Task 2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "OUGUW9Vjxdou"
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from pandas.core.common import SettingWithCopyWarning\n",
    "import warnings\n",
    "warnings.simplefilter(action=\"ignore\", category=SettingWithCopyWarning)\n",
    "\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "from sklearn.model_selection import train_test_split, cross_val_score\n",
    "\n",
    "from sklearn.preprocessing import Normalizer\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "\n",
    "#classifiers\n",
    "from sklearn.gaussian_process.kernels import RBF\n",
    "from sklearn.gaussian_process import GaussianProcessClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "import xgboost as xgb\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import VotingClassifier, StackingClassifier, BaggingClassifier\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier, AdaBoostClassifier, GradientBoostingClassifier, ExtraTreesClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n",
    "from sklearn.ensemble import BaggingClassifier, StackingClassifier\n",
    "\n",
    "#Prformance metrics\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import roc_curve, auc, roc_auc_score, accuracy_score, confusion_matrix, classification_report\n",
    "from sklearn.metrics import make_scorer\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import recall_score\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.model_selection import cross_validate\n",
    "\n",
    "# Dimensionality reduction\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.linear_model import Lasso\n",
    "from sklearn.pipeline import Pipeline\n",
    "\n",
    "from sklearn.exceptions import ConvergenceWarning\n",
    "# import warnings filter\n",
    "from warnings import simplefilter\n",
    "# ignore all future warnings\n",
    "simplefilter(action='ignore', category=FutureWarning)\n",
    "\n",
    "simplefilter(action='ignore', category=ConvergenceWarning)\n",
    "\n",
    "simplefilter(action='ignore', category=UserWarning)\n",
    "\n",
    "from statistics import mean\n",
    "\n",
    "# Explainable AI\n",
    "import shap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "0C-fdgXRp5Xl"
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "plt.rcParams['figure.figsize']=15,10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "57F9c0q5FZEd"
   },
   "outputs": [],
   "source": [
    "def Algo_search(models , params, X_train, y_train, X_test, y_test):\n",
    "\n",
    "    max_score = 0\n",
    "    max_model = None\n",
    "    max_model_params = None\n",
    "    estimators_gridsearch = pd.DataFrame()\n",
    "    models_est = []\n",
    "    parametres = []\n",
    "    lscore = []\n",
    "    lroc = []\n",
    "    lfscore = []\n",
    "    lprecision = []\n",
    "    lrecall = []\n",
    "    \n",
    "    scoring = {'accuracy': make_scorer(accuracy_score), \n",
    "               'roc_auc': make_scorer(roc_auc_score, needs_threshold=True),\n",
    "               'f1': make_scorer(f1_score), \n",
    "               'precision': make_scorer(precision_score), \n",
    "               'recall': make_scorer(recall_score),}\n",
    "\n",
    "    for i,j in zip(models.keys() , models.values() ):\n",
    "\n",
    "        gs = GridSearchCV(estimator=j,param_grid=params[i], scoring=scoring, refit = \"accuracy\", )\n",
    "        a = gs.fit(X_train,y_train)\n",
    "        score = gs.score(X_test,y_test )\n",
    "        mauc = []\n",
    "        mprecision = []\n",
    "        mrecall = []\n",
    "        mfscore = []\n",
    "        maccuracy = []\n",
    "        for i in range(5):\n",
    "          maccuracy.append(mean(gs.cv_results_['split'+str(i)+'_test_accuracy']))\n",
    "          mauc.append(mean(gs.cv_results_['split'+str(i)+'_test_roc_auc']))\n",
    "          mfscore.append(mean(gs.cv_results_['split'+str(i)+'_test_f1']))\n",
    "          mprecision.append(mean(gs.cv_results_['split'+str(i)+'_test_precision']))\n",
    "          mrecall.append(mean(gs.cv_results_['split'+str(i)+'_test_recall']))\n",
    "        models_est.append(gs.best_estimator_)\n",
    "        parametres.append(gs.best_params_)\n",
    "        lscore.append(round(score,2))\n",
    "        lroc.append(round(mean(mauc),2))\n",
    "        lfscore.append(round(mean(mfscore),2))\n",
    "        lprecision.append(round(mean(mprecision),2))\n",
    "        lrecall.append(round(mean(mrecall),2))\n",
    "\n",
    "        if score > max_score:\n",
    "            max_score = score\n",
    "            max_model = gs.best_estimator_\n",
    "            max_model_params = gs.best_params_\n",
    "\n",
    "    estimators_gridsearch['Models'] = list(models_est)\n",
    "    estimators_gridsearch['Best parameters'] = list(parametres)\n",
    "    estimators_gridsearch['Score'] = list(lscore)\n",
    "    estimators_gridsearch['AUC'] = list(lroc)\n",
    "    estimators_gridsearch['F1-score'] = list(lfscore)\n",
    "    estimators_gridsearch['Precision'] = list(lprecision)\n",
    "    estimators_gridsearch['Recall'] = list(lrecall)\n",
    "   \n",
    "\n",
    "    # print(estimators_gridsearch)\n",
    "\n",
    "    return [max_score, max_model, max_model_params], estimators_gridsearch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "dVs3bvgn7Da5"
   },
   "outputs": [],
   "source": [
    "#Data points\n",
    "\n",
    "models = {'model_gbm':GradientBoostingClassifier(), 'model_rf':RandomForestClassifier(), \n",
    "  'model_dt':DecisionTreeClassifier(), 'model_svm':SVC(), 'model_ada':AdaBoostClassifier(), 'model_knn': KNeighborsClassifier(), \n",
    "  'model_xgb': xgb.XGBClassifier(),'model_ext': ExtraTreesClassifier(), 'model_mlp': MLPClassifier()}\n",
    "params_gbm = {'learning_rate':[0.1,0.2,0.3,0.4], 'n_estimators':[50,100,500,1000,2000]}\n",
    "params_rf = {'n_estimators':[50,100,500,1000,2000], 'max_depth' : [5, 10, 20], 'class_weight' : [None, {0:1,1:5}, {0:1,1:10}, {0:1,1:25}]}\n",
    "params_dt = {'splitter':['best','random'], 'max_depth':[1, 5, 10, 50, 100]}\n",
    "params_svm = {'C':[1,2,5,10,50,100,500], 'kernel':['rbf','poly','sigmoid','linear']}\n",
    "params_ada = {'n_estimators':[10,20,30,50,100,500,1000], 'learning_rate':[0.5,1,2,5,10]}\n",
    "params_knn = {'n_neighbors' : [2,5,10,25,50]}\n",
    "params_xgb = {'gamma': [0,0.1,0.2,0.4,0.8,1.6,3.2,6.4,12.8,25.6,51.2,102.4, 200],\n",
    "              'learning_rate': [0.01, 0.03, 0.06, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7],\n",
    "              'max_depth': [5,6,7,8,9,10,11,12,13,14],\n",
    "              'n_estimators': [50,65,80,100,115,130,150],\n",
    "              'reg_alpha': [0,0.1,0.2,0.4,0.8,1.6,3.2,6.4,12.8,25.6,51.2,102.4,200],\n",
    "              'reg_lambda': [0,0.1,0.2,0.4,0.8,1.6,3.2,6.4,12.8,25.6,51.2,102.4,200]}\n",
    "params_ext = {'n_estimators':[50,100,500,1000,2000], 'max_depth' : [5, 10, 20], 'class_weight' : [None, {0:1,1:5}, {0:1,1:10}, {0:1,1:25}]}\n",
    "model_mlp = {'solver': ['lbfgs', 'sgd', 'adam'], 'activation': ['identity', 'logistic', 'tanh', 'relu'], \n",
    "             'learning_rate':['constant','adaptive'], 'alpha': [0.0001, 0.05],\n",
    "             'hidden_layer_sizes': [(50,50,50), (50,100,50), (100,)]}\n",
    "params = {'model_gbm':params_gbm, 'model_rf':params_rf, 'model_dt':params_dt, 'model_svm':params_svm, 'model_ada':params_ada, 'model_knn': params_knn,\n",
    "          'model_xgb':params_xgb, 'model_ext':params_ext, 'model_mlp': model_mlp}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "ijCRy0q1LQCs"
   },
   "outputs": [],
   "source": [
    "mgh = pd.read_excel('mghML.xlsx')\n",
    "icl = pd.read_excel('iclML.xlsx')\n",
    "yps = pd.read_excel('ypsML.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "_0Yx1hDhq9m6"
   },
   "outputs": [],
   "source": [
    "common_columns = list(set(mgh) & set(icl) & set(yps))\n",
    "mgh_c = mgh[common_columns]\n",
    "icl_c = icl[common_columns]\n",
    "yps_c = yps[common_columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 81
    },
    "executionInfo": {
     "elapsed": 315,
     "status": "ok",
     "timestamp": 1652081195929,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "4NP4xF8jKXHX",
    "outputId": "f3b22664-0323-4081-9f97-d5af76547575"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-d2d6feb0-b572-4601-8192-4900c6d19e7c\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>column_name</th>\n",
       "      <th>percent_missing</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>CD6</th>\n",
       "      <td>CD6</td>\n",
       "      <td>100.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-d2d6feb0-b572-4601-8192-4900c6d19e7c')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-d2d6feb0-b572-4601-8192-4900c6d19e7c button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-d2d6feb0-b572-4601-8192-4900c6d19e7c');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "    column_name  percent_missing\n",
       "CD6         CD6            100.0"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "yps_c = yps_c.drop([124, 125, 126])\n",
    "percent_missing = yps_c.isnull().sum() * 100 / len(yps_c)\n",
    "missing_value_df = pd.DataFrame({'column_name': yps_c.columns,\n",
    "                                 'percent_missing': percent_missing})\n",
    "missing_value_df.loc[missing_value_df['percent_missing']> 0.000] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "kvUoq5sugngx"
   },
   "outputs": [],
   "source": [
    "mgh_con = mgh_c.drop('CD6', axis=1)\n",
    "icl_con = icl_c.drop('CD6', axis=1)\n",
    "yps_con = yps_c.drop('CD6', axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "pu8sFnRlhBbj"
   },
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "col = mgh_con.drop(['WHO max', 'KIDNEY', 'Age cat'], axis=1).columns\n",
    "\n",
    "sc = StandardScaler()\n",
    "\n",
    "for i in col:\n",
    "    mgh_con[[i]] = sc.fit_transform(mgh_con[[i]])\n",
    "    icl_con[[i]] = sc.fit_transform(icl_con[[i]])\n",
    "    yps_con[[i]] = sc.fit_transform(yps_con[[i]])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "pl8jYkdj9SMb"
   },
   "source": [
    "# MGH dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "VJMy-ADFbE2l"
   },
   "outputs": [],
   "source": [
    "X = mgh_con.drop(['WHO max'], axis=1)\n",
    "y = mgh_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "onghoHvNweNO"
   },
   "outputs": [],
   "source": [
    "X_test1 = icl_con.drop(['WHO max'], axis=1)\n",
    "y_test1 = icl_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "tBSSEQnHwgU8"
   },
   "outputs": [],
   "source": [
    "X_test2 = yps_con.drop(['WHO max'], axis=1)\n",
    "y_test2 = yps_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "yH17xSerBtV8"
   },
   "outputs": [],
   "source": [
    "pca = PCA(.90)\n",
    " \n",
    "\n",
    "X = pca.fit_transform(X[col])\n",
    "X = pd.DataFrame(X)\n",
    "X['KIDNEY'] = list(mgh_con['KIDNEY'])\n",
    "X['Age cat'] = list(mgh_con['Age cat'])\n",
    "\n",
    "X_test1 = pca.transform(X_test1[col])\n",
    "X_test1 = pd.DataFrame(X_test1)\n",
    "X_test1['KIDNEY'] = list(icl_con['KIDNEY'])\n",
    "X_test1['Age cat'] = list(icl_con['Age cat'])\n",
    "\n",
    "X_test2 = pca.transform(X_test2[col])\n",
    "X_test2 = pd.DataFrame(X_test2)\n",
    "X_test2['KIDNEY'] = list(yps_con['KIDNEY'])\n",
    "X_test2['Age cat'] = list(yps_con['Age cat'])\n",
    " \n",
    "explained_variance = pca.explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 488
    },
    "executionInfo": {
     "elapsed": 359,
     "status": "ok",
     "timestamp": 1651839734607,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "JND-C4MmOloR",
    "outputId": "fa7759a1-4906-4409-e960-9c028783cb9f"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-8b4e003d-7372-40cf-8b08-9576fdd4d0e4\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>49</th>\n",
       "      <th>50</th>\n",
       "      <th>51</th>\n",
       "      <th>52</th>\n",
       "      <th>53</th>\n",
       "      <th>54</th>\n",
       "      <th>55</th>\n",
       "      <th>56</th>\n",
       "      <th>57</th>\n",
       "      <th>58</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>TNFRSF10A</th>\n",
       "      <td>0.797710</td>\n",
       "      <td>-0.204731</td>\n",
       "      <td>-0.234727</td>\n",
       "      <td>-0.115882</td>\n",
       "      <td>-0.061635</td>\n",
       "      <td>0.013233</td>\n",
       "      <td>-0.056326</td>\n",
       "      <td>-0.008419</td>\n",
       "      <td>0.150010</td>\n",
       "      <td>-0.037909</td>\n",
       "      <td>...</td>\n",
       "      <td>0.076022</td>\n",
       "      <td>0.036356</td>\n",
       "      <td>-0.095167</td>\n",
       "      <td>0.042071</td>\n",
       "      <td>0.003061</td>\n",
       "      <td>0.037474</td>\n",
       "      <td>-0.011173</td>\n",
       "      <td>0.013604</td>\n",
       "      <td>-0.024180</td>\n",
       "      <td>0.051163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CXCL1</th>\n",
       "      <td>0.253141</td>\n",
       "      <td>0.138941</td>\n",
       "      <td>0.044449</td>\n",
       "      <td>0.143732</td>\n",
       "      <td>-0.339333</td>\n",
       "      <td>-0.106429</td>\n",
       "      <td>0.207265</td>\n",
       "      <td>-0.130248</td>\n",
       "      <td>-0.402272</td>\n",
       "      <td>-0.122429</td>\n",
       "      <td>...</td>\n",
       "      <td>0.080746</td>\n",
       "      <td>0.040365</td>\n",
       "      <td>-0.014334</td>\n",
       "      <td>0.061180</td>\n",
       "      <td>-0.087790</td>\n",
       "      <td>-0.108023</td>\n",
       "      <td>-0.008110</td>\n",
       "      <td>0.003904</td>\n",
       "      <td>-0.057877</td>\n",
       "      <td>0.128499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ITM2A</th>\n",
       "      <td>0.633502</td>\n",
       "      <td>-0.350919</td>\n",
       "      <td>-0.048621</td>\n",
       "      <td>-0.102820</td>\n",
       "      <td>-0.030706</td>\n",
       "      <td>0.148995</td>\n",
       "      <td>0.122867</td>\n",
       "      <td>-0.088981</td>\n",
       "      <td>0.088043</td>\n",
       "      <td>0.058009</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008549</td>\n",
       "      <td>-0.069752</td>\n",
       "      <td>-0.167011</td>\n",
       "      <td>0.122260</td>\n",
       "      <td>0.086425</td>\n",
       "      <td>0.142849</td>\n",
       "      <td>0.107072</td>\n",
       "      <td>0.001171</td>\n",
       "      <td>-0.006032</td>\n",
       "      <td>-0.056704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SIT1</th>\n",
       "      <td>0.130831</td>\n",
       "      <td>0.179770</td>\n",
       "      <td>0.253993</td>\n",
       "      <td>0.248916</td>\n",
