2-fs1_10ml_metrics.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c7340c89-8eae-4df5-ac5c-dc9e545c2b0d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# -*- coding: utf-8 -*-\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import lightgbm as lgb\n",
    "from sklearn import metrics\n",
    "from sklearn.model_selection import train_test_split\n",
    "#import shap\n",
    "import pickle\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import StratifiedKFold\n",
    "from sklearn.svm import LinearSVC, SVC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "import xgboost\n",
    "import os\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.naive_bayes import GaussianNB\n",
    "from sklearn.metrics import precision_recall_fscore_support\n",
    "from sklearn.neural_network import MLPClassifier\n",
    "import catboost\n",
    "from catboost import *\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "3e8ec8b9-ebef-42fb-a54a-062edcc16e25",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load and process data function\n",
    "def dataloader(filepath):\n",
    "    raw_data = pd.read_csv(filepath)\n",
    "    med_columns = raw_data.columns[raw_data.columns.str.contains('Med_')]\n",
    "    raw_data[med_columns] = raw_data[med_columns].fillna(0)\n",
    "    columns_to_rename = raw_data.columns[65:273]\n",
    "    new_column_names = [col.rsplit('_', 1)[0] for col in columns_to_rename]\n",
    "    raw_data.rename(columns=dict(zip(columns_to_rename, new_column_names)), inplace=True)\n",
    "    filtered_df = raw_data[raw_data['Is transferred to ICU'] == 1].drop('Is transferred to ICU', axis=1, errors='ignore')\n",
    "    transform_dtypes_data = filtered_df.copy().iloc[:, 2:]\n",
    "    category_columns_1 = transform_dtypes_data.columns[\n",
    "        transform_dtypes_data.columns.str.contains('Med_') | transform_dtypes_data.columns.str.startswith('Diagnosis_')\n",
    "    ]\n",
    "    transform_dtypes_data[category_columns_1] = transform_dtypes_data[category_columns_1].applymap(lambda x: 0 if x == 0.0 else 1)\n",
    "    category_columns_2 = ['Gender', 'Insurance', 'Race', 'triage_Acuity', 'Arrival transport']\n",
    "    transform_dtypes_data[category_columns_1] = transform_dtypes_data[category_columns_1].astype('category')\n",
    "    transform_dtypes_data[category_columns_2] = transform_dtypes_data[category_columns_2].astype('category')\n",
    "    # print(transform_dtypes_data.dtypes)\n",
    "\n",
    "    def remove_outliers(df):\n",
    "        for column in df.columns:\n",
    "            if pd.api.types.is_numeric_dtype(df[column]):\n",
    "                lower_bound = df[column].quantile(0.01)\n",
    "                upper_bound = df[column].quantile(0.99)\n",
    "                df[column] = df[column].apply(lambda v: lower_bound if v < lower_bound else upper_bound if v > upper_bound else v)\n",
    "            if column == 'Age':\n",
    "                df[column] = df[column].apply(lambda v: 91 if v > 91 else v)\n",
    "        return df\n",
    "    \n",
    "    data = remove_outliers(transform_dtypes_data)\n",
    "    X = data.iloc[:, :-1]\n",
    "    y = data.iloc[:, -1].squeeze()  # Ensure y is a 1D Series\n",
    "    return X, y\n",
    "\n",
    "# Feature selection function\n",
    "def select_features(X, selected_features):\n",
    "    print(f'Selected feature count: {len(selected_features)}')\n",
    "    return X[selected_features]\n",
    "\n",
    "\n",
    "def fillna_mean(X):\n",
    "    for column in list(X.columns[X.isnull().sum()>0]):\n",
    "        mean_val = X[column].mean()\n",
    "        X[column].fillna(mean_val,inplace=True)\n",
    "    return X\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "62661672-1605-456b-8199-2646fc6d1b48",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Selected feature count: 168\n"
     ]
    }
   ],
   "source": [
    "cd = '/Users/hnwang/bmc/ihm_icu/' \n",
    "search_path = cd + '/fs_corr_step1/'\n",
    "save_path = cd +'/fs_imp_step2-1/ML_models_compare/'\n",
    "\n",
    "raw_data_path = '/Users/hnwang/ihm/without_fillna/alldata.csv'\n",
    "X, y = dataloader(raw_data_path)\n",
    "\n",
    "fs1 = pd.read_excel('/Users/hnwang/bmc/ihm_icu/fs_corr_step1' + '/significantPvalue_ihm.xlsx') #encoding_errors='ignore',,encoding=\"utf-8\"\n",
    "selected_features=fs1['ID'].tolist()\n",
    "X_fs1 = select_features(X, selected_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "ff7490be-8439-42a3-a972-10a8b1deb0a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.7725244932795893\n",
      "auc_te =  0.7609568277557032\n",
      "auc_te =  0.7664412185197403\n",
      "auc_te =  0.7754352368520456\n",
      "auc_te =  0.7589629385241394\n",
      "auc_te =  0.78235886180961\n",
      "auc_te =  0.7775307331487671\n",
      "auc_te =  0.7813235198894909\n",
      "auc_te =  0.7697630089185562\n",
      "auc_te =  0.7729965535962422\n",
      "====Logistic Regression model: meanAUC=0.7718293392293883\n",
      "====Logistic Regression model: 95% CI=[0.7594115636012413, 0.7821259098775832]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_LR(X,y):\n",
    "    #from sklearn.linear_model import LogisticRegression\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    X = fillna_mean(X)\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    model = LogisticRegression(max_iter=500)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====Logistic Regression model: meanAUC={}'.format(meanAUC))\n",
    "    print('====Logistic Regression model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====Logistic Regression model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====Logistic Regression model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/LR/')==False:\n",
    "        os.makedirs(save_path+'/LR/')\n",
    "    predict_result.to_csv(save_path+'/LR/predict_result_LR.csv',index=False)    \n",
    "    return predict_result  \n",
    "\n",
    "predict_result = model_10cv_LR(X_fs1, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1c29ab4a-b635-4aee-ae25-c16ae7a57ce0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.6837653426668387\n",
      "auc_te =  0.6452981031754\n",
      "auc_te =  0.6687179340056134\n",
      "auc_te =  0.6634475157670258\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_SVM(X,y):\n",
    "    #from sklearn.svm import LinearSVC\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    X = fillna_mean(X)\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    '''\n",
    "    model = LinearSVC(penalty='l2', loss='squared_hinge', \n",
    "                dual=False, tol=0.0001, C=2.0, multi_class='ovr', \n",
