       "      <td>SPE</td>\n",
       "      <td>0.794118</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.666667</td>\n",
       "      <td>0.800000</td>\n",
       "      <td>0.815196</td>\n",
       "      <td>0.119257</td>\n",
       "      <td>0.82 ± 0.12</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           target method   metric         1         2         3         4  \\\n",
       "0  combined_assay    ErG  AUC-ROC  0.563025  0.529091  0.590476  0.862500   \n",
       "1  combined_assay    ErG      MCC -0.059841  0.234521  0.000000  0.516398   \n",
       "2  combined_assay    ErG       BA  0.468487  0.580000  0.500000  0.900000   \n",
       "3  combined_assay    ErG      SEN  0.142857  0.160000  0.333333  1.000000   \n",
       "4  combined_assay    ErG      SPE  0.794118  1.000000  0.666667  0.800000   \n",
       "\n",
       "       mean       std        value  \n",
       "0  0.636273  0.132410  0.64 ± 0.13  \n",
       "1  0.172769  0.226854  0.17 ± 0.23  \n",
       "2  0.612122  0.171105  0.61 ± 0.17  \n",
       "3  0.409048  0.349228  0.41 ± 0.35  \n",
       "4  0.815196  0.119257  0.82 ± 0.12  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "file_name=\"Metrics_\"+method+\".csv\"\n",
    "df.to_csv(file_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (myenv)",
   "language": "python",
   "name": "my-rdkit-env"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}