       "      <td>0.209400</td>\n",
       "      <td>0.472950</td>\n",
       "      <td>-0.289355</td>\n",
       "      <td>0.135301</td>\n",
       "      <td>-0.146441</td>\n",
       "      <td>0.347028</td>\n",
       "      <td>...</td>\n",
       "      <td>0.042623</td>\n",
       "      <td>-0.037030</td>\n",
       "      <td>0.006670</td>\n",
       "      <td>0.016537</td>\n",
       "      <td>0.010709</td>\n",
       "      <td>-0.067238</td>\n",
       "      <td>0.010842</td>\n",
       "      <td>-0.065685</td>\n",
       "      <td>-0.032939</td>\n",
       "      <td>0.026445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>EGLN1</th>\n",
       "      <td>0.734180</td>\n",
       "      <td>0.196859</td>\n",
       "      <td>-0.257842</td>\n",
       "      <td>0.216777</td>\n",
       "      <td>0.225424</td>\n",
       "      <td>0.171290</td>\n",
       "      <td>-0.044748</td>\n",
       "      <td>0.174698</td>\n",
       "      <td>0.074217</td>\n",
       "      <td>0.167909</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.040446</td>\n",
       "      <td>0.013609</td>\n",
       "      <td>0.026512</td>\n",
       "      <td>-0.009803</td>\n",
       "      <td>-0.006016</td>\n",
       "      <td>-0.023270</td>\n",
       "      <td>0.021470</td>\n",
       "      <td>0.033686</td>\n",
       "      <td>-0.026712</td>\n",
       "      <td>-0.023275</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>GLO1</th>\n",
       "      <td>0.381934</td>\n",
       "      <td>0.292640</td>\n",
       "      <td>-0.390845</td>\n",
       "      <td>0.591383</td>\n",
       "      <td>0.285192</td>\n",
       "      <td>-0.080550</td>\n",
       "      <td>0.045263</td>\n",
       "      <td>0.013389</td>\n",
       "      <td>-0.109860</td>\n",
       "      <td>-0.087573</td>\n",
       "      <td>...</td>\n",
       "      <td>0.004563</td>\n",
       "      <td>-0.005082</td>\n",
       "      <td>-0.019318</td>\n",
       "      <td>-0.022084</td>\n",
       "      <td>0.006968</td>\n",
       "      <td>0.027341</td>\n",
       "      <td>-0.007367</td>\n",
       "      <td>-0.016814</td>\n",
       "      <td>0.017348</td>\n",
       "      <td>-0.018840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>PPP1R9B</th>\n",
       "      <td>0.385466</td>\n",
       "      <td>0.749914</td>\n",
       "      <td>0.283715</td>\n",
       "      <td>-0.232300</td>\n",
       "      <td>-0.138367</td>\n",
       "      <td>-0.053719</td>\n",
       "      <td>0.078377</td>\n",
       "      <td>-0.043379</td>\n",
       "      <td>0.053315</td>\n",
       "      <td>0.008124</td>\n",
       "      <td>...</td>\n",
       "      <td>0.004730</td>\n",
       "      <td>-0.020397</td>\n",
       "      <td>-0.024401</td>\n",
       "      <td>0.024605</td>\n",
       "      <td>-0.019153</td>\n",
       "      <td>0.008554</td>\n",
       "      <td>0.027525</td>\n",
       "      <td>0.017290</td>\n",
       "      <td>-0.000192</td>\n",
       "      <td>0.028478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TANK</th>\n",
       "      <td>0.362025</td>\n",
       "      <td>0.709495</td>\n",
       "      <td>0.150870</td>\n",
       "      <td>-0.030039</td>\n",
       "      <td>-0.001636</td>\n",
       "      <td>-0.064957</td>\n",
       "      <td>0.090942</td>\n",
       "      <td>0.061639</td>\n",
       "      <td>-0.027805</td>\n",
       "      <td>0.092919</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.114584</td>\n",
       "      <td>0.027340</td>\n",
       "      <td>0.150913</td>\n",
       "      <td>0.116223</td>\n",
       "      <td>0.020587</td>\n",
       "      <td>-0.060132</td>\n",
       "      <td>-0.083657</td>\n",
       "      <td>0.016547</td>\n",
       "      <td>-0.100436</td>\n",
       "      <td>0.122788</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ITGA11</th>\n",
       "      <td>-0.092685</td>\n",
       "      <td>-0.117027</td>\n",
       "      <td>0.391809</td>\n",
       "      <td>0.285527</td>\n",
       "      <td>-0.314618</td>\n",
       "      <td>0.171078</td>\n",
       "      <td>-0.037022</td>\n",
       "      <td>-0.231539</td>\n",
       "      <td>0.387119</td>\n",
       "      <td>0.082766</td>\n",
       "      <td>...</td>\n",
       "      <td>0.053248</td>\n",
       "      <td>-0.002439</td>\n",
       "      <td>0.146997</td>\n",
       "      <td>-0.031662</td>\n",
       "      <td>-0.003097</td>\n",
       "      <td>0.014115</td>\n",
       "      <td>0.106198</td>\n",
       "      <td>0.079781</td>\n",
       "      <td>0.032728</td>\n",
       "      <td>-0.037619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TPSAB1</th>\n",
       "      <td>0.460057</td>\n",
       "      <td>-0.293991</td>\n",
       "      <td>0.196820</td>\n",
       "      <td>0.048831</td>\n",
       "      <td>0.106954</td>\n",
       "      <td>-0.071006</td>\n",
       "      <td>-0.097098</td>\n",
       "      <td>0.053695</td>\n",
       "      <td>-0.103876</td>\n",
       "      <td>-0.042282</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.049576</td>\n",
       "      <td>-0.073240</td>\n",
       "      <td>0.096579</td>\n",
       "      <td>0.039670</td>\n",
       "      <td>0.174584</td>\n",
       "      <td>0.063012</td>\n",
       "      <td>-0.093843</td>\n",
       "      <td>0.012792</td>\n",
       "      <td>0.173550</td>\n",
       "      <td>0.057449</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>168 rows × 59 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8b4e003d-7372-40cf-8b08-9576fdd4d0e4')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-8b4e003d-7372-40cf-8b08-9576fdd4d0e4 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-8b4e003d-7372-40cf-8b08-9576fdd4d0e4');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                 0         1         2         3         4         5   \\\n",
       "TNFRSF10A  0.797710 -0.204731 -0.234727 -0.115882 -0.061635  0.013233   \n",
       "CXCL1      0.253141  0.138941  0.044449  0.143732 -0.339333 -0.106429   \n",
       "ITM2A      0.633502 -0.350919 -0.048621 -0.102820 -0.030706  0.148995   \n",
       "SIT1       0.130831  0.179770  0.253993  0.248916  0.209400  0.472950   \n",
       "EGLN1      0.734180  0.196859 -0.257842  0.216777  0.225424  0.171290   \n",
       "...             ...       ...       ...       ...       ...       ...   \n",
       "GLO1       0.381934  0.292640 -0.390845  0.591383  0.285192 -0.080550   \n",
       "PPP1R9B    0.385466  0.749914  0.283715 -0.232300 -0.138367 -0.053719   \n",
       "TANK       0.362025  0.709495  0.150870 -0.030039 -0.001636 -0.064957   \n",
       "ITGA11    -0.092685 -0.117027  0.391809  0.285527 -0.314618  0.171078   \n",
       "TPSAB1     0.460057 -0.293991  0.196820  0.048831  0.106954 -0.071006   \n",
       "\n",
       "                 6         7         8         9   ...        49        50  \\\n",
       "TNFRSF10A -0.056326 -0.008419  0.150010 -0.037909  ...  0.076022  0.036356   \n",
       "CXCL1      0.207265 -0.130248 -0.402272 -0.122429  ...  0.080746  0.040365   \n",
       "ITM2A      0.122867 -0.088981  0.088043  0.058009  ... -0.008549 -0.069752   \n",
       "SIT1      -0.289355  0.135301 -0.146441  0.347028  ...  0.042623 -0.037030   \n",
       "EGLN1     -0.044748  0.174698  0.074217  0.167909  ... -0.040446  0.013609   \n",
       "...             ...       ...       ...       ...  ...       ...       ...   \n",
       "GLO1       0.045263  0.013389 -0.109860 -0.087573  ...  0.004563 -0.005082   \n",
       "PPP1R9B    0.078377 -0.043379  0.053315  0.008124  ...  0.004730 -0.020397   \n",
       "TANK       0.090942  0.061639 -0.027805  0.092919  ... -0.114584  0.027340   \n",
       "ITGA11    -0.037022 -0.231539  0.387119  0.082766  ...  0.053248 -0.002439   \n",
       "TPSAB1    -0.097098  0.053695 -0.103876 -0.042282  ... -0.049576 -0.073240   \n",
       "\n",
       "                 51        52        53        54        55        56  \\\n",
       "TNFRSF10A -0.095167  0.042071  0.003061  0.037474 -0.011173  0.013604   \n",
       "CXCL1     -0.014334  0.061180 -0.087790 -0.108023 -0.008110  0.003904   \n",
       "ITM2A     -0.167011  0.122260  0.086425  0.142849  0.107072  0.001171   \n",
       "SIT1       0.006670  0.016537  0.010709 -0.067238  0.010842 -0.065685   \n",
       "EGLN1      0.026512 -0.009803 -0.006016 -0.023270  0.021470  0.033686   \n",
       "...             ...       ...       ...       ...       ...       ...   \n",
       "GLO1      -0.019318 -0.022084  0.006968  0.027341 -0.007367 -0.016814   \n",
       "PPP1R9B   -0.024401  0.024605 -0.019153  0.008554  0.027525  0.017290   \n",
       "TANK       0.150913  0.116223  0.020587 -0.060132 -0.083657  0.016547   \n",
       "ITGA11     0.146997 -0.031662 -0.003097  0.014115  0.106198  0.079781   \n",
       "TPSAB1     0.096579  0.039670  0.174584  0.063012 -0.093843  0.012792   \n",
       "\n",
       "                 57        58  \n",
       "TNFRSF10A -0.024180  0.051163  \n",
       "CXCL1     -0.057877  0.128499  \n",
       "ITM2A     -0.006032 -0.056704  \n",
       "SIT1      -0.032939  0.026445  \n",
       "EGLN1     -0.026712 -0.023275  \n",
       "...             ...       ...  \n",
       "GLO1       0.017348 -0.018840  \n",
       "PPP1R9B   -0.000192  0.028478  \n",
       "TANK      -0.100436  0.122788  \n",
       "ITGA11     0.032728 -0.037619  \n",
       "TPSAB1     0.173550  0.057449  \n",
       "\n",
       "[168 rows x 59 columns]"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loadings = pca.components_.T * np.sqrt(pca.explained_variance_)\n",
    "loading_matrix = pd.DataFrame(loadings, index=mgh_con.drop(['WHO max'], axis=1)[col].columns)\n",
    "loading_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "9EyeHOCBWwzf"
   },
   "outputs": [],
   "source": [
    "load_trans = loading_matrix.T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "    magnitude = load_trans[i].pow(2).sum()\n",
    "    loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowMGHtask2.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ao97g3c9-UQi"
   },
   "source": [
    "## ICL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 301
    },
    "executionInfo": {
     "elapsed": 20949,
     "status": "ok",
     "timestamp": 1649059050662,
     "user": {
      "displayName": "PV Clinical",
      "userId": "13901498519864370078"
     },
     "user_tz": -180
    },
    "id": "OgcKTpdSc5FP",
    "outputId": "ac7b7ae8-1cf3-499f-c817-5628d9480788"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-94129c1b-c32a-4e4a-b711-94daf4c50c9d\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>Models</th>\n",
       "      <th>Best parameters</th>\n",
       "      <th>Score</th>\n",
       "      <th>AUC</th>\n",
       "      <th>F1-score</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>([DecisionTreeRegressor(criterion='friedman_ms...</td>\n",
       "      <td>{'learning_rate': 0.1, 'n_estimators': 100}</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.68</td>\n",
       "      <td>0.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=20, max_feat...</td>\n",
       "      <td>{'class_weight': None, 'max_depth': 20, 'n_est...</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DecisionTreeClassifier(max_depth=5)</td>\n",
       "      <td>{'max_depth': 5, 'splitter': 'best'}</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.47</td>\n",
       "      <td>0.49</td>\n",
       "      <td>0.45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC(C=1)</td>\n",
       "      <td>{'C': 1, 'kernel': 'rbf'}</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=1, random_st...</td>\n",
       "      <td>{'learning_rate': 0.5, 'n_estimators': 20}</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.56</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KNeighborsClassifier(n_neighbors=10)</td>\n",
       "      <td>{'n_neighbors': 10}</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>(ExtraTreeClassifier(max_depth=20, random_stat...</td>\n",
       "      <td>{'class_weight': None, 'max_depth': 20, 'n_est...</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.79</td>\n",
       "      <td>0.36</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MLPClassifier(activation='tanh', solver='sgd')</td>\n",
       "      <td>{'activation': 'tanh', 'alpha': 0.0001, 'hidde...</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.61</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>XGBClassifier(max_depth=8, n_estimators=50, re...</td>\n",
       "      <td>{'gamma': 0, 'learning_rate': 0.1, 'max_depth'...</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.79</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.53</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-94129c1b-c32a-4e4a-b711-94daf4c50c9d')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-94129c1b-c32a-4e4a-b711-94daf4c50c9d button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-94129c1b-c32a-4e4a-b711-94daf4c50c9d');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                                              Models  \\\n",
       "0  ([DecisionTreeRegressor(criterion='friedman_ms...   \n",
       "1  (DecisionTreeClassifier(max_depth=20, max_feat...   \n",
       "2                DecisionTreeClassifier(max_depth=5)   \n",
       "3                                           SVC(C=1)   \n",
       "4  (DecisionTreeClassifier(max_depth=1, random_st...   \n",
       "5               KNeighborsClassifier(n_neighbors=10)   \n",
       "6  (ExtraTreeClassifier(max_depth=20, random_stat...   \n",
       "7     MLPClassifier(activation='tanh', solver='sgd')   \n",
       "8  XGBClassifier(max_depth=8, n_estimators=50, re...   \n",
       "\n",
       "                                     Best parameters  Score   AUC  F1-score  \\\n",
       "0        {'learning_rate': 0.1, 'n_estimators': 100}   0.76  0.77      0.55   \n",
       "1  {'class_weight': None, 'max_depth': 20, 'n_est...   0.80  0.77      0.40   \n",
       "2               {'max_depth': 5, 'splitter': 'best'}   0.67  0.55      0.47   \n",
       "3                          {'C': 1, 'kernel': 'rbf'}   0.78  0.81      0.61   \n",
       "4         {'learning_rate': 0.5, 'n_estimators': 20}   0.73  0.75      0.56   \n",
       "5                                {'n_neighbors': 10}   0.84  0.78      0.48   \n",
       "6  {'class_weight': None, 'max_depth': 20, 'n_est...   0.71  0.79      0.36   \n",
       "7  {'activation': 'tanh', 'alpha': 0.0001, 'hidde...   0.75  0.80      0.62   \n",
       "8  {'gamma': 0, 'learning_rate': 0.1, 'max_depth'...   0.71  0.79      0.61   \n",
       "\n",
       "   Precision  Recall  \n",
       "0       0.68    0.47  \n",
       "1       0.69    0.28  \n",
       "2       0.49    0.45  \n",