    "                fit_intercept=True, intercept_scaling=1, class_weight='balanced', \n",
    "                verbose=0, random_state=0, max_iter=100)\n",
    "    '''\n",
    "    # model = LinearSVC(max_iter=500)\n",
    "    model = SVC(probability=True)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        '''\n",
    "        since 'LinearSVC' object has no attribute 'predict_proba'\n",
    "        we use its decision function (predict confidence scores for samples)\n",
    "        '''\n",
    "        # prob_pos = model.decision_function(X_test)\n",
    "        # y_tepred = (prob_pos - prob_pos.min()) / (prob_pos.max() - prob_pos.min())\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====SVM model: meanAUC={}'.format(meanAUC))\n",
    "    print('====SVM model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====SVM model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====SVM model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/SVM/')==False:\n",
    "        os.makedirs(save_path+'/SVM/')\n",
    "    predict_result.to_csv(save_path+'/SVM/predict_result_SVM.csv',index=False)    \n",
    "    return predict_result    \n",
    "\n",
    "predict_result = model_10cv_SVM(X_fs1, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "69becaea-83c0-4233-8d41-edde743bd3b5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.7091922118782246\n",
      "auc_te =  0.7074714663453492\n",
      "auc_te =  0.6720129913598134\n",
      "auc_te =  0.7082950812617446\n",
      "auc_te =  0.675406667336078\n",
      "auc_te =  0.7190259827196266\n",
      "auc_te =  0.7040377452316402\n",
      "auc_te =  0.7089151971682028\n",
      "auc_te =  0.7311625693286807\n",
      "auc_te =  0.6859300148044687\n",
      "====Decision Tree Classifier model: meanAUC=0.7021449927433829\n",
      "====Decision Tree Classifier model: 95% CI=[0.672776568454473, 0.7284318373416435]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_DTC(X,y):\n",
    "    #from sklearn.tree import DecisionTreeClassifier\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    #import graphviz\n",
    "    X = fillna_mean(X)\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    model = DecisionTreeClassifier(min_samples_split=20, min_samples_leaf=50)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====Decision Tree Classifier model: meanAUC={}'.format(meanAUC))\n",
    "    print('====Decision Tree Classifier model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====Decision Tree Classifier model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====Decision Tree Classifier model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/DTC/')==False:\n",
    "        os.makedirs(save_path+'/DTC/')\n",
    "    predict_result.to_csv(save_path+'/DTC/predict_result_DTC.csv',index=False)    \n",
    "    return predict_result  \n",
    "\n",
    "predict_result = model_10cv_DTC(X_fs1, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "74fce9c3-c109-448f-953f-50904e997562",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.6926481909160894\n",
      "auc_te =  0.6737170016113686\n",
      "auc_te =  0.6606061040743237\n",
      "auc_te =  0.6994630365450729\n",
      "auc_te =  0.6542939145735611\n",
      "auc_te =  0.6843468936927115\n",
      "auc_te =  0.6960825097184169\n",
      "auc_te =  0.6847209331884817\n",
      "auc_te =  0.6823718002377325\n",
      "auc_te =  0.6979045338984065\n",
      "====KNN Classifier model: meanAUC=0.6826154918456165\n",
      "====KNN Classifier model: 95% CI=[0.6557141572112327, 0.699112373449573]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_KNN(X,y):\n",
    "    #from sklearn.neighbors import KNeighborsClassifier\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    X = fillna_mean(X)\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    model = KNeighborsClassifier(60)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====KNN Classifier model: meanAUC={}'.format(meanAUC))\n",
    "    print('====KNN Classifier model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====KNN Classifier model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====KNN Classifier model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/KNN/')==False:\n",
    "        os.makedirs(save_path+'/KNN/')\n",
    "    predict_result.to_csv(save_path+'/KNN/predict_result_KNN.csv',index=False)    \n",
    "    return predict_result  \n",
    "\n",
    "predict_result = model_10cv_KNN(X_fs1, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "4ca9c75e-5d24-45d3-9f99-53d15dc45ea2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.7129821554086039\n",
      "auc_te =  0.6997624785429786\n",
      "auc_te =  0.6929577936205108\n",
      "auc_te =  0.725258679634119\n",
      "auc_te =  0.6842936994056584\n",
      "auc_te =  0.7075653034584654\n",
      "auc_te =  0.7166764687600111\n",
      "auc_te =  0.7052579386491885\n",
      "auc_te =  0.7219103743688966\n",
      "auc_te =  0.6918195146334638\n",
      "====Naive Bayes model: meanAUC=0.7058484406481895\n",
      "====Naive Bayes model: 95% CI=[0.6859870078319147, 0.724505310949444]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_NB(X,y):\n",
    "    #from sklearn.naive_bayes import GaussianNB\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    X = fillna_mean(X)\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    model = GaussianNB(var_smoothing=1e-8)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====Naive Bayes model: meanAUC={}'.format(meanAUC))\n",
    "    print('====Naive Bayes model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====Naive Bayes model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====Naive Bayes model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/NB/')==False:\n",
    "        os.makedirs(save_path+'/NB/')\n",
    "    predict_result.to_csv(save_path+'/NB/predict_result_NB.csv',index=False)    \n",
    "    return predict_result  \n",
    "\n",
    "predict_result = model_10cv_NB(X_fs1, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "1094d09c-a61a-4190-a0bf-2c3d41f19b64",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.7847615700562786\n",
      "auc_te =  0.735333916677425\n",
      "auc_te =  0.7397233418922343\n",
      "auc_te =  0.7296475071603097\n",
      "auc_te =  0.7412833091867133\n",
      "auc_te =  0.7719423445650979\n",
      "auc_te =  0.7846025848837377\n",
      "auc_te =  0.7639341549574439\n",
      "auc_te =  0.7756094585965173\n",
      "auc_te =  0.7580294510866575\n",
      "====neural network MLP Classifier model: meanAUC=0.7584867639062415\n",