       "3       0.67    0.58  \n",
       "4       0.66    0.50  \n",
       "5       0.77    0.35  \n",
       "6       0.73    0.24  \n",
       "7       0.63    0.61  \n",
       "8       0.74    0.53  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_mX_pca, grid_ml_pca = Algo_search(models = models, params = params, X_train=X, y_train=y, X_test=X_test1, y_test=y_test1)\n",
    "\n",
    "grid_ml_pca"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6WZ4XeQS0YVM"
   },
   "source": [
    "### Explainable AI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 66,
     "referenced_widgets": [
      "a961064cabca48edb4f50d91ff1b386d",
      "9348543a8cd84980809203081c777509",
      "64483751c11e4e4fbde1e1a3e6391eae",
      "0ed49d6fc73f4ee7930604ded0487c23",
      "0579c1ffe3734c3e87b041dfc89a5048",
      "a92d7ad7d07949a09e6de9057cb27e4f",
      "15cba8dead5b45bdbfc3ef49b9ef1d2e",
      "f757b7edd0df4cbfb3513fe595663e08",
      "bc95e906d0a749479b881f0747e5ad23",
      "7ad4773718da4fd8a8bc165497d181de",
      "e9c1c09b268a4fe18a535c14d36afa85"
     ]
    },
    "executionInfo": {
     "elapsed": 399623,
     "status": "ok",
     "timestamp": 1651840557599,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "B7mHatePw-jP",
    "outputId": "0b3281e3-29bf-4eea-bc0d-ee8dfe298684"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using 305 background data samples could cause slower run times. Consider using shap.sample(data, K) or shap.kmeans(data, K) to summarize the background as K samples.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a961064cabca48edb4f50d91ff1b386d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/55 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = MLPClassifier(activation= 'tanh',\n",
    "hidden_layer_sizes= (100,),\n",
    "alpha= 0.0001,\n",
    "learning_rate= 'constant',\n",
    "solver= 'sgd'\n",
    ")\n",
    "\n",
    "clf.fit(X, y)\n",
    "\n",
    "explainer = shap.KernelExplainer(clf.predict,X)\n",
    "shap_values = explainer.shap_values(X_test1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 585
    },
    "executionInfo": {
     "elapsed": 814,
     "status": "ok",
     "timestamp": 1651840558383,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "KJ3ouaAnxEL3",
    "outputId": "c387d4dc-b90b-4fc5-b8b4-5d22af173608"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "shap.summary_plot(shap_values,X_test1,feature_names=X.columns, max_display =20,show=False )\n",
    "plt.savefig('mlp_xai_mgh_icl.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "Sr0r5Rxt61OR"
   },
   "outputs": [],
   "source": [
    "vals= np.abs(shap_values).mean(0)\n",
    "feature_importance = pd.DataFrame(list(zip(X.columns,vals)),columns=['col_name','feature_importance_vals'])\n",
    "feature_importance.sort_values(by=['feature_importance_vals'],ascending=False,inplace=True)\n",
    "feature_importance.to_excel('mgh_mlp_icl_xai.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "KLD5TsAv7gwD"
   },
   "outputs": [],
   "source": [
    "lp =  [i for i in list(feature_importance['col_name'])[:21] if str(i).isdigit()]\n",
    "\n",
    "load_trans = loading_matrix[lp].T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowMGHtoICLmpl.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Z_FNpxpG-d-Y"
   },
   "source": [
    "## YPS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 301
    },
    "executionInfo": {
     "elapsed": 36637,
     "status": "ok",
     "timestamp": 1649059688565,
     "user": {
      "displayName": "PV Clinical",
      "userId": "13901498519864370078"
     },
     "user_tz": -180
    },
    "id": "D4dBwMA7mwgz",
    "outputId": "afec3d84-a930-4aad-e152-63c64f62f024"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-fe6df6e5-060c-4b9b-b48d-12e98bb139cb\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>Models</th>\n",
       "      <th>Best parameters</th>\n",
       "      <th>Score</th>\n",
       "      <th>AUC</th>\n",
       "      <th>F1-score</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>([DecisionTreeRegressor(criterion='friedman_ms...</td>\n",
       "      <td>{'learning_rate': 0.4, 'n_estimators': 100}</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=20, max_feat...</td>\n",
       "      <td>{'class_weight': None, 'max_depth': 20, 'n_est...</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.42</td>\n",
       "      <td>0.72</td>\n",
       "      <td>0.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DecisionTreeClassifier(max_depth=1, splitter='...</td>\n",
       "      <td>{'max_depth': 1, 'splitter': 'random'}</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.58</td>\n",
       "      <td>0.32</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC(C=1)</td>\n",
       "      <td>{'C': 1, 'kernel': 'rbf'}</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=1, random_st...</td>\n",
       "      <td>{'learning_rate': 0.5, 'n_estimators': 20}</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.56</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KNeighborsClassifier(n_neighbors=10)</td>\n",
       "      <td>{'n_neighbors': 10}</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.77</td>\n",
       "      <td>0.35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>(ExtraTreeClassifier(max_depth=20, random_stat...</td>\n",
       "      <td>{'class_weight': None, 'max_depth': 20, 'n_est...</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MLPClassifier(activation='tanh', hidden_layer_...</td>\n",
       "      <td>{'activation': 'tanh', 'alpha': 0.0001, 'hidde...</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.60</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>XGBClassifier(max_depth=8, n_estimators=50, re...</td>\n",
       "      <td>{'gamma': 0, 'learning_rate': 0.1, 'max_depth'...</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.79</td>\n",
       "      <td>0.61</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.53</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-fe6df6e5-060c-4b9b-b48d-12e98bb139cb')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-fe6df6e5-060c-4b9b-b48d-12e98bb139cb button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-fe6df6e5-060c-4b9b-b48d-12e98bb139cb');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                                              Models  \\\n",
       "0  ([DecisionTreeRegressor(criterion='friedman_ms...   \n",
       "1  (DecisionTreeClassifier(max_depth=20, max_feat...   \n",
       "2  DecisionTreeClassifier(max_depth=1, splitter='...   \n",
       "3                                           SVC(C=1)   \n",
       "4  (DecisionTreeClassifier(max_depth=1, random_st...   \n",
       "5               KNeighborsClassifier(n_neighbors=10)   \n",
       "6  (ExtraTreeClassifier(max_depth=20, random_stat...   \n",
       "7  MLPClassifier(activation='tanh', hidden_layer_...   \n",
       "8  XGBClassifier(max_depth=8, n_estimators=50, re...   \n",
       "\n",
       "                                     Best parameters  Score   AUC  F1-score  \\\n",
       "0        {'learning_rate': 0.4, 'n_estimators': 100}   0.84  0.77      0.55   \n",
       "1  {'class_weight': None, 'max_depth': 20, 'n_est...   0.85  0.77      0.42   \n",
       "2             {'max_depth': 1, 'splitter': 'random'}   0.84  0.58      0.32   \n",
       "3                          {'C': 1, 'kernel': 'rbf'}   0.85  0.81      0.61   \n",
       "4         {'learning_rate': 0.5, 'n_estimators': 20}   0.84  0.75      0.56   \n",
       "5                                {'n_neighbors': 10}   0.86  0.78      0.48   \n",
       "6  {'class_weight': None, 'max_depth': 20, 'n_est...   0.83  0.80      0.40   \n",
       "7  {'activation': 'tanh', 'alpha': 0.0001, 'hidde...   0.82  0.81      0.62   \n",
       "8  {'gamma': 0, 'learning_rate': 0.1, 'max_depth'...   0.86  0.79      0.61   \n",
       "\n",
       "   Precision  Recall  \n",
       "0       0.67    0.48  \n",
       "1       0.72    0.30  \n",
       "2       0.63    0.23  \n",
       "3       0.67    0.58  \n",
       "4       0.66    0.50  \n",
       "5       0.77    0.35  \n",
       "6       0.76    0.28  \n",
       "7       0.65    0.60  \n",
       "8       0.74    0.53  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_mX_pca, grid_ml_pca_yps = Algo_search(models = models, params = params, X_train=X, y_train=y, X_test=X_test2, y_test=y_test2)\n",
    "\n",
    "grid_ml_pca_yps"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "_RorZjskWgDR"
   },
   "source": [
    "### Explainable AI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 66,
     "referenced_widgets": [
      "6e13b236fb9f4b3aad08893bf5812e23",
      "10ebfb73b7494bed9a698b5024b2636e",
      "4d5bf205a7694ef081ecc7a56305307d",
      "2aca8bd289c34a5c8b49bd64cfc9f2af",
      "7f3109affa534cffa948126c54ba3f0d",
      "6585c84dd0c140c6868ac60efa01a461",
      "e7df0eeb8b364496b832d1bb5b89f854",
      "663105b452d54ed983932a81b885adb2",
      "30b057d3ed124a6fbc1e5922c74deb55",
      "e1fed3d156a14076b5c1a47f75be79df",
      "3c70c8e3cd49478dbf3988e53054edc1"
     ]
    },
    "executionInfo": {
     "elapsed": 1968574,
     "status": "ok",
     "timestamp": 1651843877982,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "jghsbHvWWgDR",
    "outputId": "afd18082-b7fc-47c7-a6eb-433566a56770"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using 305 background data samples could cause slower run times. Consider using shap.sample(data, K) or shap.kmeans(data, K) to summarize the background as K samples.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e13b236fb9f4b3aad08893bf5812e23",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/256 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = MLPClassifier(activation= 'tanh',\n",
    "hidden_layer_sizes= (50, 50, 50),\n",
    "alpha= 0.0001,\n",
    "learning_rate= 'adaptive',\n",
    "solver= 'sgd'\n",
    ")\n",
    "\n",
    "clf.fit(X, y)\n",
    "\n",
    "explainer = shap.KernelExplainer(clf.predict,X)\n",
    "shap_values = explainer.shap_values(X_test2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "executionInfo": {
     "elapsed": 1318,
     "status": "ok",
     "timestamp": 1651843879277,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "bnP_olNvWgDS",
    "outputId": "a9b7b7cb-a92f-48d5-e072-a4602d4d9e8d"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "shap.summary_plot(shap_values,X_test2,feature_names=X.columns, max_display =20,show=False )\n",
    "plt.savefig('mlp_xai_mgh_yps.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "9_N2933QWgDT"
   },
   "outputs": [],
   "source": [
    "vals= np.abs(shap_values).mean(0)\n",
    "feature_importance = pd.DataFrame(list(zip(X.columns,vals)),columns=['col_name','feature_importance_vals'])\n",
    "feature_importance.sort_values(by=['feature_importance_vals'],ascending=False,inplace=True)\n",
    "feature_importance.to_excel('mgh_mlp_yps_xai.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "2GYi4qBtWgDT"
   },
   "outputs": [],
   "source": [
    "lp =  [i for i in list(feature_importance['col_name'])[:21] if str(i).isdigit()]\n",
    "\n",
    "load_trans = loading_matrix[lp].T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowMGHtoYPSmpl.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "i0sQPBtE_Kuz"
   },
   "source": [
    "# YPS dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "oFxmrkEE_iEM"
   },
   "outputs": [],
   "source": [
    "X = yps_con.drop(['WHO max'], axis=1)\n",
    "y = yps_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "CgPSYly2_iEL"
   },
   "outputs": [],
   "source": [
    "X_test1 = icl_con.drop(['WHO max'], axis=1)\n",
    "y_test1 = icl_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "9PT0pQyo_iEA"
   },
   "outputs": [],
   "source": [
    "X_test2 = mgh_con.drop(['WHO max'], axis=1)\n",
    "y_test2 = mgh_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "BvK4xrf5ouFw"
   },
   "outputs": [],
   "source": [
    "pca = PCA(.90)\n",
    " \n",
    "X = pca.fit_transform(X[col])\n",
    "X = pd.DataFrame(X)\n",
    "X['KIDNEY'] = list(yps_con['KIDNEY'])\n",
    "X['Age cat'] = list(yps_con['Age cat'])\n",
    "\n",
    "\n",
    "X_test1 = pca.transform(X_test1[col])\n",
    "X_test1 = pd.DataFrame(X_test1)\n",
    "X_test1['KIDNEY'] = list(icl_con['KIDNEY'])\n",
    "X_test1['Age cat'] = list(icl_con['Age cat'])\n",
    "\n",
    "\n",
    "X_test2 = pca.transform(X_test2[col])\n",
    "X_test2 = pd.DataFrame(X_test2)\n",
    "X_test2['KIDNEY'] = list(mgh_con['KIDNEY'])\n",
    "X_test2['Age cat'] = list(mgh_con['Age cat'])\n",
    " \n",
    "explained_variance = pca.explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 488
    },
    "executionInfo": {
     "elapsed": 322,
     "status": "ok",
     "timestamp": 1652090065699,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "uCb7ja7UeDPA",
    "outputId": "e295d495-13a2-41e7-872b-607b86f09ab3"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-808ac278-9cba-4aa4-9c37-42cfe48e30c9\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>36</th>\n",
       "      <th>37</th>\n",
       "      <th>38</th>\n",
       "      <th>39</th>\n",
       "      <th>40</th>\n",
       "      <th>41</th>\n",
       "      <th>42</th>\n",
       "      <th>43</th>\n",
       "      <th>44</th>\n",
       "      <th>45</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PADI2</th>\n",
       "      <td>0.185265</td>\n",
       "      <td>0.290259</td>\n",
       "      <td>0.096207</td>\n",
       "      <td>0.177796</td>\n",
       "      <td>0.060971</td>\n",
       "      <td>-0.132316</td>\n",
       "      <td>0.162581</td>\n",
       "      <td>0.032703</td>\n",
       "      <td>0.159832</td>\n",
       "      <td>0.148374</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.015878</td>\n",
       "      <td>0.103218</td>\n",
       "      <td>0.014140</td>\n",
       "      <td>-0.085143</td>\n",
       "      <td>0.013206</td>\n",
       "      <td>0.168408</td>\n",
       "      <td>-0.037142</td>\n",
       "      <td>-0.069784</td>\n",
       "      <td>-0.008192</td>\n",
       "      <td>0.094283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TNF</th>\n",
       "      <td>0.664572</td>\n",
       "      <td>0.566817</td>\n",
       "      <td>0.099295</td>\n",
       "      <td>-0.058488</td>\n",
       "      <td>-0.033647</td>\n",
       "      <td>0.138605</td>\n",
       "      <td>-0.198789</td>\n",
       "      <td>0.037580</td>\n",
       "      <td>0.061956</td>\n",
       "      <td>-0.053142</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.038470</td>\n",
       "      <td>-0.053540</td>\n",
       "      <td>-0.084142</td>\n",
       "      <td>-0.030940</td>\n",
       "      <td>0.056747</td>\n",
       "      <td>0.003489</td>\n",
       "      <td>-0.059054</td>\n",
       "      <td>-0.026892</td>\n",
       "      <td>-0.016471</td>\n",
       "      <td>-0.030832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NTF4</th>\n",
       "      <td>-0.299257</td>\n",
       "      <td>0.036657</td>\n",
       "      <td>-0.117001</td>\n",
       "      <td>0.276262</td>\n",
       "      <td>-0.197796</td>\n",
       "      <td>0.085638</td>\n",
       "      <td>-0.099776</td>\n",
       "      <td>-0.002739</td>\n",
       "      <td>0.269674</td>\n",
       "      <td>0.022818</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.017002</td>\n",
       "      <td>-0.062447</td>\n",
       "      <td>0.073164</td>\n",
       "      <td>0.107295</td>\n",
       "      <td>0.060386</td>\n",
       "      <td>-0.139602</td>\n",
       "      <td>0.051448</td>\n",
       "      <td>0.068168</td>\n",
       "      <td>-0.008835</td>\n",
       "      <td>-0.075251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SH2D1A</th>\n",
       "      <td>-0.500462</td>\n",
       "      <td>0.256107</td>\n",
       "      <td>-0.039449</td>\n",
       "      <td>0.264996</td>\n",
       "      <td>-0.166362</td>\n",
       "      <td>0.077634</td>\n",
       "      <td>0.241581</td>\n",
       "      <td>-0.045299</td>\n",
       "      <td>-0.327944</td>\n",
       "      <td>0.130274</td>\n",
       "      <td>...</td>\n",
       "      <td>0.082902</td>\n",
       "      <td>0.045541</td>\n",
       "      <td>-0.032675</td>\n",
       "      <td>-0.017191</td>\n",
       "      <td>0.099906</td>\n",
       "      <td>-0.034097</td>\n",
       "      <td>-0.020038</td>\n",
       "      <td>-0.060043</td>\n",
       "      <td>0.023619</td>\n",
       "      <td>0.065753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TNFRSF9</th>\n",
       "      <td>0.388303</td>\n",
       "      <td>0.462019</td>\n",
       "      <td>-0.476707</td>\n",
       "      <td>0.323234</td>\n",
       "      <td>-0.076156</td>\n",
       "      <td>0.087476</td>\n",
       "      <td>-0.267976</td>\n",
       "      <td>-0.046952</td>\n",
       "      <td>-0.067336</td>\n",
       "      <td>-0.197087</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.032266</td>\n",
       "      <td>0.015113</td>\n",
       "      <td>-0.005006</td>\n",
       "      <td>-0.026084</td>\n",
       "      <td>0.033368</td>\n",
       "      <td>0.094410</td>\n",
       "      <td>0.025215</td>\n",
       "      <td>-0.040964</td>\n",
       "      <td>-0.039637</td>\n",
       "      <td>0.090872</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>PGF</th>\n",
       "      <td>0.407320</td>\n",
       "      <td>0.632454</td>\n",
       "      <td>-0.130870</td>\n",
       "      <td>0.157479</td>\n",
       "      <td>0.155316</td>\n",
       "      <td>-0.198072</td>\n",
       "      <td>-0.012667</td>\n",
       "      <td>-0.026586</td>\n",
       "      <td>0.050156</td>\n",
       "      <td>-0.288868</td>\n",
       "      <td>...</td>\n",
       "      <td>0.026915</td>\n",
       "      <td>-0.041211</td>\n",
       "      <td>0.116871</td>\n",
       "      <td>-0.029310</td>\n",
       "      <td>0.048338</td>\n",
       "      <td>-0.032054</td>\n",
       "      <td>-0.044420</td>\n",
       "      <td>0.097505</td>\n",
       "      <td>-0.055912</td>\n",
       "      <td>-0.012641</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FCRL3</th>\n",
       "      <td>0.021800</td>\n",
       "      <td>0.212162</td>\n",
       "      <td>-0.368506</td>\n",
       "      <td>0.232978</td>\n",
       "      <td>-0.245505</td>\n",
       "      <td>0.294919</td>\n",
       "      <td>-0.039214</td>\n",
       "      <td>-0.037851</td>\n",
       "      <td>-0.062067</td>\n",
       "      <td>0.026484</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.045252</td>\n",
       "      <td>0.103538</td>\n",
       "      <td>0.029314</td>\n",
       "      <td>0.094755</td>\n",
       "      <td>-0.120752</td>\n",
       "      <td>0.034494</td>\n",
       "      <td>0.010170</td>\n",
       "      <td>-0.042781</td>\n",
       "      <td>-0.065560</td>\n",
       "      <td>-0.045369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VEGFA</th>\n",
       "      <td>0.127509</td>\n",
       "      <td>0.665666</td>\n",
       "      <td>0.264007</td>\n",
       "      <td>0.101151</td>\n",
       "      <td>0.036236</td>\n",
       "      <td>-0.143981</td>\n",
       "      <td>-0.197576</td>\n",
       "      <td>0.082658</td>\n",
       "      <td>0.043985</td>\n",
       "      <td>-0.031689</td>\n",
       "      <td>...</td>\n",
       "      <td>0.023745</td>\n",
       "      <td>0.069864</td>\n",
       "      <td>0.051979</td>\n",
       "      <td>-0.094880</td>\n",
       "      <td>0.060571</td>\n",
       "      <td>0.078159</td>\n",
       "      <td>-0.117521</td>\n",
       "      <td>0.164456</td>\n",
       "      <td>-0.018921</td>\n",
       "      <td>-0.002658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KRT19</th>\n",
       "      <td>0.601978</td>\n",
       "      <td>0.448903</td>\n",
       "      <td>0.444960</td>\n",
       "      <td>-0.077246</td>\n",
       "      <td>0.009860</td>\n",
       "      <td>-0.202920</td>\n",
       "      <td>0.128936</td>\n",
       "      <td>0.037780</td>\n",
       "      <td>-0.035462</td>\n",
       "      <td>0.019507</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.025515</td>\n",
       "      <td>-0.003225</td>\n",
       "      <td>0.015843</td>\n",
       "      <td>0.012196</td>\n",
       "      <td>-0.046100</td>\n",
       "      <td>-0.089813</td>\n",
       "      <td>0.020325</td>\n",
       "      <td>0.001998</td>\n",
       "      <td>-0.010889</td>\n",
       "      <td>0.119265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NRTN</th>\n",
       "      <td>-0.083886</td>\n",
       "      <td>0.455005</td>\n",
       "      <td>0.013628</td>\n",
       "      <td>-0.017429</td>\n",
       "      <td>0.047384</td>\n",
       "      <td>0.059611</td>\n",
       "      <td>0.098507</td>\n",
       "      <td>0.077692</td>\n",
       "      <td>0.185100</td>\n",
       "      <td>-0.129237</td>\n",
       "      <td>...</td>\n",
       "      <td>0.009347</td>\n",
       "      <td>-0.045934</td>\n",
       "      <td>-0.176941</td>\n",
       "      <td>0.050089</td>\n",
       "      <td>-0.066265</td>\n",
       "      <td>0.123045</td>\n",
       "      <td>-0.038087</td>\n",
       "      <td>0.146057</td>\n",
       "      <td>0.067918</td>\n",
       "      <td>0.042168</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>168 rows × 46 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-808ac278-9cba-4aa4-9c37-42cfe48e30c9')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-808ac278-9cba-4aa4-9c37-42cfe48e30c9 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-808ac278-9cba-4aa4-9c37-42cfe48e30c9');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "               0         1         2         3         4         5         6   \\\n",
       "PADI2    0.185265  0.290259  0.096207  0.177796  0.060971 -0.132316  0.162581   \n",
       "TNF      0.664572  0.566817  0.099295 -0.058488 -0.033647  0.138605 -0.198789   \n",
       "NTF4    -0.299257  0.036657 -0.117001  0.276262 -0.197796  0.085638 -0.099776   \n",
       "SH2D1A  -0.500462  0.256107 -0.039449  0.264996 -0.166362  0.077634  0.241581   \n",
       "TNFRSF9  0.388303  0.462019 -0.476707  0.323234 -0.076156  0.087476 -0.267976   \n",
       "...           ...       ...       ...       ...       ...       ...       ...   \n",
       "PGF      0.407320  0.632454 -0.130870  0.157479  0.155316 -0.198072 -0.012667   \n",
       "FCRL3    0.021800  0.212162 -0.368506  0.232978 -0.245505  0.294919 -0.039214   \n",
       "VEGFA    0.127509  0.665666  0.264007  0.101151  0.036236 -0.143981 -0.197576   \n",
       "KRT19    0.601978  0.448903  0.444960 -0.077246  0.009860 -0.202920  0.128936   \n",
       "NRTN    -0.083886  0.455005  0.013628 -0.017429  0.047384  0.059611  0.098507   \n",
       "\n",
       "               7         8         9   ...        36        37        38  \\\n",
       "PADI2    0.032703  0.159832  0.148374  ... -0.015878  0.103218  0.014140   \n",
       "TNF      0.037580  0.061956 -0.053142  ... -0.038470 -0.053540 -0.084142   \n",
       "NTF4    -0.002739  0.269674  0.022818  ... -0.017002 -0.062447  0.073164   \n",
       "SH2D1A  -0.045299 -0.327944  0.130274  ...  0.082902  0.045541 -0.032675   \n",
       "TNFRSF9 -0.046952 -0.067336 -0.197087  ... -0.032266  0.015113 -0.005006   \n",
       "...           ...       ...       ...  ...       ...       ...       ...   \n",
       "PGF     -0.026586  0.050156 -0.288868  ...  0.026915 -0.041211  0.116871   \n",
       "FCRL3   -0.037851 -0.062067  0.026484  ... -0.045252  0.103538  0.029314   \n",
       "VEGFA    0.082658  0.043985 -0.031689  ...  0.023745  0.069864  0.051979   \n",
       "KRT19    0.037780 -0.035462  0.019507  ... -0.025515 -0.003225  0.015843   \n",
       "NRTN     0.077692  0.185100 -0.129237  ...  0.009347 -0.045934 -0.176941   \n",
       "\n",
       "               39        40        41        42        43        44        45  \n",
       "PADI2   -0.085143  0.013206  0.168408 -0.037142 -0.069784 -0.008192  0.094283  \n",
       "TNF     -0.030940  0.056747  0.003489 -0.059054 -0.026892 -0.016471 -0.030832  \n",
       "NTF4     0.107295  0.060386 -0.139602  0.051448  0.068168 -0.008835 -0.075251  \n",
       "SH2D1A  -0.017191  0.099906 -0.034097 -0.020038 -0.060043  0.023619  0.065753  \n",
       "TNFRSF9 -0.026084  0.033368  0.094410  0.025215 -0.040964 -0.039637  0.090872  \n",
       "...           ...       ...       ...       ...       ...       ...       ...  \n",
       "PGF     -0.029310  0.048338 -0.032054 -0.044420  0.097505 -0.055912 -0.012641  \n",
       "FCRL3    0.094755 -0.120752  0.034494  0.010170 -0.042781 -0.065560 -0.045369  \n",
       "VEGFA   -0.094880  0.060571  0.078159 -0.117521  0.164456 -0.018921 -0.002658  \n",
       "KRT19    0.012196 -0.046100 -0.089813  0.020325  0.001998 -0.010889  0.119265  \n",
       "NRTN     0.050089 -0.066265  0.123045 -0.038087  0.146057  0.067918  0.042168  \n",
       "\n",
       "[168 rows x 46 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loadings = pca.components_.T * np.sqrt(pca.explained_variance_)\n",
    "loading_matrix = pd.DataFrame(loadings, index=yps_con.drop(['WHO max'], axis=1)[col].columns)\n",
    "loading_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "DJ3revw2eDPB"
   },
   "outputs": [],
   "source": [
    "load_trans = loading_matrix.T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowYPStask2.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eBjXOiuc_iEN"
   },
   "source": [
    "## ICL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 301
    },
    "executionInfo": {
     "elapsed": 22951,
     "status": "ok",
     "timestamp": 1649060609285,
     "user": {
      "displayName": "PV Clinical",
      "userId": "13901498519864370078"
     },
     "user_tz": -180
    },
    "id": "VBaV3uBOo5az",
    "outputId": "e7cc200f-592e-4614-89fd-b2763349ebf4"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-8d6b2f36-fa53-4814-a290-304a65c67ece\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>Models</th>\n",
       "      <th>Best parameters</th>\n",
       "      <th>Score</th>\n",
       "      <th>AUC</th>\n",
       "      <th>F1-score</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>([DecisionTreeRegressor(criterion='friedman_ms...</td>\n",
       "      <td>{'learning_rate': 0.2, 'n_estimators': 50}</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.52</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=20, max_feat...</td>\n",
       "      <td>{'class_weight': None, 'max_depth': 20, 'n_est...</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.84</td>\n",
       "      <td>0.22</td>\n",
       "      <td>0.55</td>\n",
       "      <td>0.14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DecisionTreeClassifier(max_depth=5)</td>\n",
       "      <td>{'max_depth': 5, 'splitter': 'best'}</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.51</td>\n",
       "      <td>0.62</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC(C=5)</td>\n",
       "      <td>{'C': 5, 'kernel': 'rbf'}</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.94</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=1, random_st...</td>\n",
       "      <td>{'learning_rate': 0.5, 'n_estimators': 50}</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.47</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KNeighborsClassifier(n_neighbors=25)</td>\n",
       "      <td>{'n_neighbors': 25}</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>(ExtraTreeClassifier(max_depth=10, random_stat...</td>\n",
       "      <td>{'class_weight': {0: 1, 1: 10}, 'max_depth': 1...</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.19</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MLPClassifier(solver='sgd')</td>\n",
       "      <td>{'activation': 'relu', 'alpha': 0.0001, 'hidde...</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.94</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>XGBClassifier(gamma=0.1, learning_rate=0.7, ma...</td>\n",
       "      <td>{'gamma': 0.1, 'learning_rate': 0.7, 'max_dept...</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.57</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-8d6b2f36-fa53-4814-a290-304a65c67ece')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-8d6b2f36-fa53-4814-a290-304a65c67ece button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-8d6b2f36-fa53-4814-a290-304a65c67ece');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                                              Models  \\\n",
       "0  ([DecisionTreeRegressor(criterion='friedman_ms...   \n",
       "1  (DecisionTreeClassifier(max_depth=20, max_feat...   \n",
       "2                DecisionTreeClassifier(max_depth=5)   \n",
       "3                                           SVC(C=5)   \n",
       "4  (DecisionTreeClassifier(max_depth=1, random_st...   \n",
       "5               KNeighborsClassifier(n_neighbors=25)   \n",
       "6  (ExtraTreeClassifier(max_depth=10, random_stat...   \n",
       "7                        MLPClassifier(solver='sgd')   \n",
       "8  XGBClassifier(gamma=0.1, learning_rate=0.7, ma...   \n",
       "\n",
       "                                     Best parameters  Score   AUC  F1-score  \\\n",
       "0         {'learning_rate': 0.2, 'n_estimators': 50}   0.76  0.84      0.52   \n",
       "1  {'class_weight': None, 'max_depth': 20, 'n_est...   0.76  0.84      0.22   \n",
       "2               {'max_depth': 5, 'splitter': 'best'}   0.69  0.71      0.53   \n",