      "====neural network MLP Classifier model: 95% CI=[0.7309269493016606, 0.7847257983924568]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_MLP(X,y):\n",
    "    #from sklearn.neural_network import MLPClassifier\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    X = fillna_mean(X)\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    #model = MLPClassifier(solver='adam',alpha=1e-5,activation='logisic', hidden_layer_sizes=(10,2),max_iter=2000,tol=1e-4) #solver{｡ｮlbfgs｡ｯ, ｡ｮsgd｡ｯ, ｡ｮadam｡ｯ}\n",
    "    model = MLPClassifier(max_iter=500)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====neural network MLP Classifier model: meanAUC={}'.format(meanAUC))\n",
    "    print('====neural network MLP Classifier model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====neural network MLP Classifier model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====neural network MLP Classifier model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/MLP/')==False:\n",
    "        os.makedirs(save_path+'/MLP/')\n",
    "    predict_result.to_csv(save_path+'/MLP/predict_result_MLP.csv',index=False)    \n",
    "    return predict_result  \n",
    "\n",
    "predict_result = model_10cv_MLP(X_fs1, y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "0d009746-492e-40d6-b690-9d4c59f44b84",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.7911753666221986\n",
      "auc_te =  0.7651514064808572\n",
      "auc_te =  0.764876469716313\n",
      "auc_te =  0.7803583980032419\n",
      "auc_te =  0.7459452804115924\n",
      "auc_te =  0.7842362017605516\n",
      "auc_te =  0.7712884732163776\n",
      "auc_te =  0.7832330884768777\n",
      "auc_te =  0.7820711666830097\n",
      "auc_te =  0.7785566001860722\n",
      "====Random Forest model: meanAUC=0.7746892451557092\n",
      "====Random Forest model: 95% CI=[0.7502047980051545, 0.789614054528328]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_RF(X,y):\n",
    "    #from sklearn.ensemble import RandomForestClassifier\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    model = RandomForestClassifier(n_estimators=500)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====Random Forest model: meanAUC={}'.format(meanAUC))\n",
    "    print('====Random Forest model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====Random Forest model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====Random Forest model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/RF/')==False:\n",
    "        os.makedirs(save_path+'/RF/')\n",
    "    predict_result.to_csv(save_path+'/RF/predict_result_RF.csv',index=False)      \n",
    "    return predict_result    \n",
    "\n",
    "predict_result = model_10cv_RF(X_fs1, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "e0e1ee65-da7b-4052-9b0b-178d7ac53ea4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.8072824772041561\n",
      "auc_te =  0.7799681073353129\n",
      "auc_te =  0.7855092785181147\n",
      "auc_te =  0.8038350092522198\n",
      "auc_te =  0.7762624379001526\n",
      "auc_te =  0.8116025705392109\n",
      "auc_te =  0.8023252478017013\n",
      "auc_te =  0.8026207554816909\n",
      "auc_te =  0.7986891229010062\n",
      "auc_te =  0.7951805355917123\n",
      "====XGBoost model: meanAUC=0.7963275542525279\n",
      "====XGBoost model: 95% CI=[0.7770962135230637, 0.8106305495388236]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_XGBoost(X,y):\n",
    "    #import xgboost\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    skf = StratifiedKFold(n_splits=10)#, random_state=0\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        D_train = xgboost.DMatrix(X_train,label=y_train, enable_categorical=True)\n",
    "        D_test = xgboost.DMatrix(X_test,label=y_test, enable_categorical=True)\n",
    "        ###\n",
    "        params = {\n",
    "            \"eta\": 0.01,\n",
    "            \"booster\": \"gbtree\",\n",
    "            \"objective\": \"binary:logistic\",\n",
    "            \"eval_metric\":\"auc\",\n",
    "            \"min_child_weight\": 150,\n",
    "        }\n",
    "        steps = 1000 \n",
    "        model = xgboost.train(params, D_train, steps, verbose_eval=False)#\n",
    "        y_tepred = model.predict(D_test)\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====XGBoost model: meanAUC={}'.format(meanAUC))\n",
    "    print('====XGBoost model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====XGBoost model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====XGBoost model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/XGBoost/')==False:\n",
    "        os.makedirs(save_path+'/XGBoost/')\n",
    "    predict_result.to_csv(save_path+'/XGBoost/predict_result_XGBoost.csv',index=False)    \n",
    "    return predict_result    \n",
    "\n",
    "\n",
    "predict_result = model_10cv_XGBoost(X_fs1, y)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "22d925ac-c271-4abe-a172-5147450152b9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.8086631379130628\n",
      "auc_te =  0.7899734148102457\n",
      "auc_te =  0.788855737093512\n",
      "auc_te =  0.8077008592371581\n",
      "auc_te =  0.7791122172335144\n",
      "auc_te =  0.821256437106422\n",
      "auc_te =  0.8048881365981476\n",
      "auc_te =  0.8042803114785155\n",
      "auc_te =  0.8017540544871411\n",
      "auc_te =  0.7940564483147061\n",
      "====LightGBM model: meanAUC=0.8000540754272425\n",
      "====LightGBM model: 95% CI=[0.7813045092020139, 0.8184229447879162]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_LightGBM(X,y):\n",
    "    #import lightgbm as lgb\n",
    "    #from sklearn.model_selection import StratifiedKFold\n",
    "    #from sklearn import metrics\n",
    "    skf = StratifiedKFold(n_splits=10)#, random_state=0\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        d_train = lgb.Dataset(X_train,label=y_train)\n",
    "        d_test = lgb.Dataset(X_test,label=y_test)\n",
    "        ###\n",
    "        params = {\n",
    "            \"learning_rate\": 0.01,\n",
    "            \"boosting_type\": \"gbdt\",\n",
    "            \"objective\": \"binary\",\n",
    "            \"metric\": ['auc', 'binary_logloss'],\n",
    "            'max_depth': 5,\n",
    "            'verbosity': -1,\n",
    "        }\n",
    "        model = lgb.train(params, d_train, 1000)\n",
    "        y_tepred = model.predict(X_test)\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====LightGBM model: meanAUC={}'.format(meanAUC))\n",
    "    print('====LightGBM model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====LightGBM model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====LightGBM model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/LightGBM/')==False:\n",
    "        os.makedirs(save_path+'/LightGBM/')\n",
    "    predict_result.to_csv(save_path+'/LightGBM/predict_result_LightGBM.csv',index=False)    \n",
    "    return predict_result    \n",
    "\n",
    "predict_result = model_10cv_LightGBM(X_fs1, y)  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "9fa412e8-abb9-422c-84ff-29653c4073db",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "auc_te =  0.8057081653828316\n",
      "auc_te =  0.7870602804833149\n",
      "auc_te =  0.787954422656702\n",
      "auc_te =  0.8095859691401412\n",
      "auc_te =  0.776427399958879\n",
      "auc_te =  0.8103701366077107\n",
      "auc_te =  0.8057727157536374\n",
      "auc_te =  0.8048187078965606\n",
      "auc_te =  0.8030240339426525\n",
      "auc_te =  0.7942525656694178\n",
      "====Catboost model: meanAUC=0.7984974397491849\n",
      "====Catboost model: 95% CI=[0.7788197980768771, 0.8101936989275076]\n"
     ]
    }
   ],
   "source": [
    "def model_10cv_Catboost(X,y):\n",
    "    #import catboost\n",
    "    #from catboost import *\n",
    "    #from sklearn import metrics\n",
    "    skf = StratifiedKFold(n_splits=10)\n",
    "    cat_features = X.select_dtypes(['category']).columns.tolist()\n",
    "    model = CatBoostClassifier(iterations=1000, learning_rate=0.01, cat_features=cat_features, verbose=0)\n",
    "    ## note X,y is dataFrame, not array, so .split(X,y) is not right\n",
    "    t = y.values\n",
    "    allAUC = []\n",
    "    allY = []\n",
    "    allPY = []\n",
    "    for train_index, test_index in skf.split(np.zeros(len(t)),t):\n",
    "        X_train, X_test = X.iloc[train_index,:], X.iloc[test_index,:]\n",
    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
    "        model.fit(X_train,y_train)\n",
    "        probs = model.predict_proba(X_test)\n",
    "        y_tepred = probs[:,1]\n",
    "        auc_te = metrics.roc_auc_score(y_test, y_tepred)        \n",
    "        print('auc_te = ', auc_te)\n",
    "        allAUC.append(auc_te)\n",
    "        allY.extend(y_test)\n",
    "        allPY.extend(y_tepred)\n",
    "    meanAUC = np.mean(allAUC)\n",
    "    print('====Catboost model: meanAUC={}'.format(meanAUC))\n",
    "    print('====Catboost model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5))) \n",
    "    # print('====Catboost model: meanAUC={}'.format(meanAUC),file=log)\n",
    "    # print('====Catboost model: 95% CI=[{}, {}]'.format(np.percentile(allAUC,2.5), np.percentile(allAUC,97.5)),file=log)\n",
    "    predict_result = pd.DataFrame({'y_test':allY,'y_tepred':allPY}) \n",
    "    if os.path.exists(save_path+'/Catboost/')==False:\n",
    "        os.makedirs(save_path+'/Catboost/')\n",
    "    predict_result.to_csv(save_path+'/Catboost/predict_result_Catboost.csv',index=False)      \n",
    "    return predict_result    \n",
    "\n",
    "predict_result = model_10cv_Catboost(X_fs1, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ea0ebdb1-c136-4bb3-a446-063b95944619",
   "metadata": {},
   "source": [
    "### 绘制结果图"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "e0ca56e9-e0dc-4004-acfa-65d577279fd3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import pickle\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from sklearn import metrics\n",
    "import itertools\n",
    "#import plot_roc_pr_calibration_curves # this is myself function\n",
    "from joblib import Parallel, delayed\n",
    "#from function_Delong_test import DelongTest\n",
    "\n",
    "cd = '/Users/hnwang/bmc/ihm_icu/fs_imp_step2-1' \n",
    "search_path = cd + '/ML_models_compare/'\n",
    "save_path = cd + '/prediction_figure/'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "743c14a7-db3b-4d89-b53d-68f04428362a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "ce151014-f50b-47dd-9971-888e24bdce3f",
   "metadata": {},
   "outputs": [],
   "source": [
    "## evaluation\n",
    "from matplotlib import pyplot as plt\n",
    "import matplotlib\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn import metrics, utils\n",
    "from joblib import Parallel, delayed\n",
    "from tqdm import tqdm\n",
    "import scipy\n",
    "import scipy.special\n",
    "import itertools\n",
    "import yaml\n",
    "import pickle\n",
    "\n",
    "##############\n",
    "#models = ['LR', 'RF', 'CNN_V3', 'RNN_V2', ]\n",
    "#colors = ['#2ca02c', '#d62728', '#1f77b4', '#ff7f0e']\n",
    "fig_size = 3\n",
    "q=10\n",
    "scipy.interp = np.interp\n",
    "#################### function only the curve #################\n",
    "\n",
    "def get_roc_curve(y_true,y_pred):\n",
    "    roc_curve = metrics.roc_curve(y_true, y_pred)\n",
    "    auc = metrics.roc_auc_score(y_true, y_pred)\n",
    "    return roc_curve, auc\n",
    "\n",
    "\n",
    "def get_pr_curve(y_true,y_pred):\n",
    "    pr_curve = metrics.precision_recall_curve(y_true, y_pred)\n",
    "    return pr_curve\n",
    "\n",
    "\n",
    "def get_calibration_curve(y_true,y_pred,q):\n",
    "    def get_calibration_plot(y_true, y_prob, q):\n",
    "        try:\n",
    "            out, bins = pd.qcut(y_prob.to_numpy(), q=q, labels=False, retbins=True)\n",
    "            #bins = np.percentile(y_prob.to_numpy(), [0, 20, 40, 60, 80, 100])\n",
    "            P_true, P_pred = [], []\n",
    "            for bin_idx in range(q):\n",
    "                m = (out == bin_idx)\n",
    "                P_true.append(np.mean(y_true[m]))\n",
    "                P_pred.append(np.mean(y_prob[m]))\n",
    "        except:\n",
    "            print(q)\n",
    "            print(out)\n",
    "            raise\n",
    "        return P_true, P_pred\n",
    "\n",
    "    p_true, p_pred = get_calibration_plot(y_true, y_pred, q)\n",
    "    brier_score = np.mean(np.square(np.array(p_true) - np.array(p_pred)))\n",
    "    print('brier_score = %.3f'%brier_score) \n",
    "    cali_curve = [p_true, p_pred]      \n",
    "    return cali_curve, brier_score \n",
    "\n",
    "################ function  the curve and 95%CI ###################\n",
    "#======= ROC curves\n",
    "def get_roc_CI_curve(y_true,y_pred):\n",
    "\n",
    "    def get_roc_curve(y_true,y_pred):\n",
    "        roc_curve = metrics.roc_curve(y_true, y_pred)\n",
    "        auc = metrics.roc_auc_score(y_true, y_pred)\n",
    "        return roc_curve, auc\n",
    "    \n",
    "    def bootstrap_func(i,y_true, y_pred):\n",
    "        y_true_i, y_pred_i = utils.resample(y_true,y_pred,replace=True, random_state=i)\n",
    "        roc_curve_i = metrics.roc_curve(y_true_i,y_pred_i)\n",