       "3                          {'C': 5, 'kernel': 'rbf'}   0.78  0.94      0.67   \n",
       "4         {'learning_rate': 0.5, 'n_estimators': 50}   0.78  0.83      0.47   \n",
       "5                                {'n_neighbors': 25}   0.78  0.89      0.33   \n",
       "6  {'class_weight': {0: 1, 1: 10}, 'max_depth': 1...   0.75  0.82      0.19   \n",
       "7  {'activation': 'relu', 'alpha': 0.0001, 'hidde...   0.82  0.94      0.65   \n",
       "8  {'gamma': 0.1, 'learning_rate': 0.7, 'max_dept...   0.80  0.86      0.57   \n",
       "\n",
       "   Precision  Recall  \n",
       "0       0.53    0.54  \n",
       "1       0.55    0.14  \n",
       "2       0.51    0.62  \n",
       "3       0.78    0.68  \n",
       "4       0.44    0.50  \n",
       "5       0.75    0.28  \n",
       "6       0.50    0.12  \n",
       "7       0.69    0.68  \n",
       "8       0.76    0.50  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_mX_pca, grid_ml_pca = Algo_search(models = models, params = params, X_train=X, y_train=y, X_test=X_test1, y_test=y_test1)\n",
    "\n",
    "grid_ml_pca"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "te6I8UWydcTP"
   },
   "source": [
    "### Explainable AI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 66,
     "referenced_widgets": [
      "6626a85c8dac4734b581824f25fc3d1b",
      "c3840367809d463d9c694ef2a92e0ff5",
      "69045c7a03864e48b20282fababbfe1f",
      "5de3e8a13918447fb856d796bc3e4305",
      "6bc5e43ffbf249658755d4844dec9525",
      "57a3b742ef2d47039cd217b8a4557bb4",
      "2b8a7b88de8243758573144ed63aadd2",
      "1e29de520a3a4c3dad23aee7a8169ac2",
      "a170d8c8cf7f49e8aa86520bc406fbed",
      "1b90209079f749a8bef476882a854e27",
      "f9d7d05478bf4cee998fe3c38a0069c6"
     ]
    },
    "executionInfo": {
     "elapsed": 195572,
     "status": "ok",
     "timestamp": 1652090315648,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "OdDSRaN-dcTP",
    "outputId": "8d623a51-032d-4088-8cf1-1e3d63563730"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using 256 background data samples could cause slower run times. Consider using shap.sample(data, K) or shap.kmeans(data, K) to summarize the background as K samples.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6626a85c8dac4734b581824f25fc3d1b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/55 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = MLPClassifier(activation= 'relu',\n",
    "hidden_layer_sizes= (100,),\n",
    "alpha= 0.0001,\n",
    "learning_rate= 'constant',\n",
    "solver= 'sgd'\n",
    ")\n",
    "\n",
    "clf.fit(X, y)\n",
    "\n",
    "explainer = shap.KernelExplainer(clf.predict,X)\n",
    "shap_values = explainer.shap_values(X_test1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 585
    },
    "executionInfo": {
     "elapsed": 1083,
     "status": "ok",
     "timestamp": 1652090316707,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "qbN8QAQXdcTQ",
    "outputId": "977e4ce6-7e80-4fd5-87c5-ca0c8df230ef"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "shap.summary_plot(shap_values,X_test1,feature_names=X.columns, max_display =20,show=False )\n",
    "plt.savefig('mlp_xai_yps_icl.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "32a8Sg7_dcTR"
   },
   "outputs": [],
   "source": [
    "vals= np.abs(shap_values).mean(0)\n",
    "feature_importance = pd.DataFrame(list(zip(X.columns,vals)),columns=['col_name','feature_importance_vals'])\n",
    "feature_importance.sort_values(by=['feature_importance_vals'],ascending=False,inplace=True)\n",
    "feature_importance.to_excel('mgh_yps_icl_xai.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "BKfYWRcqdcTR"
   },
   "outputs": [],
   "source": [
    "lp =  [i for i in list(feature_importance['col_name'])[:21] if str(i).isdigit()]\n",
    "\n",
    "load_trans = loading_matrix[lp].T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowYPStoICLmpl.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XMfQaOja_iEQ"
   },
   "source": [
    "## MGH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 301
    },
    "executionInfo": {
     "elapsed": 23384,
     "status": "ok",
     "timestamp": 1649061295058,
     "user": {
      "displayName": "PV Clinical",
      "userId": "13901498519864370078"
     },
     "user_tz": -180
    },
    "id": "-v66VZiepBRM",
    "outputId": "42355ee7-4b43-4b03-d668-de4500b1eb28"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-e895219b-2130-4730-8344-5fa55394f5b6\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>Models</th>\n",
       "      <th>Best parameters</th>\n",
       "      <th>Score</th>\n",
       "      <th>AUC</th>\n",
       "      <th>F1-score</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>([DecisionTreeRegressor(criterion='friedman_ms...</td>\n",
       "      <td>{'learning_rate': 0.2, 'n_estimators': 50}</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.85</td>\n",
       "      <td>0.51</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=20, max_feat...</td>\n",
       "      <td>{'class_weight': None, 'max_depth': 20, 'n_est...</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DecisionTreeClassifier(max_depth=1, splitter='...</td>\n",
       "      <td>{'max_depth': 1, 'splitter': 'random'}</td>\n",
       "      <td>0.64</td>\n",
       "      <td>0.54</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC(C=5)</td>\n",
       "      <td>{'C': 5, 'kernel': 'rbf'}</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.94</td>\n",
       "      <td>0.67</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=1, random_st...</td>\n",
       "      <td>{'learning_rate': 0.5, 'n_estimators': 50}</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.47</td>\n",
       "      <td>0.44</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KNeighborsClassifier(n_neighbors=25)</td>\n",
       "      <td>{'n_neighbors': 25}</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.89</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.75</td>\n",
       "      <td>0.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>(ExtraTreeClassifier(max_depth=5, random_state...</td>\n",
       "      <td>{'class_weight': {0: 1, 1: 5}, 'max_depth': 5,...</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.87</td>\n",
       "      <td>0.42</td>\n",
       "      <td>0.51</td>\n",
       "      <td>0.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MLPClassifier(hidden_layer_sizes=(50, 50, 50),...</td>\n",
       "      <td>{'activation': 'relu', 'alpha': 0.0001, 'hidde...</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.60</td>\n",
       "      <td>0.62</td>\n",
       "      <td>0.70</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>XGBClassifier(gamma=0.1, learning_rate=0.7, ma...</td>\n",
       "      <td>{'gamma': 0.1, 'learning_rate': 0.7, 'max_dept...</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.86</td>\n",
       "      <td>0.57</td>\n",
       "      <td>0.76</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-e895219b-2130-4730-8344-5fa55394f5b6')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-e895219b-2130-4730-8344-5fa55394f5b6 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-e895219b-2130-4730-8344-5fa55394f5b6');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                                              Models  \\\n",
       "0  ([DecisionTreeRegressor(criterion='friedman_ms...   \n",
       "1  (DecisionTreeClassifier(max_depth=20, max_feat...   \n",
       "2  DecisionTreeClassifier(max_depth=1, splitter='...   \n",
       "3                                           SVC(C=5)   \n",
       "4  (DecisionTreeClassifier(max_depth=1, random_st...   \n",
       "5               KNeighborsClassifier(n_neighbors=25)   \n",
       "6  (ExtraTreeClassifier(max_depth=5, random_state...   \n",
       "7  MLPClassifier(hidden_layer_sizes=(50, 50, 50),...   \n",
       "8  XGBClassifier(gamma=0.1, learning_rate=0.7, ma...   \n",
       "\n",
       "                                     Best parameters  Score   AUC  F1-score  \\\n",
       "0         {'learning_rate': 0.2, 'n_estimators': 50}   0.67  0.85      0.51   \n",
       "1  {'class_weight': None, 'max_depth': 20, 'n_est...   0.66  0.82      0.21   \n",
       "2             {'max_depth': 1, 'splitter': 'random'}   0.64  0.54      0.00   \n",
       "3                          {'C': 5, 'kernel': 'rbf'}   0.70  0.94      0.67   \n",
       "4         {'learning_rate': 0.5, 'n_estimators': 50}   0.69  0.83      0.47   \n",
       "5                                {'n_neighbors': 25}   0.65  0.89      0.33   \n",
       "6  {'class_weight': {0: 1, 1: 5}, 'max_depth': 5,...   0.70  0.87      0.42   \n",
       "7  {'activation': 'relu', 'alpha': 0.0001, 'hidde...   0.71  0.86      0.60   \n",
       "8  {'gamma': 0.1, 'learning_rate': 0.7, 'max_dept...   0.69  0.86      0.57   \n",
       "\n",
       "   Precision  Recall  \n",
       "0       0.53    0.52  \n",
       "1       0.80    0.12  \n",
       "2       0.00    0.00  \n",
       "3       0.78    0.68  \n",
       "4       0.44    0.50  \n",
       "5       0.75    0.28  \n",
       "6       0.51    0.42  \n",
       "7       0.62    0.70  \n",
       "8       0.76    0.50  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_mX_pca, grid_ml_pca = Algo_search(models = models, params = params, X_train=X, y_train=y, X_test=X_test2, y_test=y_test2)\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "grid_ml_pca"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bkNsFWQBsmW6"
   },
   "source": [
    "### Explainable AI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 66,
     "referenced_widgets": [
      "059e9ac27df342d99ca05d049089d7c4",
      "a20279445b4c4c72a73779f54da3ae67",
      "ebd580561949448e8d0ccf5ba4453ff9",
      "ff9a01ae87124d55a125bb0a31982a8f",
      "644b671db7aa4002bc7d9d0e4b730934",
      "e396b657844b4593a301f78f9bc67b63",
      "959b2741caa14a4887d764761df12548",
      "3484006847c245b78e3397f32c30ffed",
      "583a4136a3784ca6a35950627b684fc6",
      "674cdb8623bd4e31b3732cc67afc2fd0",
      "e4493369d8f843b69d0e962a071f460a"
     ]
    },
    "executionInfo": {
     "elapsed": 1272285,
     "status": "ok",
     "timestamp": 1652082538173,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "SkhZMSeEsmW7",
    "outputId": "1c21412b-4ec6-4fd9-fc20-1c3bba65bf48"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using 256 background data samples could cause slower run times. Consider using shap.sample(data, K) or shap.kmeans(data, K) to summarize the background as K samples.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "059e9ac27df342d99ca05d049089d7c4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/305 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = MLPClassifier(activation= 'relu',\n",
    "hidden_layer_sizes= (50, 50, 50),\n",
    "alpha= 0.0001,\n",
    "learning_rate= 'adaptive',\n",
    "solver= 'adam'\n",
    ")\n",
    "\n",
    "clf.fit(X, y)\n",
    "\n",
    "explainer = shap.KernelExplainer(clf.predict,X)\n",
    "shap_values = explainer.shap_values(X_test2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 585
    },
    "executionInfo": {
     "elapsed": 2076,
     "status": "ok",
     "timestamp": 1652082540923,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "OsK-oPU-smW9",
    "outputId": "b361b60d-ad77-473b-a448-1991cb031cc4"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "shap.summary_plot(shap_values,X_test2,feature_names=X.columns, max_display =20,show=False )\n",
    "plt.savefig('mlp_xai_yps_mgh.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "aS2EMRVysmW9"
   },
   "outputs": [],
   "source": [
    "vals= np.abs(shap_values).mean(0)\n",
    "feature_importance = pd.DataFrame(list(zip(X.columns,vals)),columns=['col_name','feature_importance_vals'])\n",
    "feature_importance.sort_values(by=['feature_importance_vals'],ascending=False,inplace=True)\n",
    "feature_importance.to_excel('mgh_yps_mgh_xai.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "tFUscgK0smW-"
   },
   "outputs": [],
   "source": [
    "lp =  [i for i in list(feature_importance['col_name'])[:21] if str(i).isdigit()]\n",
    "\n",
    "load_trans = loading_matrix[lp].T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowYPStoMGHmpl.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EgXtDBt5ABnp"
   },
   "source": [
    "# ICL dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "A_RW1U7FAKuw"
   },
   "outputs": [],
   "source": [
    "X = icl_con.drop(['WHO max'], axis=1)\n",
    "y = icl_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "3SGA_i5LAKux"
   },
   "outputs": [],
   "source": [
    "X_test1 = yps_con.drop(['WHO max'], axis=1)\n",
    "y_test1 = yps_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "PbCh-T4gAKug"
   },
   "outputs": [],
   "source": [
    "X_test2 = mgh_con.drop(['WHO max'], axis=1)\n",
    "y_test2 = mgh_con['WHO max']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "rvUC3KvQ700P"
   },
   "outputs": [],
   "source": [
    "pca = PCA(.90)\n",
    " \n",
    "X = pca.fit_transform(X[col])\n",
    "X = pd.DataFrame(X)\n",
    "X['KIDNEY'] = list(icl_con['KIDNEY'])\n",
    "X['Age cat'] = list(icl_con['Age cat'])\n",
    "\n",
    "X_test1 = pca.transform(X_test1[col])\n",
    "X_test1 = pd.DataFrame(X_test1)\n",
    "X_test1['KIDNEY'] = list(yps_con['KIDNEY'])\n",
    "X_test1['Age cat'] = list(yps_con['Age cat'])\n",
    "\n",
    "X_test2 = pca.transform(X_test2[col])\n",
    "X_test2 = pd.DataFrame(X_test2)\n",
    "X_test2['KIDNEY'] = list(mgh_con['KIDNEY'])\n",
    "X_test2['Age cat'] = list(mgh_con['Age cat'])\n",
    " \n",
    "explained_variance = pca.explained_variance_ratio_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 488
    },
    "executionInfo": {
     "elapsed": 219,
     "status": "ok",
     "timestamp": 1652088434914,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "9NwHm_ahug8A",
    "outputId": "c5fe601a-79e3-4575-fcd5-920fb4e0f33a"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-0245635a-c416-43de-83b0-340f99da042e\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>...</th>\n",
       "      <th>18</th>\n",
       "      <th>19</th>\n",
       "      <th>20</th>\n",
       "      <th>21</th>\n",
       "      <th>22</th>\n",
       "      <th>23</th>\n",
       "      <th>24</th>\n",
       "      <th>25</th>\n",
       "      <th>26</th>\n",
       "      <th>27</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PADI2</th>\n",
       "      <td>0.449112</td>\n",
       "      <td>-0.351198</td>\n",
       "      <td>0.372750</td>\n",
       "      <td>-0.080031</td>\n",
       "      <td>-0.022563</td>\n",
       "      <td>0.212563</td>\n",
       "      <td>-0.125401</td>\n",
       "      <td>0.167963</td>\n",
       "      <td>-0.006620</td>\n",
       "      <td>0.315036</td>\n",
       "      <td>...</td>\n",
       "      <td>0.149044</td>\n",
       "      <td>0.015410</td>\n",
       "      <td>-0.006099</td>\n",
       "      <td>-0.114161</td>\n",
       "      <td>0.065606</td>\n",
       "      <td>-0.044286</td>\n",
       "      <td>-0.017437</td>\n",
       "      <td>0.105818</td>\n",
       "      <td>-0.079829</td>\n",
       "      <td>0.074926</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TNF</th>\n",
       "      <td>0.222402</td>\n",
       "      <td>-0.644291</td>\n",
       "      <td>-0.088030</td>\n",
       "      <td>0.166297</td>\n",
       "      <td>0.317129</td>\n",
       "      <td>-0.017309</td>\n",
       "      <td>0.034202</td>\n",
       "      <td>-0.109869</td>\n",
       "      <td>-0.321428</td>\n",
       "      <td>0.011333</td>\n",
       "      <td>...</td>\n",
       "      <td>0.076572</td>\n",
       "      <td>0.017154</td>\n",
       "      <td>0.230089</td>\n",
       "      <td>0.050087</td>\n",
       "      <td>-0.064315</td>\n",
       "      <td>0.083880</td>\n",
       "      <td>0.001578</td>\n",
       "      <td>0.034597</td>\n",
       "      <td>0.004371</td>\n",
       "      <td>-0.087636</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NTF4</th>\n",
       "      <td>0.192649</td>\n",
       "      <td>-0.286860</td>\n",
       "      <td>-0.184748</td>\n",
       "      <td>-0.422365</td>\n",
       "      <td>-0.280658</td>\n",
       "      <td>-0.177679</td>\n",
       "      <td>0.140001</td>\n",
       "      <td>0.204948</td>\n",
       "      <td>-0.243305</td>\n",
       "      <td>-0.210081</td>\n",
       "      <td>...</td>\n",
       "      <td>0.202971</td>\n",
       "      <td>-0.020633</td>\n",
       "      <td>0.039845</td>\n",
       "      <td>-0.098247</td>\n",
       "      <td>-0.114510</td>\n",
       "      <td>-0.200051</td>\n",
       "      <td>-0.095271</td>\n",
       "      <td>0.042072</td>\n",
       "      <td>0.081888</td>\n",
       "      <td>-0.011702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SH2D1A</th>\n",
       "      <td>0.288056</td>\n",
       "      <td>-0.144189</td>\n",
       "      <td>-0.267626</td>\n",
       "      <td>0.125419</td>\n",
       "      <td>-0.130242</td>\n",
       "      <td>0.145341</td>\n",
       "      <td>0.355644</td>\n",
       "      <td>0.189862</td>\n",
       "      <td>-0.241801</td>\n",
       "      <td>-0.152290</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.187571</td>\n",
       "      <td>-0.017491</td>\n",
       "      <td>-0.108204</td>\n",
       "      <td>-0.310805</td>\n",
       "      <td>-0.100792</td>\n",
       "      <td>-0.034822</td>\n",
       "      <td>0.022162</td>\n",
       "      <td>-0.177597</td>\n",
       "      <td>-0.012979</td>\n",
       "      <td>-0.042379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>TNFRSF9</th>\n",
       "      <td>-0.027562</td>\n",
       "      <td>0.225177</td>\n",
       "      <td>-0.350171</td>\n",
       "      <td>0.629305</td>\n",
       "      <td>-0.180002</td>\n",
       "      <td>0.055368</td>\n",
       "      <td>-0.059731</td>\n",
       "      <td>-0.033287</td>\n",
       "      <td>-0.183046</td>\n",
       "      <td>0.128914</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.044119</td>\n",
       "      <td>-0.003965</td>\n",
       "      <td>-0.061224</td>\n",
       "      <td>0.020185</td>\n",
       "      <td>-0.161343</td>\n",
       "      <td>0.116450</td>\n",
       "      <td>-0.050233</td>\n",
       "      <td>0.004984</td>\n",
       "      <td>0.027244</td>\n",
       "      <td>-0.002204</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>PGF</th>\n",
       "      <td>0.449012</td>\n",
       "      <td>-0.223170</td>\n",
       "      <td>-0.543811</td>\n",
       "      <td>-0.090254</td>\n",
       "      <td>-0.422730</td>\n",
       "      <td>0.271714</td>\n",
       "      <td>-0.077709</td>\n",
       "      <td>0.009349</td>\n",
       "      <td>-0.055924</td>\n",
       "      <td>0.021341</td>\n",
       "      <td>...</td>\n",
       "      <td>0.011338</td>\n",
       "      <td>-0.155276</td>\n",
       "      <td>0.072631</td>\n",
       "      <td>0.061127</td>\n",
       "      <td>0.070614</td>\n",
       "      <td>-0.094500</td>\n",
       "      <td>0.104132</td>\n",
       "      <td>-0.073430</td>\n",
       "      <td>0.026209</td>\n",
       "      <td>0.047733</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FCRL3</th>\n",
       "      <td>-0.168464</td>\n",
       "      <td>-0.075079</td>\n",
       "      <td>0.048051</td>\n",
       "      <td>0.544284</td>\n",
       "      <td>0.164175</td>\n",
       "      <td>-0.208679</td>\n",
       "      <td>0.142044</td>\n",
       "      <td>0.137180</td>\n",
       "      <td>0.182361</td>\n",
       "      <td>0.011864</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.194516</td>\n",
       "      <td>-0.184986</td>\n",
       "      <td>0.060718</td>\n",
       "      <td>-0.040108</td>\n",
       "      <td>-0.120731</td>\n",
       "      <td>-0.109990</td>\n",
       "      <td>0.020461</td>\n",
       "      <td>-0.000041</td>\n",
       "      <td>0.147353</td>\n",
       "      <td>0.029331</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>VEGFA</th>\n",
       "      <td>0.744103</td>\n",
       "      <td>-0.332496</td>\n",
       "      <td>-0.151798</td>\n",
       "      <td>-0.163632</td>\n",
       "      <td>-0.111182</td>\n",
       "      <td>0.119943</td>\n",
       "      <td>-0.007310</td>\n",
       "      <td>-0.150059</td>\n",
       "      <td>-0.025193</td>\n",
       "      <td>-0.031916</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.118396</td>\n",
       "      <td>-0.011655</td>\n",
       "      <td>0.038851</td>\n",
       "      <td>-0.025670</td>\n",
       "      <td>0.034913</td>\n",
       "      <td>-0.000988</td>\n",
       "      <td>-0.033778</td>\n",
       "      <td>0.012884</td>\n",
       "      <td>-0.100690</td>\n",
       "      <td>0.058078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>KRT19</th>\n",
       "      <td>0.443195</td>\n",
       "      <td>-0.641168</td>\n",
       "      <td>0.253166</td>\n",
       "      <td>-0.344109</td>\n",
       "      <td>0.120253</td>\n",
       "      <td>-0.035220</td>\n",
       "      <td>-0.020874</td>\n",
       "      <td>-0.133898</td>\n",
       "      <td>-0.116732</td>\n",
       "      <td>0.028670</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.153926</td>\n",
       "      <td>0.014587</td>\n",
       "      <td>-0.004665</td>\n",
       "      <td>0.190300</td>\n",
       "      <td>-0.019797</td>\n",
       "      <td>-0.037615</td>\n",
       "      <td>0.049000</td>\n",
       "      <td>0.044832</td>\n",
       "      <td>0.042877</td>\n",
       "      <td>-0.035642</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>NRTN</th>\n",
       "      <td>0.053936</td>\n",
       "      <td>-0.280731</td>\n",
       "      <td>0.057808</td>\n",
       "      <td>0.251609</td>\n",
       "      <td>-0.389682</td>\n",
       "      <td>-0.606336</td>\n",
       "      <td>-0.226725</td>\n",
       "      <td>0.079948</td>\n",
       "      <td>-0.173546</td>\n",
       "      <td>-0.240695</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.108487</td>\n",
       "      <td>-0.018950</td>\n",
       "      <td>0.052568</td>\n",
       "      <td>-0.059051</td>\n",
       "      <td>-0.081873</td>\n",
       "      <td>-0.078550</td>\n",
       "      <td>-0.039111</td>\n",
       "      <td>-0.032164</td>\n",
       "      <td>-0.118319</td>\n",
       "      <td>0.040510</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>168 rows × 28 columns</p>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0245635a-c416-43de-83b0-340f99da042e')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-0245635a-c416-43de-83b0-340f99da042e button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-0245635a-c416-43de-83b0-340f99da042e');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "               0         1         2         3         4         5         6   \\\n",
       "PADI2    0.449112 -0.351198  0.372750 -0.080031 -0.022563  0.212563 -0.125401   \n",
       "TNF      0.222402 -0.644291 -0.088030  0.166297  0.317129 -0.017309  0.034202   \n",
       "NTF4     0.192649 -0.286860 -0.184748 -0.422365 -0.280658 -0.177679  0.140001   \n",
       "SH2D1A   0.288056 -0.144189 -0.267626  0.125419 -0.130242  0.145341  0.355644   \n",
       "TNFRSF9 -0.027562  0.225177 -0.350171  0.629305 -0.180002  0.055368 -0.059731   \n",
       "...           ...       ...       ...       ...       ...       ...       ...   \n",
       "PGF      0.449012 -0.223170 -0.543811 -0.090254 -0.422730  0.271714 -0.077709   \n",
       "FCRL3   -0.168464 -0.075079  0.048051  0.544284  0.164175 -0.208679  0.142044   \n",
       "VEGFA    0.744103 -0.332496 -0.151798 -0.163632 -0.111182  0.119943 -0.007310   \n",
       "KRT19    0.443195 -0.641168  0.253166 -0.344109  0.120253 -0.035220 -0.020874   \n",
       "NRTN     0.053936 -0.280731  0.057808  0.251609 -0.389682 -0.606336 -0.226725   \n",
       "\n",
       "               7         8         9   ...        18        19        20  \\\n",
       "PADI2    0.167963 -0.006620  0.315036  ...  0.149044  0.015410 -0.006099   \n",
       "TNF     -0.109869 -0.321428  0.011333  ...  0.076572  0.017154  0.230089   \n",
       "NTF4     0.204948 -0.243305 -0.210081  ...  0.202971 -0.020633  0.039845   \n",
       "SH2D1A   0.189862 -0.241801 -0.152290  ... -0.187571 -0.017491 -0.108204   \n",
       "TNFRSF9 -0.033287 -0.183046  0.128914  ... -0.044119 -0.003965 -0.061224   \n",
       "...           ...       ...       ...  ...       ...       ...       ...   \n",
       "PGF      0.009349 -0.055924  0.021341  ...  0.011338 -0.155276  0.072631   \n",
       "FCRL3    0.137180  0.182361  0.011864  ... -0.194516 -0.184986  0.060718   \n",
       "VEGFA   -0.150059 -0.025193 -0.031916  ... -0.118396 -0.011655  0.038851   \n",
       "KRT19   -0.133898 -0.116732  0.028670  ... -0.153926  0.014587 -0.004665   \n",
       "NRTN     0.079948 -0.173546 -0.240695  ... -0.108487 -0.018950  0.052568   \n",
       "\n",
       "               21        22        23        24        25        26        27  \n",
       "PADI2   -0.114161  0.065606 -0.044286 -0.017437  0.105818 -0.079829  0.074926  \n",
       "TNF      0.050087 -0.064315  0.083880  0.001578  0.034597  0.004371 -0.087636  \n",
       "NTF4    -0.098247 -0.114510 -0.200051 -0.095271  0.042072  0.081888 -0.011702  \n",
       "SH2D1A  -0.310805 -0.100792 -0.034822  0.022162 -0.177597 -0.012979 -0.042379  \n",
       "TNFRSF9  0.020185 -0.161343  0.116450 -0.050233  0.004984  0.027244 -0.002204  \n",
       "...           ...       ...       ...       ...       ...       ...       ...  \n",
       "PGF      0.061127  0.070614 -0.094500  0.104132 -0.073430  0.026209  0.047733  \n",
       "FCRL3   -0.040108 -0.120731 -0.109990  0.020461 -0.000041  0.147353  0.029331  \n",
       "VEGFA   -0.025670  0.034913 -0.000988 -0.033778  0.012884 -0.100690  0.058078  \n",
       "KRT19    0.190300 -0.019797 -0.037615  0.049000  0.044832  0.042877 -0.035642  \n",
       "NRTN    -0.059051 -0.081873 -0.078550 -0.039111 -0.032164 -0.118319  0.040510  \n",
       "\n",
       "[168 rows x 28 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loadings = pca.components_.T * np.sqrt(pca.explained_variance_)\n",
    "loading_matrix = pd.DataFrame(loadings, index=icl_con.drop(['WHO max'], axis=1)[col].columns)\n",
    "loading_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "mzlVuWkDJrq9"
   },
   "outputs": [],
   "source": [
    "load_trans = loading_matrix.T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowICLtask2.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lCcoJG9_AKuy"
   },
   "source": [
    "## YPS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 301
    },
    "executionInfo": {
     "elapsed": 6486,
     "status": "ok",
     "timestamp": 1649061926537,
     "user": {
      "displayName": "PV Clinical",
      "userId": "13901498519864370078"
     },
     "user_tz": -180
    },
    "id": "PvGf1Dxx6Ndi",
    "outputId": "39885f33-2cb9-41ee-8eea-46b9a9d3dfe1"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-00c64157-6bed-4772-8d9e-2f3206b754e4\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>Models</th>\n",
       "      <th>Best parameters</th>\n",
       "      <th>Score</th>\n",
       "      <th>AUC</th>\n",
       "      <th>F1-score</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>([DecisionTreeRegressor(criterion='friedman_ms...</td>\n",
       "      <td>{'learning_rate': 0.4, 'n_estimators': 50}</td>\n",
       "      <td>0.79</td>\n",
       "      <td>0.26</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=20, max_feat...</td>\n",
       "      <td>{'class_weight': {0: 1, 1: 5}, 'max_depth': 20...</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DecisionTreeClassifier(max_depth=1, splitter='...</td>\n",
       "      <td>{'max_depth': 1, 'splitter': 'random'}</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC(C=5, kernel='sigmoid')</td>\n",
       "      <td>{'C': 5, 'kernel': 'sigmoid'}</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=1, random_st...</td>\n",
       "      <td>{'learning_rate': 0.5, 'n_estimators': 30}</td>\n",
       "      <td>0.78</td>\n",
       "      <td>0.51</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KNeighborsClassifier(n_neighbors=10)</td>\n",
       "      <td>{'n_neighbors': 10}</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>(ExtraTreeClassifier(max_depth=5, random_state...</td>\n",
       "      <td>{'class_weight': {0: 1, 1: 5}, 'max_depth': 5,...</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MLPClassifier(activation='logistic', hidden_la...</td>\n",
       "      <td>{'activation': 'logistic', 'alpha': 0.0001, 'h...</td>\n",
       "      <td>0.80</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.35</td>\n",
       "      <td>0.45</td>\n",
       "      <td>0.33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>XGBClassifier(gamma=1.6, max_depth=5, n_estima...</td>\n",
       "      <td>{'gamma': 1.6, 'learning_rate': 0.1, 'max_dept...</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.29</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-00c64157-6bed-4772-8d9e-2f3206b754e4')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-00c64157-6bed-4772-8d9e-2f3206b754e4 button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-00c64157-6bed-4772-8d9e-2f3206b754e4');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                                              Models  \\\n",