    "        auc_i = metrics.roc_auc_score(y_true_i, y_pred_i)\n",
    "        return roc_curve_i, auc_i\n",
    "    \n",
    "    def get_roc_CI(y_true, y_pred):\n",
    "        roc_curves, auc_scores = zip(*Parallel(n_jobs=-1)(delayed(bootstrap_func)(i,y_true,y_pred) for i in range(1000)))\n",
    "        #roc_curves, auc_scores = [], []\n",
    "        #for i in range(1000):\n",
    "        #    roc_curve_i, auc_i = bootstrap_func(i,y_true,y_pred)\n",
    "        #    roc_curves.append(roc_curve_i)\n",
    "        #    auc_scores.append(auc_i)\n",
    "        print('Test AUC: {:.3f}'.format(metrics.roc_auc_score(y_true,y_pred)))\n",
    "        print('Test AUC:({:.3f},{:.3f}) percentile 95% CI'.format(np.percentile(auc_scores, 2.5), np.percentile(auc_scores, 97.5)))\n",
    "        #print('Test AUC: {.3f} ({:.3f},{:.3f}) percentile'.format(np.median(auc_scores), np.percentile(auc_scores,2.5),np.percentile(auc_scores,97.5)))\n",
    "        #print('Test AUC: {.3f} ({:.3f},{:.3f}) empirical'.format(np.mean(auc_scores),np.mean(auc_scores)-1.96 * np.std(auc_scores),np.mean(auc_scores)+1.96 * np.std(auc_scores)))\n",
    "        \n",
    "        uci = np.percentile(auc_scores, 97.5)\n",
    "        dci = np.percentile(auc_scores, 2.5)\n",
    "        \n",
    "        tprs = []\n",
    "        mean_fpr = np.linspace(0,1,100)\n",
    "        for fpr, tpr,_ in roc_curves:\n",
    "            tprs.append(scipy.interp(mean_fpr,fpr, tpr)) # np.interp ?\n",
    "            tprs[-1][0] = 0.0 #?｣ｿ\n",
    "        \n",
    "        mean_tpr = np.mean(tprs, axis=0)\n",
    "        std_tpr = np.std(tprs, axis=0)\n",
    "        tprs_upper = np.minimum(mean_tpr+1.96*std_tpr,1)\n",
    "        tprs_lower = np.maximum(mean_tpr-1.96*std_tpr,0)\n",
    "        return uci,dci,auc_scores, mean_fpr, tprs_lower, tprs_upper\n",
    "    \n",
    "        \n",
    "    roc_curve, auc = get_roc_curve(y_true,y_pred)\n",
    "    uci,dci,auc_scores, mean_fpr, tprs_lower, tprs_upper = get_roc_CI(y_true, y_pred)\n",
    "   \n",
    "    ## save\n",
    "    '''\n",
    "    result_roc_CI = {'roc_curve':roc_curve,'auc':auc,'dci':dci,'uci':uci,'auc_scores':auc_scores,\n",
    "              'mean_fpr':mean_fpr,'tprs_lower':tprs_lower,'tprs_upper':tprs_upper}\n",
    "    f = open(save_path+'/result_roc_CI.pkl','rb')\n",
    "    pickle.dump(result_roc_CI, f)\n",
    "    f.close()\n",
    "    '''  \n",
    "    return roc_curve, auc, uci,dci, auc_scores, mean_fpr, tprs_lower, tprs_upper\n",
    "\n",
    "\n",
    "\n",
    "    \n",
    "def plot_roc_CI(y_true,y_pred_1,ypred_2):\n",
    "    fig, ax = plt.subplots(figsize=(fig_size,fig_size))\n",
    "    #fpr,tpr,thresholds = metrics.roc_curve(y_true,y_score)\n",
    "    # for model_1\n",
    "    roc_curve, auc_1, uci,dci, auc_scores_1, mean_fpr, tprs_lower, tprs_upper = get_roc_CI_curve(y_true,y_pred_1)\n",
    "    fpr = roc_curve[0];tpr = roc_curve[1]\n",
    "    plt.plot(fpr,tpr,lw=1.25, label='model_1', color='green')\n",
    "    plt.fill_between(mean_fpr,tprs_lower,tprs_upper, alpha=0.1, color='green')\n",
    "\n",
    "    # for model_2\n",
    "    roc_curve, auc_2, uci,dci,auc_scores_2, mean_fpr, tprs_lower, tprs_upper = get_roc_CI_curve(y_true,y_pred_2)\n",
    "    fpr = roc_curve[0], tpr = roc_curve[1]\n",
    "    plt.plot(fpr,tpr,lw=1.25, label='model_2', color='blue')\n",
    "    plt.fill_between(mean_fpr,tprs_lower,tprs_upper, alpha=0.1, color='blue')\n",
    "    \n",
    "    \n",
    "    # statistical tests\n",
    "    pvalue = 2 * min((auc_scores_1 < auc_scores_2).mean(), (auc_scores_1 > auc_scores_2).mean())\n",
    "    display(pvalue)\n",
    "    \n",
    "    \n",
    "    lg = plt.legend(\n",
    "        [\n",
    "            'model_1: {.3f}'.format(auc_1),\n",
    "            'model_2: {.3f}'.format(auc_2),\n",
    "        ],\n",
    "        title = 'AUROC',\n",
    "        loc = 'lower right',\n",
    "        handlelength = 1.0,\n",
    "        fontsize = 9,\n",
    "    )\n",
    "    lg.get_title().set_fontsize(10)\n",
    "    \n",
    "    plt.plot([0,1],[0,1],'k:')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.xlim(0,1)\n",
    "    plt.ylim(0,1)\n",
    "    #plt.title('{}.T={}.dt={}'.format(tast,T,dt))\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(save_path+'ROC_curve.png')\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "    \n",
    "################ function  the curve and 95%CI ###################\n",
    "#======= PR curves\n",
    "def get_pr_CI_curve(y_true,y_pred):\n",
    "\n",
    "    def get_pr_curve(y_true,y_pred):\n",
    "        pr_curve = metrics.precision_recall_curve(y_true, y_pred)\n",
    "        pr_auc = metrics.auc(*metrics.precision_recall_curve(y_true,y_pred)[1::-1])\n",
    "        return pr_curve, pr_auc\n",
    "    \n",
    "    def bootstrap_func(i,y_true, y_pred):\n",
    "        y_true_i, y_pred_i = utils.resample(y_true,y_pred,replace=True, random_state=i)\n",
    "        pr_curve_i = metrics.precision_recall_curve(y_true_i,y_pred_i)\n",
    "        pr_auc_i = metrics.auc(*metrics.precision_recall_curve(y_true_i,y_pred_i)[1::-1])\n",
    "        return pr_curve_i, pr_auc_i\n",
    "    \n",
    "    def get_pr_CI(y_true, y_pred):\n",
    "        pr_curves, pr_auc_scores = zip(*Parallel(n_jobs=-1)(delayed(bootstrap_func)(i,y_true,y_pred) for i in range(1000)))\n",
    "        #pr_curves, auc_scores = [], []\n",
    "        #for i in range(1000):\n",
    "        #    pr_curve_i, pr_auc_i = bootstrap_func(i,y_true,y_pred)\n",
    "        #    pr_curves.append(pr_curve_i)\n",
    "        #    pr_auc_scores.append(pr_auc_i)\n",
    "        print('Test AUPRC: {:.3f}'.format(metrics.auc(*metrics.precision_recall_curve(y_true,y_pred)[1::-1])))\n",
    "        print('Test AUPRC:({:.3f},{:.3f}) percentile 95% CI'.format(np.percentile(pr_auc_scores, 2.5), np.percentile(pr_auc_scores, 97.5)))\n",
    "        #print('Test AUPRC: {.3f} ({:.3f},{:.3f}) percentile'.format(np.median(pr_auc_scores), np.percentile(pr_auc_scores,2.5),np.percentile(pr_auc_scores,97.5)))\n",
    "        #print('Test AUPRC: {.3f} ({:.3f},{:.3f}) empirical'.format(np.mean(pr_auc_scores),np.mean(pr_auc_scores)-1.96 * np.std(pr_auc_scores),np.mean(pr_auc_scores)+1.96 * np.std(pr_auc_scores)))\n",
    "        \n",
    "        pr_uci = np.percentile(pr_auc_scores, 97.5)\n",
    "        pr_dci = np.percentile(pr_auc_scores, 2.5)\n",
    "        \n",
    "        precs = []\n",
    "        mean_rec = np.linspace(0,1,101)\n",
    "        for prec, rec,_ in pr_curves:\n",
    "            rec_sorted = rec[np.argsort(rec)]\n",