       "0  ([DecisionTreeRegressor(criterion='friedman_ms...   \n",
       "1  (DecisionTreeClassifier(max_depth=20, max_feat...   \n",
       "2  DecisionTreeClassifier(max_depth=1, splitter='...   \n",
       "3                         SVC(C=5, kernel='sigmoid')   \n",
       "4  (DecisionTreeClassifier(max_depth=1, random_st...   \n",
       "5               KNeighborsClassifier(n_neighbors=10)   \n",
       "6  (ExtraTreeClassifier(max_depth=5, random_state...   \n",
       "7  MLPClassifier(activation='logistic', hidden_la...   \n",
       "8  XGBClassifier(gamma=1.6, max_depth=5, n_estima...   \n",
       "\n",
       "                                     Best parameters  Score   AUC  F1-score  \\\n",
       "0         {'learning_rate': 0.4, 'n_estimators': 50}   0.79  0.26      0.07   \n",
       "1  {'class_weight': {0: 1, 1: 5}, 'max_depth': 20...   0.81  0.63      0.13   \n",
       "2             {'max_depth': 1, 'splitter': 'random'}   0.80  0.50      0.13   \n",
       "3                      {'C': 5, 'kernel': 'sigmoid'}   0.82  0.81      0.69   \n",
       "4         {'learning_rate': 0.5, 'n_estimators': 30}   0.78  0.51      0.20   \n",
       "5                                {'n_neighbors': 10}   0.81  0.82      0.00   \n",
       "6  {'class_weight': {0: 1, 1: 5}, 'max_depth': 5,...   0.81  0.50      0.13   \n",
       "7  {'activation': 'logistic', 'alpha': 0.0001, 'h...   0.80  0.82      0.35   \n",
       "8  {'gamma': 1.6, 'learning_rate': 0.1, 'max_dept...   0.81  0.29      0.00   \n",
       "\n",
       "   Precision  Recall  \n",
       "0       0.05    0.10  \n",
       "1       0.20    0.10  \n",
       "2       0.20    0.10  \n",
       "3       0.83    0.63  \n",
       "4       0.20    0.20  \n",
       "5       0.00    0.00  \n",
       "6       0.20    0.10  \n",
       "7       0.45    0.33  \n",
       "8       0.00    0.00  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_mX_pca, grid_ml_pca = Algo_search(models = models, params = params, X_train=X, y_train=y, X_test=X_test1, y_test=y_test1)\n",
    "\n",
    "grid_ml_pca"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "PnGJf1ibuNP6"
   },
   "source": [
    "### Explainable AI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 49,
     "referenced_widgets": [
      "9eec49ded4394f89961493b913ca2cbb",
      "410a75bfc02b478ea5ee00048a740ca9",
      "cecb237fe40b4c8fa4c9da0d8d524d04",
      "a31f97d00ef34e629fd30c2c2389e878",
      "9148ddb7d90b4b64bc570d6a687369c4",
      "7d3f69803a8641cd8e56e86501e03a50",
      "6a84c4fd5b3a432ab9deb67864ae56ef",
      "e3bf145254de4fe4b6fc42acb4c55bd1",
      "5c890b497a194ffaa5ff64c296f3d745",
      "21494dff3dc54b938bb3c4af7bddc2bf",
      "32a6171162984f68b7e465a3d12f3494"
     ]
    },
    "executionInfo": {
     "elapsed": 302946,
     "status": "ok",
     "timestamp": 1652088964816,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "ZgbXvAdkuNP6",
    "outputId": "2bc12e28-c3aa-4690-c975-b236122b87e2"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9eec49ded4394f89961493b913ca2cbb",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/256 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = SVC(C=5,\n",
    "          kernel=\"sigmoid\"\n",
    ")\n",
    "\n",
    "\n",
    "clf.fit(X, y)\n",
    "\n",
    "explainer = shap.KernelExplainer(clf.predict,X)\n",
    "shap_values = explainer.shap_values(X_test1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 585
    },
    "executionInfo": {
     "elapsed": 2003,
     "status": "ok",
     "timestamp": 1652088966788,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "uqUHATO_uNP7",
    "outputId": "4133128a-83db-4909-c837-8867fb15da3c"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "shap.summary_plot(shap_values,X_test1,feature_names=X.columns, max_display =20,show=False )\n",
    "plt.savefig('mlp_xai_icl_yps.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "hsh-CaybuNP8"
   },
   "outputs": [],
   "source": [
    "vals= np.abs(shap_values).mean(0)\n",
    "feature_importance = pd.DataFrame(list(zip(X.columns,vals)),columns=['col_name','feature_importance_vals'])\n",
    "feature_importance.sort_values(by=['feature_importance_vals'],ascending=False,inplace=True)\n",
    "feature_importance.to_excel('mgh_icl_yps_xai.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "wnvLSvMXuNP8"
   },
   "outputs": [],
   "source": [
    "lp =  [i for i in list(feature_importance['col_name'])[:21] if str(i).isdigit()]\n",
    "\n",
    "load_trans = loading_matrix[lp].T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowICLtoYPSmpl.xlsx')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ruhQu9RdAKu1"
   },
   "source": [
    "## MGH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 301
    },
    "executionInfo": {
     "elapsed": 5804,
     "status": "ok",
     "timestamp": 1649062539597,
     "user": {
      "displayName": "PV Clinical",
      "userId": "13901498519864370078"
     },
     "user_tz": -180
    },
    "id": "DTaVLyFa84gX",
    "outputId": "1b30acb4-c50b-477f-d69c-b432ec4bccd5"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "  <div id=\"df-0726326f-4516-4f91-87a6-04042eec635a\">\n",
       "    <div class=\"colab-df-container\">\n",
       "      <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>Models</th>\n",
       "      <th>Best parameters</th>\n",
       "      <th>Score</th>\n",
       "      <th>AUC</th>\n",
       "      <th>F1-score</th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>([DecisionTreeRegressor(criterion='friedman_ms...</td>\n",
       "      <td>{'learning_rate': 0.4, 'n_estimators': 500}</td>\n",
       "      <td>0.70</td>\n",
       "      <td>0.34</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=5, max_featu...</td>\n",
       "      <td>{'class_weight': {0: 1, 1: 10}, 'max_depth': 5...</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.56</td>\n",
       "      <td>0.13</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DecisionTreeClassifier(max_depth=1, splitter='...</td>\n",
       "      <td>{'max_depth': 1, 'splitter': 'random'}</td>\n",
       "      <td>0.63</td>\n",
       "      <td>0.43</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.11</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SVC(C=5, kernel='sigmoid')</td>\n",
       "      <td>{'C': 5, 'kernel': 'sigmoid'}</td>\n",
       "      <td>0.73</td>\n",
       "      <td>0.81</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.63</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(DecisionTreeClassifier(max_depth=1, random_st...</td>\n",
       "      <td>{'learning_rate': 0.5, 'n_estimators': 30}</td>\n",
       "      <td>0.69</td>\n",
       "      <td>0.51</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>KNeighborsClassifier(n_neighbors=10)</td>\n",
       "      <td>{'n_neighbors': 10}</td>\n",
       "      <td>0.65</td>\n",
       "      <td>0.82</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>(ExtraTreeClassifier(max_depth=5, random_state...</td>\n",
       "      <td>{'class_weight': {0: 1, 1: 5}, 'max_depth': 5,...</td>\n",
       "      <td>0.71</td>\n",
       "      <td>0.50</td>\n",
       "      <td>0.20</td>\n",
       "      <td>0.30</td>\n",
       "      <td>0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>MLPClassifier(activation='tanh', hidden_layer_...</td>\n",
       "      <td>{'activation': 'tanh', 'alpha': 0.0001, 'hidde...</td>\n",
       "      <td>0.74</td>\n",
       "      <td>0.88</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.40</td>\n",
       "      <td>0.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>XGBClassifier(max_depth=5, n_estimators=50, re...</td>\n",
       "      <td>{'gamma': 0, 'learning_rate': 0.1, 'max_depth'...</td>\n",
       "      <td>0.66</td>\n",
       "      <td>0.33</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>\n",
       "      <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-0726326f-4516-4f91-87a6-04042eec635a')\"\n",
       "              title=\"Convert this dataframe to an interactive table.\"\n",
       "              style=\"display:none;\">\n",
       "        \n",
       "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
       "       width=\"24px\">\n",
       "    <path d=\"M0 0h24v24H0V0z\" fill=\"none\"/>\n",
       "    <path d=\"M18.56 5.44l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94zm-11 1L8.5 8.5l.94-2.06 2.06-.94-2.06-.94L8.5 2.5l-.94 2.06-2.06.94zm10 10l.94 2.06.94-2.06 2.06-.94-2.06-.94-.94-2.06-.94 2.06-2.06.94z\"/><path d=\"M17.41 7.96l-1.37-1.37c-.4-.4-.92-.59-1.43-.59-.52 0-1.04.2-1.43.59L10.3 9.45l-7.72 7.72c-.78.78-.78 2.05 0 2.83L4 21.41c.39.39.9.59 1.41.59.51 0 1.02-.2 1.41-.59l7.78-7.78 2.81-2.81c.8-.78.8-2.07 0-2.86zM5.41 20L4 18.59l7.72-7.72 1.47 1.35L5.41 20z\"/>\n",
       "  </svg>\n",
       "      </button>\n",
       "      \n",
       "  <style>\n",
       "    .colab-df-container {\n",
       "      display:flex;\n",
       "      flex-wrap:wrap;\n",
       "      gap: 12px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert {\n",
       "      background-color: #E8F0FE;\n",
       "      border: none;\n",
       "      border-radius: 50%;\n",
       "      cursor: pointer;\n",
       "      display: none;\n",
       "      fill: #1967D2;\n",
       "      height: 32px;\n",
       "      padding: 0 0 0 0;\n",
       "      width: 32px;\n",
       "    }\n",
       "\n",
       "    .colab-df-convert:hover {\n",
       "      background-color: #E2EBFA;\n",
       "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
       "      fill: #174EA6;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert {\n",
       "      background-color: #3B4455;\n",
       "      fill: #D2E3FC;\n",
       "    }\n",
       "\n",
       "    [theme=dark] .colab-df-convert:hover {\n",
       "      background-color: #434B5C;\n",
       "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
       "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
       "      fill: #FFFFFF;\n",
       "    }\n",
       "  </style>\n",
       "\n",
       "      <script>\n",
       "        const buttonEl =\n",
       "          document.querySelector('#df-0726326f-4516-4f91-87a6-04042eec635a button.colab-df-convert');\n",
       "        buttonEl.style.display =\n",
       "          google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
       "\n",
       "        async function convertToInteractive(key) {\n",
       "          const element = document.querySelector('#df-0726326f-4516-4f91-87a6-04042eec635a');\n",
       "          const dataTable =\n",
       "            await google.colab.kernel.invokeFunction('convertToInteractive',\n",
       "                                                     [key], {});\n",
       "          if (!dataTable) return;\n",
       "\n",
       "          const docLinkHtml = 'Like what you see? Visit the ' +\n",
       "            '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
       "            + ' to learn more about interactive tables.';\n",
       "          element.innerHTML = '';\n",
       "          dataTable['output_type'] = 'display_data';\n",
       "          await google.colab.output.renderOutput(dataTable, element);\n",
       "          const docLink = document.createElement('div');\n",
       "          docLink.innerHTML = docLinkHtml;\n",
       "          element.appendChild(docLink);\n",
       "        }\n",
       "      </script>\n",
       "    </div>\n",
       "  </div>\n",
       "  "
      ],
      "text/plain": [
       "                                              Models  \\\n",
       "0  ([DecisionTreeRegressor(criterion='friedman_ms...   \n",
       "1  (DecisionTreeClassifier(max_depth=5, max_featu...   \n",
       "2  DecisionTreeClassifier(max_depth=1, splitter='...   \n",
       "3                         SVC(C=5, kernel='sigmoid')   \n",
       "4  (DecisionTreeClassifier(max_depth=1, random_st...   \n",
       "5               KNeighborsClassifier(n_neighbors=10)   \n",
       "6  (ExtraTreeClassifier(max_depth=5, random_state...   \n",
       "7  MLPClassifier(activation='tanh', hidden_layer_...   \n",
       "8  XGBClassifier(max_depth=5, n_estimators=50, re...   \n",
       "\n",
       "                                     Best parameters  Score   AUC  F1-score  \\\n",
       "0        {'learning_rate': 0.4, 'n_estimators': 500}   0.70  0.34      0.10   \n",
       "1  {'class_weight': {0: 1, 1: 10}, 'max_depth': 5...   0.65  0.56      0.13   \n",
       "2             {'max_depth': 1, 'splitter': 'random'}   0.63  0.43      0.11   \n",
       "3                      {'C': 5, 'kernel': 'sigmoid'}   0.73  0.81      0.69   \n",
       "4         {'learning_rate': 0.5, 'n_estimators': 30}   0.69  0.51      0.20   \n",
       "5                                {'n_neighbors': 10}   0.65  0.82      0.00   \n",
       "6  {'class_weight': {0: 1, 1: 5}, 'max_depth': 5,...   0.71  0.50      0.20   \n",
       "7  {'activation': 'tanh', 'alpha': 0.0001, 'hidde...   0.74  0.88      0.33   \n",
       "8  {'gamma': 0, 'learning_rate': 0.1, 'max_depth'...   0.66  0.33      0.00   \n",
       "\n",
       "   Precision  Recall  \n",
       "0       0.10    0.10  \n",
       "1       0.20    0.10  \n",
       "2       0.11    0.12  \n",
       "3       0.83    0.63  \n",
       "4       0.20    0.20  \n",
       "5       0.00    0.00  \n",
       "6       0.30    0.17  \n",
       "7       0.40    0.30  \n",
       "8       0.00    0.00  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_mX_pca, grid_ml_pca = Algo_search(models = models, params = params, X_train=X, y_train=y, X_test=X_test2, y_test=y_test2)\n",
    "\n",
    "grid_ml_pca"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "tirhLfiquPrY"
   },
   "source": [
    "### Explainable AI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 49,
     "referenced_widgets": [
      "f1654d6e363b450da33f4e1807b19ece",
      "9c56afe1c1a449e4a6ca5f2e5ff19962",
      "a849e25a234b4580b06b496bef8e00f2",
      "62328e01c5744b239c4fde0535389ab4",
      "e6fa20e8a7bf4d0fbfbd82add6387f3c",
      "677e67e37cae4f4993bcfb90b64ea414",
      "fbea06001a724a348d8f6591648c1cc8",
      "0001d478c10e4e0793b0847fddb5b61f",
      "53615d583f7642a9bbe594454d26d54e",
      "ecec72f3253d4406809e24bf55bb7b51",
      "5f1f2e3d23e14219a25fb7ded941f7cc"
     ]
    },
    "executionInfo": {
     "elapsed": 285377,
     "status": "ok",
     "timestamp": 1652089252146,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "grffmweHuPrZ",
    "outputId": "2c898d00-9ba2-4e40-dbdd-a45bf6092cf1"
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f1654d6e363b450da33f4e1807b19ece",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/305 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "clf = SVC(C=5,\n",
    "          kernel=\"sigmoid\"\n",
    ")\n",
    "\n",
    "clf.fit(X, y)\n",