    "            prec_sorted = prec[np.argsort(rec)]\n",
    "            precs.append(np.interp(mean_rec,rec_sorted, prec_sorted))\n",
    "        \n",
    "        mean_prec = np.mean(precs, axis=0)\n",
    "        std_prec = np.std(precs, axis=0)\n",
    "        prec_upper = np.minimum(mean_prec+1.96*std_prec,1)\n",
    "        prec_lower = np.maximum(mean_prec-1.96*std_prec,0)\n",
    "        return pr_uci,pr_dci,pr_auc_scores, mean_rec, prec_lower, prec_upper\n",
    "    \n",
    "        \n",
    "    pr_curve, pr_auc = get_pr_curve(y_true,y_pred)\n",
    "    pr_uci,pr_dci,pr_auc_scores, mean_rec, prec_lower, prec_upper = get_pr_CI(y_true, y_pred)\n",
    "   \n",
    "    ## save\n",
    "    '''\n",
    "    result_pr_CI = {'pr_curve':pr_curve,'pr_auc':pr_auc,'pr_dci':pr_dci,'pr_uci':pr_uci,'pr_auc_scores':pr_auc_scores,\n",
    "              'mean_prec':mean_prec,'prec_lower':prec_lower,'prec_upper':prec_upper}\n",
    "    f = open(save_path+'/result_pr_CI.pkl','rb')\n",
    "    pickle.dump(result_pr_CI, f)\n",
    "    f.close()\n",
    "    '''  \n",
    "    return pr_curve, pr_auc, pr_uci,pr_dci, pr_auc_scores, mean_rec, prec_lower, prec_upper\n",
    "\n",
    "   \n",
    "def plot_pr_CI(y_true,y_pred_1,y_pred_2):\n",
    "    fig, ax = plt.subplots(figsize=(fig_size,fig_size))\n",
    "    #precision, recall, thresholds = metrics.precision_recall_curve(y_true,y_score)\n",
    "    # for model_1\n",
    "    pr_curve, pr_auc_1, pr_uci,pr_dci, pr_auc_scores_1, mean_prec, prec_lower, prec_uppe = get_pr_CI_curve(y_true,y_pred_1)\n",
    "    precison = pr_curve[0], recall = pr_curve[1]\n",
    "    plt.plot(recall,precision,lw=1.25, label='model_1', color='green')\n",
    "    plt.fill_between(mean_prec,prec_lower,prec_upper, alpha=0.1, color='green')\n",
    "\n",
    "    # for model_2\n",
    "    pr_curve, pr_auc_2, pr_uci,pr_dci, pr_auc_scores_2, mean_prec, prec_lower, prec_uppe = get_roc_CI_curve(y_true,y_pred)\n",
    "    precison = pr_curve[0]; recall = pr_curve[1]\n",
    "    plt.plot(recall,precision,lw=1.25, label='model_2', color='blue')\n",
    "    plt.fill_between(mean_fpr,tprs_lower,tprs_upper, alpha=0.1, color='blue')\n",
    "    \n",
    "    \n",
    "    # statistical tests\n",
    "    pvalue = 2 * min((pr_auc_scores_1 < pr_auc_scores_2).mean(), (pr_auc_scores_1 > pr_auc_scores_2).mean())\n",
    "    display(pvalue)\n",
    "    \n",
    "    \n",
    "    lg = plt.legend(\n",
    "        [\n",
    "            'model_1: {.3f}'.format(pr_auc_1),\n",
    "            'model_2: {.3f}'.format(pr_auc_2),\n",
    "        ],\n",
    "        title = 'AUPRC',\n",
    "        loc = 'lower left',\n",
    "        handlelength = 1.0,\n",
    "        fontsize = 9,\n",
    "    )\n",
    "    lg.get_title().set_fontsize(10)\n",
    "    \n",
    "    \n",
    "    plt.xlabel('Recall')\n",
    "    plt.ylabel('Precision')\n",
    "    plt.xlim(0,1)\n",
    "    plt.ylim(0,1)\n",
    "    #plt.title('{}.T={}.dt={}'.format(tast,T,dt))\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(save_path+'PR_curve.png')\n",
    "    plt.show()\n",
    "    \n",
    "\n",
    "\n",
    "################ function  the curve and 95%CI ###################\n",
    "#====Calibration\n",
    "    \n",
    "def get_calibration_CI_curve(y_true, y_pred,q):\n",
    "\n",
    "    def get_calibration_plot(y_true, y_pred, q):\n",
    "        try:\n",
    "            out, bins = pd.qcut(y_pred.to_numpy(), q=q, labels=False, retbins=True,duplicates='drop')\n",
    "            #bins = np.percentile(y_prob.to_numpy(), [0, 20, 40, 60, 80, 100])\n",
    "            P_true, P_pred = [], []\n",
    "            for bin_idx in range(q):\n",
    "                m = (out == bin_idx)\n",
    "                P_true.append(np.mean(y_true[m]))\n",
    "                P_pred.append(np.mean(y_pred[m]))\n",
    "        except:\n",
    "            print(q)\n",
    "            #print(out)\n",
    "            raise\n",
    "        return P_true, P_pred\n",
    "\n",
    "    def bootstrap_func(i,y_true,y_pred,q):\n",
    "        y_true_i, y_pred_i = utils.resample(y_true, y_pred, replace=True, random_state=i)\n",
    "        P_true, P_pred = get_calibration_plot(y_true_i, y_pred_i, q)\n",
    "        return P_true, P_pred\n",
    "        \n",
    "    def generate_calibration_CI_and_cloud(y_true,y_pred,q):\n",
    "        P_trues, P_preds = zip(*Parallel(n_jobs=-1)(delayed(bootstrap_func)(i,y_true,y_pred,q=q) for i in range(1000)))\n",
    "        return P_trues, P_preds\n",
    "    \n",
    "    \n",
    "    P_true, P_pred = get_calibration_plot(y_true, y_pred, q)\n",
    "    P_trues, P_preds = generate_calibration_CI_and_cloud(y_true,y_pred,q)\n",
    "    # save\n",
    "    '''\n",
    "    result_cali_CI = {'p_true': P_true,'P_pred':P_pred,'P_trues':P_trues,'P_preds':P_preds}\n",
    "    f = open(save_path+'/result_cali_CI.pkl','rb')\n",
    "    pickle.dump(result_cali_CI, f)\n",
    "    f.close()  \n",
    "    '''  \n",
    "    return P_true, P_pred, P_trues, P_preds\n",
    "    \n",
    "\n",
    "def plot_calibration_CI(y_true, y_pred,q):\n",
    "    fig = plt.figure(figsize=(figure_size,figure_size))\n",
    "    #brier_scores = []\n",
    "    P_true, P_pred, P_trues, P_preds = get_calibration_CI_curves(y_true, y_pred,q)\n",
    "    brier_score = np.mean(np.square(np.array(P_true)-np.array(P_pred)))\n",
    "    plt.plot(P_pred, P_true, '.', markersize=5, label=model, color='green',alpha=0.9)\n",
    "    yerrs = 1.96*np.std(P_trues,axis=0)\n",
    "    xerrs = 1.96*np.std(P_preds,axis=0)\n",
    "    markers, caps, bars = plt.errorbar(P_pred,P_true,yerrs,xerrs,'.',markersize=0,elinewidth=1,color='green')\n",
    "    for bar in bars: bar.set_alpha(0.5)\n",
    "    for cap in caps: cap.set_alpha(0.5)\n",
    "    \n",
    "    plt.xlim(0,1)\n",
    "    plt.ylim(0,1)\n",
    "    plt.xlabel('Mean predicted value')\n",
    "    plt.ylabel('Fraction of positives')\n",
    "    lg = plt.legend(\n",
    "        [' model_x:{:.4f}'.format(brier_score),\n",
    "        ],\n",
    "        loc='lower right',\n",
    "        handlelength=0.5,\n",
    "        bbox_to_anchor=(1.02,-0.02),\n",
    "        fontsize=9,\n",
    "        title='Brier score',\n",
    "    )\n",
    "    lg.get_title().set_fontsize(10)\n",
    "    plt.plot([0,1],[0,1],'k:')\n",
    "    plt.xticks([0,0.2,0.4,0.6,0.8,1.0])# for q=5\n",
    "    \n",
    "    #plt.title('{}.T={}.dt={}'.format(tast,T,dt))\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(save_path+'cali_curve.png')\n",
    "    plt.show()\n",
    "        \n",
    "    \n",
    "#################################################    \n",
    "        \n",
    "#brier_score = np.mean(np.square(np.array(P_true) - np.array(P_pred)))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "ce317e6f-bd0a-4b71-9661-7b1fb70068a6",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "==============model LR begin=============\n",
      "Test AUC: 0.770\n",
      "Test AUC:(0.762,0.779) percentile 95% CI\n",
      "Test AUPRC: 0.322\n",
      "Test AUPRC:(0.306,0.339) percentile 95% CI\n",
      "==============model LR over=============\n",
      "==============model SVM begin=============\n",
      "Test AUC: 0.640\n",
      "Test AUC:(0.630,0.651) percentile 95% CI\n",
      "Test AUPRC: 0.202\n",
      "Test AUPRC:(0.191,0.215) percentile 95% CI\n",
      "==============model SVM over=============\n",
      "==============model KNN begin=============\n",
      "Test AUC: 0.683\n",
      "Test AUC:(0.674,0.692) percentile 95% CI\n",
      "Test AUPRC: 0.209\n",
      "Test AUPRC:(0.199,0.221) percentile 95% CI\n",
      "==============model KNN over=============\n",
      "==============model DTC begin=============\n",
      "Test AUC: 0.702\n",
      "Test AUC:(0.692,0.711) percentile 95% CI\n",
      "Test AUPRC: 0.246\n",
      "Test AUPRC:(0.232,0.261) percentile 95% CI\n",
      "==============model DTC over=============\n",
      "==============model NB begin=============\n",
      "Test AUC: 0.706\n",
      "Test AUC:(0.696,0.716) percentile 95% CI\n",
      "Test AUPRC: 0.270\n",
      "Test AUPRC:(0.257,0.283) percentile 95% CI\n",
      "==============model NB over=============\n",
      "==============model MLP begin=============\n",
      "Test AUC: 0.716\n",
      "Test AUC:(0.707,0.726) percentile 95% CI\n",
      "Test AUPRC: 0.261\n",
      "Test AUPRC:(0.247,0.277) percentile 95% CI\n",
      "==============model MLP over=============\n",
      "==============model RF begin=============\n",
      "Test AUC: 0.774\n",
      "Test AUC:(0.766,0.783) percentile 95% CI\n",
      "Test AUPRC: 0.338\n",
      "Test AUPRC:(0.321,0.356) percentile 95% CI\n",
      "==============model RF over=============\n",
      "==============model XGBoost begin=============\n",
      "Test AUC: 0.796\n",
      "Test AUC:(0.789,0.804) percentile 95% CI\n",
      "Test AUPRC: 0.357\n",
      "Test AUPRC:(0.340,0.376) percentile 95% CI\n",
      "==============model XGBoost over=============\n",
      "==============model Catboost begin=============\n",
      "Test AUC: 0.798\n",
      "Test AUC:(0.791,0.806) percentile 95% CI\n",
      "Test AUPRC: 0.357\n",
      "Test AUPRC:(0.341,0.376) percentile 95% CI\n",
      "==============model Catboost over=============\n",
      "==============model LightGBM begin=============\n",
      "Test AUC: 0.800\n",
      "Test AUC:(0.793,0.808) percentile 95% CI\n",
      "Test AUPRC: 0.365\n",
      "Test AUPRC:(0.348,0.384) percentile 95% CI\n",
      "==============model LightGBM over=============\n"
     ]
    }
   ],
   "source": [
    "#### get data and result ##############\n",
    "models = ['LR','SVM','KNN','DTC','NB','MLP','RF','XGBoost','Catboost','LightGBM']#\n",
    "preData = {}\n",
    "for model in models:\n",
    "    temp = pd.read_csv(search_path +'/'+ model+'/predict_result_'+model+'.csv')\n",
    "    preData[model] = temp\n",
    "    \n",
    "\n",
    "#####################################\n",
    "def get_data(y_true, y_pred,model):\n",
    "    print('==============model %s begin============='%model) \n",
    "    roc_curve, auc, uci,dci, auc_scores, mean_fpr, tprs_lower, tprs_upper = get_roc_CI_curve(y_true, y_pred)\n",
    "    pr_curve, pr_auc, pr_uci,pr_dci, pr_auc_scores, mean_rec, prec_lower, prec_upper = get_pr_CI_curve(y_true, y_pred)\n",
    "    P_true, P_pred, P_trues, P_preds = get_calibration_CI_curve(y_true, y_pred,10) #q=5\n",
    "    roc_CI = {'roc_curve':roc_curve,'auc':auc,'dci':dci,'uci':uci,'auc_scores':auc_scores,'mean_fpr':mean_fpr,'tprs_lower':tprs_lower,'tprs_upper':tprs_upper}\n",
    "    pr_CI = {'pr_curve':pr_curve,'pr_auc':pr_auc,'pr_dci':pr_dci,'pr_uci':pr_uci,'pr_auc_scores':pr_auc_scores,\n",
    "         'mean_rec':mean_rec,'prec_lower':prec_lower,'prec_upper':prec_upper}\n",
    "    cali_CI = {'P_true': P_true,'P_pred':P_pred,'P_trues':P_trues,'P_preds':P_preds}\n",
    "    \n",
    "    data = {'roc_CI': roc_CI,'pr_CI':pr_CI,'cali_CI':cali_CI}\n",
    "    f = open(save_path+'/results/roc_pr_cali_CI_Model_'+model+'.pkl','wb')\n",
    "    pickle.dump(data, f)\n",
    "    f.close()\n",
    "    print('==============model %s over============='%model)  \n",
    "    return data\n",
    "\n",
    "   \n",
    "for model in models: \n",
    "    y_true = preData[model]['y_test']\n",
    "    y_pred = preData[model]['y_tepred']\n",
    "    rt = get_data(y_true,y_pred,model)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "35d4ebbb-b783-492b-b818-40d674c36c83",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 900x450 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "\n",
    "sns.set_theme(style=\"whitegrid\")\n",
    "plt.rcParams['font.sans-serif'] = ['Arial Unicode MS']\n",
    "\n",
    "################################### plot figures roc_pc_calibration curves ###########################\n",
    "models = ['LR','SVM','KNN','DTC','NB','MLP','RF','XGBoost','Catboost','LightGBM']\n",
    "#colors =['#2ca02c','#d62728','#1f77b4','#ff7f0e']\n",
    "\n",
    "\n",
    "allData = {}\n",
    "for model in models:\n",
    "    temp = pickle.load(open(save_path+'/results/roc_pr_cali_CI_Model_'+model+'.pkl','rb'))\n",
    "    allData[model] = temp\n",
    "\n",
    "\n",
    "def draw_roc_pr_cali_curves():\n",
    "    fig = plt.figure(figsize=(9,4.5))\n",
    "    ##=================for roc curve========================\n",
    "    ax1 = plt.subplot(1,2,1) \n",
    "    ## for model_i\n",
    "    for model in models:\n",
    "        temp = allData[model]\n",
    "        data = temp['roc_CI']\n",
    "        fpr = data['roc_curve'][0];tpr = data['roc_curve'][1];\n",
    "        mean_fpr = data['mean_fpr']; tprs_lower=data['tprs_lower']; tprs_upper=data['tprs_upper']\n",
    "        plt.plot(fpr,tpr,lw=1.25, label=model)#, color=colors[s]\n",
    "        plt.fill_between(mean_fpr,tprs_lower,tprs_upper, alpha=0.3, label=\"_nolegend_\")#, color=colors[s]\n",
    "     \n",
    "    lg = plt.legend(\n",
    "        [   \n",
    "            'LR:{:.3f} [{:.3f},{:.3f}]'.format(allData['LR']['roc_CI']['auc'],allData['LR']['roc_CI']['dci'],allData['LR']['roc_CI']['uci']),\n",
    "            'SVM:{:.3f} [{:.3f},{:.3f}]'.format(allData['SVM']['roc_CI']['auc'],allData['SVM']['roc_CI']['dci'],allData['SVM']['roc_CI']['uci']),            \n",