    "\n",
    "explainer = shap.KernelExplainer(clf.predict,X)\n",
    "shap_values = explainer.shap_values(X_test2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 585
    },
    "executionInfo": {
     "elapsed": 1935,
     "status": "ok",
     "timestamp": 1652089254062,
     "user": {
      "displayName": "STELLA DIMITSAKI",
      "userId": "10918269675847450233"
     },
     "user_tz": -180
    },
    "id": "9BIBXSBSuPra",
    "outputId": "dd92bc60-433a-45cc-90e8-0aaa7e325fc8"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x684 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "shap.summary_plot(shap_values,X_test2,feature_names=X.columns, max_display =20,show=False )\n",
    "plt.savefig('mlp_xai_icl_mgh.svg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "KxN5L8g0uPra"
   },
   "outputs": [],
   "source": [
    "vals= np.abs(shap_values).mean(0)\n",
    "feature_importance = pd.DataFrame(list(zip(X.columns,vals)),columns=['col_name','feature_importance_vals'])\n",
    "feature_importance.sort_values(by=['feature_importance_vals'],ascending=False,inplace=True)\n",
    "feature_importance.to_excel('mgh_icl_mgh_xai.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "id": "OgX9xQqCuPrb"
   },
   "outputs": [],
   "source": [
    "lp =  [i for i in list(feature_importance['col_name'])[:21] if str(i).isdigit()]\n",
    "\n",
    "load_trans = loading_matrix[lp].T\n",
    "magnitude = 0\n",
    "loading_pow = {}\n",
    "for i in load_trans.columns:\n",
    "  magnitude = load_trans[i].pow(2).sum()\n",
    "  loading_pow[i] = magnitude\n",
    "\n",
    "pd.DataFrame.from_dict(loading_pow, orient='index').to_excel('loadingspowICLtoMGHmpl.xlsx')"
   ]
  }
 ],
 "metadata": {
  "colab": {
   "name": "Task 2.ipynb",
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "0001d478c10e4e0793b0847fddb5b61f": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "0579c1ffe3734c3e87b041dfc89a5048": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "059e9ac27df342d99ca05d049089d7c4": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_a20279445b4c4c72a73779f54da3ae67",
       "IPY_MODEL_ebd580561949448e8d0ccf5ba4453ff9",
       "IPY_MODEL_ff9a01ae87124d55a125bb0a31982a8f"
      ],
      "layout": "IPY_MODEL_644b671db7aa4002bc7d9d0e4b730934"
     }
    },
    "0ed49d6fc73f4ee7930604ded0487c23": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_7ad4773718da4fd8a8bc165497d181de",
      "placeholder": "​",
      "style": "IPY_MODEL_e9c1c09b268a4fe18a535c14d36afa85",
      "value": " 55/55 [06:38&lt;00:00,  6.29s/it]"
     }
    },
    "10ebfb73b7494bed9a698b5024b2636e": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_6585c84dd0c140c6868ac60efa01a461",
      "placeholder": "​",
      "style": "IPY_MODEL_e7df0eeb8b364496b832d1bb5b89f854",
      "value": "100%"
     }
    },
    "15cba8dead5b45bdbfc3ef49b9ef1d2e": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "1b90209079f749a8bef476882a854e27": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "1e29de520a3a4c3dad23aee7a8169ac2": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "21494dff3dc54b938bb3c4af7bddc2bf": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "2aca8bd289c34a5c8b49bd64cfc9f2af": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_e1fed3d156a14076b5c1a47f75be79df",
      "placeholder": "​",
      "style": "IPY_MODEL_3c70c8e3cd49478dbf3988e53054edc1",
      "value": " 256/256 [32:44&lt;00:00,  7.21s/it]"
     }
    },
    "2b8a7b88de8243758573144ed63aadd2": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "30b057d3ed124a6fbc1e5922c74deb55": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "32a6171162984f68b7e465a3d12f3494": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "3484006847c245b78e3397f32c30ffed": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "3c70c8e3cd49478dbf3988e53054edc1": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "410a75bfc02b478ea5ee00048a740ca9": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_7d3f69803a8641cd8e56e86501e03a50",
      "placeholder": "​",
      "style": "IPY_MODEL_6a84c4fd5b3a432ab9deb67864ae56ef",
      "value": "100%"
     }
    },
    "4d5bf205a7694ef081ecc7a56305307d": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_663105b452d54ed983932a81b885adb2",
      "max": 256,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_30b057d3ed124a6fbc1e5922c74deb55",
      "value": 256
     }
    },
    "53615d583f7642a9bbe594454d26d54e": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "57a3b742ef2d47039cd217b8a4557bb4": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "583a4136a3784ca6a35950627b684fc6": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "5c890b497a194ffaa5ff64c296f3d745": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "5de3e8a13918447fb856d796bc3e4305": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_1b90209079f749a8bef476882a854e27",
      "placeholder": "​",
      "style": "IPY_MODEL_f9d7d05478bf4cee998fe3c38a0069c6",
      "value": " 55/55 [03:14&lt;00:00,  3.53s/it]"
     }
    },
    "5f1f2e3d23e14219a25fb7ded941f7cc": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "62328e01c5744b239c4fde0535389ab4": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_ecec72f3253d4406809e24bf55bb7b51",
      "placeholder": "​",
      "style": "IPY_MODEL_5f1f2e3d23e14219a25fb7ded941f7cc",
      "value": " 305/305 [04:45&lt;00:00,  1.12it/s]"
     }
    },
    "64483751c11e4e4fbde1e1a3e6391eae": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_f757b7edd0df4cbfb3513fe595663e08",
      "max": 55,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_bc95e906d0a749479b881f0747e5ad23",
      "value": 55
     }
    },
    "644b671db7aa4002bc7d9d0e4b730934": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "6585c84dd0c140c6868ac60efa01a461": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "6626a85c8dac4734b581824f25fc3d1b": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_c3840367809d463d9c694ef2a92e0ff5",
       "IPY_MODEL_69045c7a03864e48b20282fababbfe1f",
       "IPY_MODEL_5de3e8a13918447fb856d796bc3e4305"
      ],
      "layout": "IPY_MODEL_6bc5e43ffbf249658755d4844dec9525"
     }
    },
    "663105b452d54ed983932a81b885adb2": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "674cdb8623bd4e31b3732cc67afc2fd0": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "677e67e37cae4f4993bcfb90b64ea414": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "69045c7a03864e48b20282fababbfe1f": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_1e29de520a3a4c3dad23aee7a8169ac2",
      "max": 55,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_a170d8c8cf7f49e8aa86520bc406fbed",
      "value": 55
     }
    },
    "6a84c4fd5b3a432ab9deb67864ae56ef": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "6bc5e43ffbf249658755d4844dec9525": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "6e13b236fb9f4b3aad08893bf5812e23": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_10ebfb73b7494bed9a698b5024b2636e",
       "IPY_MODEL_4d5bf205a7694ef081ecc7a56305307d",
       "IPY_MODEL_2aca8bd289c34a5c8b49bd64cfc9f2af"
      ],
      "layout": "IPY_MODEL_7f3109affa534cffa948126c54ba3f0d"
     }
    },
    "7ad4773718da4fd8a8bc165497d181de": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "7d3f69803a8641cd8e56e86501e03a50": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "7f3109affa534cffa948126c54ba3f0d": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "9148ddb7d90b4b64bc570d6a687369c4": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "9348543a8cd84980809203081c777509": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_a92d7ad7d07949a09e6de9057cb27e4f",
      "placeholder": "​",
      "style": "IPY_MODEL_15cba8dead5b45bdbfc3ef49b9ef1d2e",
      "value": "100%"
     }
    },
    "959b2741caa14a4887d764761df12548": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "9c56afe1c1a449e4a6ca5f2e5ff19962": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_677e67e37cae4f4993bcfb90b64ea414",
      "placeholder": "​",
      "style": "IPY_MODEL_fbea06001a724a348d8f6591648c1cc8",
      "value": "100%"
     }
    },
    "9eec49ded4394f89961493b913ca2cbb": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_410a75bfc02b478ea5ee00048a740ca9",
       "IPY_MODEL_cecb237fe40b4c8fa4c9da0d8d524d04",
       "IPY_MODEL_a31f97d00ef34e629fd30c2c2389e878"
      ],
      "layout": "IPY_MODEL_9148ddb7d90b4b64bc570d6a687369c4"
     }
    },
    "a170d8c8cf7f49e8aa86520bc406fbed": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "a20279445b4c4c72a73779f54da3ae67": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_e396b657844b4593a301f78f9bc67b63",
      "placeholder": "​",
      "style": "IPY_MODEL_959b2741caa14a4887d764761df12548",
      "value": "100%"
     }
    },
    "a31f97d00ef34e629fd30c2c2389e878": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_21494dff3dc54b938bb3c4af7bddc2bf",
      "placeholder": "​",
      "style": "IPY_MODEL_32a6171162984f68b7e465a3d12f3494",
      "value": " 256/256 [05:02&lt;00:00,  1.09it/s]"
     }
    },
    "a849e25a234b4580b06b496bef8e00f2": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_0001d478c10e4e0793b0847fddb5b61f",
      "max": 305,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_53615d583f7642a9bbe594454d26d54e",
      "value": 305
     }
    },
    "a92d7ad7d07949a09e6de9057cb27e4f": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "a961064cabca48edb4f50d91ff1b386d": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_9348543a8cd84980809203081c777509",
       "IPY_MODEL_64483751c11e4e4fbde1e1a3e6391eae",
       "IPY_MODEL_0ed49d6fc73f4ee7930604ded0487c23"
      ],
      "layout": "IPY_MODEL_0579c1ffe3734c3e87b041dfc89a5048"
     }
    },
    "bc95e906d0a749479b881f0747e5ad23": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "ProgressStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "ProgressStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "bar_color": null,
      "description_width": ""
     }
    },
    "c3840367809d463d9c694ef2a92e0ff5": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_57a3b742ef2d47039cd217b8a4557bb4",
      "placeholder": "​",
      "style": "IPY_MODEL_2b8a7b88de8243758573144ed63aadd2",
      "value": "100%"
     }
    },
    "cecb237fe40b4c8fa4c9da0d8d524d04": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_e3bf145254de4fe4b6fc42acb4c55bd1",
      "max": 256,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_5c890b497a194ffaa5ff64c296f3d745",
      "value": 256
     }
    },
    "e1fed3d156a14076b5c1a47f75be79df": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "e396b657844b4593a301f78f9bc67b63": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "e3bf145254de4fe4b6fc42acb4c55bd1": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "e4493369d8f843b69d0e962a071f460a": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "e6fa20e8a7bf4d0fbfbd82add6387f3c": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "e7df0eeb8b364496b832d1bb5b89f854": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "e9c1c09b268a4fe18a535c14d36afa85": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "ebd580561949448e8d0ccf5ba4453ff9": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "FloatProgressModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "FloatProgressModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "ProgressView",
      "bar_style": "success",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_3484006847c245b78e3397f32c30ffed",
      "max": 305,
      "min": 0,
      "orientation": "horizontal",
      "style": "IPY_MODEL_583a4136a3784ca6a35950627b684fc6",
      "value": 305
     }
    },
    "ecec72f3253d4406809e24bf55bb7b51": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "f1654d6e363b450da33f4e1807b19ece": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HBoxModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HBoxModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HBoxView",
      "box_style": "",
      "children": [
       "IPY_MODEL_9c56afe1c1a449e4a6ca5f2e5ff19962",
       "IPY_MODEL_a849e25a234b4580b06b496bef8e00f2",
       "IPY_MODEL_62328e01c5744b239c4fde0535389ab4"
      ],
      "layout": "IPY_MODEL_e6fa20e8a7bf4d0fbfbd82add6387f3c"
     }
    },
    "f757b7edd0df4cbfb3513fe595663e08": {
     "model_module": "@jupyter-widgets/base",
     "model_module_version": "1.2.0",
     "model_name": "LayoutModel",
     "state": {
      "_model_module": "@jupyter-widgets/base",
      "_model_module_version": "1.2.0",
      "_model_name": "LayoutModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "LayoutView",
      "align_content": null,
      "align_items": null,
      "align_self": null,
      "border": null,
      "bottom": null,
      "display": null,
      "flex": null,
      "flex_flow": null,
      "grid_area": null,
      "grid_auto_columns": null,
      "grid_auto_flow": null,
      "grid_auto_rows": null,
      "grid_column": null,
      "grid_gap": null,
      "grid_row": null,
      "grid_template_areas": null,
      "grid_template_columns": null,
      "grid_template_rows": null,
      "height": null,
      "justify_content": null,
      "justify_items": null,
      "left": null,
      "margin": null,
      "max_height": null,
      "max_width": null,
      "min_height": null,
      "min_width": null,
      "object_fit": null,
      "object_position": null,
      "order": null,
      "overflow": null,
      "overflow_x": null,
      "overflow_y": null,
      "padding": null,
      "right": null,
      "top": null,
      "visibility": null,
      "width": null
     }
    },
    "f9d7d05478bf4cee998fe3c38a0069c6": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "fbea06001a724a348d8f6591648c1cc8": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "DescriptionStyleModel",
     "state": {
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "DescriptionStyleModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/base",
      "_view_module_version": "1.2.0",
      "_view_name": "StyleView",
      "description_width": ""
     }
    },
    "ff9a01ae87124d55a125bb0a31982a8f": {
     "model_module": "@jupyter-widgets/controls",
     "model_module_version": "1.5.0",
     "model_name": "HTMLModel",
     "state": {
      "_dom_classes": [],
      "_model_module": "@jupyter-widgets/controls",
      "_model_module_version": "1.5.0",
      "_model_name": "HTMLModel",
      "_view_count": null,
      "_view_module": "@jupyter-widgets/controls",
      "_view_module_version": "1.5.0",
      "_view_name": "HTMLView",
      "description": "",
      "description_tooltip": null,
      "layout": "IPY_MODEL_674cdb8623bd4e31b3732cc67afc2fd0",
      "placeholder": "​",
      "style": "IPY_MODEL_e4493369d8f843b69d0e962a071f460a",
      "value": " 305/305 [21:10&lt;00:00,  3.85s/it]"
     }
    }
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}