    "            'KNN:{:.3f} [{:.3f},{:.3f}]'.format(allData['KNN']['roc_CI']['auc'],allData['KNN']['roc_CI']['dci'],allData['KNN']['roc_CI']['uci']),\n",
    "            'DT:{:.3f} [{:.3f},{:.3f}]'.format(allData['DTC']['roc_CI']['auc'],allData['DTC']['roc_CI']['dci'],allData['DTC']['roc_CI']['uci']),            \n",
    "            'NB:{:.3f} [{:.3f},{:.3f}]'.format(allData['NB']['roc_CI']['auc'],allData['NB']['roc_CI']['dci'],allData['NB']['roc_CI']['uci']),\n",
    "            'MLP:{:.3f} [{:.3f},{:.3f}]'.format(allData['MLP']['roc_CI']['auc'],allData['MLP']['roc_CI']['dci'],allData['MLP']['roc_CI']['uci']),\n",
    "            'RF:{:.3f} [{:.3f},{:.3f}]'.format(allData['RF']['roc_CI']['auc'],allData['RF']['roc_CI']['dci'],allData['RF']['roc_CI']['uci']),\n",
    "            'XGBoost:{:.3f} [{:.3f},{:.3f}]'.format(allData['XGBoost']['roc_CI']['auc'],allData['XGBoost']['roc_CI']['dci'],allData['XGBoost']['roc_CI']['uci']),\n",
    "            'CatBoost:{:.3f} [{:.3f},{:.3f}]'.format(allData['Catboost']['roc_CI']['auc'],allData['Catboost']['roc_CI']['dci'],allData['Catboost']['roc_CI']['uci']),\n",
    "            'LightGBM:{:.3f} [{:.3f},{:.3f}]'.format(allData['LightGBM']['roc_CI']['auc'],allData['LightGBM']['roc_CI']['dci'],allData['LightGBM']['roc_CI']['uci']),\n",
    "            #'BFSM  : {:.3f}'.format(allData['step'+str(0)]['roc_CI']['auc']),\n",
    "            #'AFSM-1: {:.3f}'.format(allData['step'+str(1)]['roc_CI']['auc']),\n",
    "            #'AFSM-2: {:.3f}'.format(allData['step'+str(2)]['roc_CI']['auc']),\n",
    "            #'AFSM-3: {:.3f}'.format(allData['step'+str(3)]['roc_CI']['auc']),\n",
    "        ],\n",
    "        title = 'AUROC [95% CI]',\n",
    "        loc = 'lower right',\n",
    "        handlelength = 1.0,\n",
    "        fontsize = 9,\n",
    "    )\n",
    "    lg.get_title().set_fontsize(10)\n",
    "    \n",
    "    plt.plot([0,1],[0,1],'k:')\n",
    "    plt.xlabel('False Positive Rate')\n",
    "    plt.ylabel('True Positive Rate')\n",
    "    plt.xlim(0,1)\n",
    "    plt.ylim(0,1)\n",
    "    plt.title('(c) ROC curves for Cohort 2', fontsize=12, pad=12)\n",
    "\n",
    "    ##=================for pr curve========================\n",
    "    ax2 = plt.subplot(1,2,2)     \n",
    "   ## for model_i\n",
    "    for model in models:\n",
    "        temp = allData[model]\n",
    "        data = temp['pr_CI']\n",
    "        precision = data['pr_curve'][0]; recall = data['pr_curve'][1]\n",
    "        mean_rec = data['mean_rec']; prec_lower=data['prec_lower']; prec_upper=data['prec_upper']\n",
    "        plt.plot(recall,precision,lw=1.25, label=model)#, color=colors[s]\n",
    "        plt.fill_between(mean_rec,prec_lower,prec_upper, alpha=0.3, label=\"_nolegend_\")#, color=colors[s]\n",
    "     \n",
    "    lg = plt.legend(\n",
    "        [\n",
    "            #'BFSM   : {:.3f}'.format(allData['step'+str(0)]['pr_CI']['pr_auc']),\n",
    "            #'AFSM-1: {:.3f}'.format(allData['step'+str(1)]['pr_CI']['pr_auc']),\n",
    "            #'AFSM-2: {:.3f}'.format(allData['step'+str(2)]['pr_CI']['pr_auc']),\n",
    "            #'AFSM-3: {:.3f}'.format(allData['step'+str(3)]['pr_CI']['pr_auc']),\n",
    "            #models = ['LR','SVM','KNN','DTC','NB','MLP','RF','XGBoost','Catboost','LightGBM']\n",
    "            'LR:{:.3f} [{:.3f},{:.3f}]'.format(allData['LR']['pr_CI']['pr_auc'],allData['LR']['pr_CI']['pr_dci'],allData['LR']['pr_CI']['pr_uci']),\n",
    "            'SVM:{:.3f} [{:.3f},{:.3f}]'.format(allData['SVM']['pr_CI']['pr_auc'],allData['SVM']['pr_CI']['pr_dci'],allData['SVM']['pr_CI']['pr_uci']),            \n",
    "            'KNN:{:.3f} [{:.3f},{:.3f}]'.format(allData['KNN']['pr_CI']['pr_auc'],allData['KNN']['pr_CI']['pr_dci'],allData['KNN']['pr_CI']['pr_uci']),\n",
    "            'DT:{:.3f} [{:.3f},{:.3f}]'.format(allData['DTC']['pr_CI']['pr_auc'],allData['DTC']['pr_CI']['pr_dci'],allData['DTC']['pr_CI']['pr_uci']),            \n",
    "            'NB:{:.3f} [{:.3f},{:.3f}]'.format(allData['NB']['pr_CI']['pr_auc'],allData['NB']['pr_CI']['pr_dci'],allData['NB']['pr_CI']['pr_uci']),\n",
    "            'MLP:{:.3f} [{:.3f},{:.3f}]'.format(allData['MLP']['pr_CI']['pr_auc'],allData['MLP']['pr_CI']['pr_dci'],allData['MLP']['pr_CI']['pr_uci']),\n",
    "            'RF:{:.3f} [{:.3f},{:.3f}]'.format(allData['RF']['pr_CI']['pr_auc'],allData['RF']['pr_CI']['pr_dci'],allData['RF']['pr_CI']['pr_uci']),\n",
    "            'XGBoost:{:.3f} [{:.3f},{:.3f}]'.format(allData['XGBoost']['pr_CI']['pr_auc'],allData['XGBoost']['pr_CI']['pr_dci'],allData['XGBoost']['pr_CI']['pr_uci']),\n",
    "            'CatBoost:{:.3f} [{:.3f},{:.3f}]'.format(allData['Catboost']['pr_CI']['pr_auc'],allData['Catboost']['pr_CI']['pr_dci'],allData['Catboost']['pr_CI']['pr_uci']),\n",
    "            'LightGBM:{:.3f} [{:.3f},{:.3f}]'.format(allData['LightGBM']['pr_CI']['pr_auc'],allData['LightGBM']['pr_CI']['pr_dci'],allData['LightGBM']['pr_CI']['pr_uci']),  \n",
    "        ],\n",
    "        title = 'AUPRC [95% CI]',\n",
    "        loc = 'upper right',#'lower left',\n",
    "        handlelength = 1.0,\n",
    "        fontsize = 9,\n",
    "    )\n",
    "    lg.get_title().set_fontsize(10)    \n",
    "    plt.xlabel('Recall')\n",
    "    plt.ylabel('Precision')\n",
    "    plt.xlim(0,1)\n",
    "    plt.ylim(0,1)\n",
    "    #plt.title('{}.T={}.dt={}'.format(tast,T,dt))    \n",
    "    plt.title('(d) PR curves for Cohort 2', fontsize=12, pad=12)\n",
    "    \n",
    "    #########################################################        \n",
    "    plt.tight_layout()\n",
    "    plt.savefig(save_path+'/figure/figure_roc_pr_curves_MIMIC.png',dpi=300)\n",
    "    #plt.savefig(save_path+'/figure/figure_roc_pr_cali_curves.eps',format='eps',dpi=300)\n",
    "    #plt.show()\n",
    "         \n",
    "    \n",
    "        \n",
    "draw_roc_pr_cali_curves()    \n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1d12ca5f-593a-418b-aa34-80cb8b4b330b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "28116a96-f16d-4976-a96a-5676842cfd7d",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "efdd7ba2-0058-436b-9826-ac4ab7f2a5f6",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5a2deb44-26b7-4be4-9503-006de0d37d9c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ae8a1ecb-feb0-4ddd-9e7b-6b50cd4f8705",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f8eb7b0-7d8d-4529-b86b-1e9237152f07",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "dev",
   "language": "python",
   "name": "dev"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}