machineLearning_diabetes_220112.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "7c04b1a7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0    500\n",
      "1    268\n",
      "Name: Outcome, dtype: int64\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pregnancies</th>\n",
       "      <th>Glucose</th>\n",
       "      <th>BloodPressure</th>\n",
       "      <th>SkinThickness</th>\n",
       "      <th>Insulin</th>\n",
       "      <th>BMI</th>\n",
       "      <th>DiabetesPedigreeFunction</th>\n",
       "      <th>Age</th>\n",
       "      <th>Outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>6</td>\n",
       "      <td>148</td>\n",
       "      <td>72</td>\n",
       "      <td>35</td>\n",
       "      <td>0</td>\n",
       "      <td>33.6</td>\n",
       "      <td>0.627</td>\n",
       "      <td>50</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>85</td>\n",
       "      <td>66</td>\n",
       "      <td>29</td>\n",
       "      <td>0</td>\n",
       "      <td>26.6</td>\n",
       "      <td>0.351</td>\n",
       "      <td>31</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8</td>\n",
       "      <td>183</td>\n",
       "      <td>64</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>23.3</td>\n",
       "      <td>0.672</td>\n",
       "      <td>32</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pregnancies  Glucose  BloodPressure  SkinThickness  Insulin   BMI  \\\n",
       "0            6      148             72             35        0  33.6   \n",
       "1            1       85             66             29        0  26.6   \n",
       "2            8      183             64              0        0  23.3   \n",
       "\n",
       "   DiabetesPedigreeFunction  Age  Outcome  \n",
       "0                     0.627   50        1  \n",
       "1                     0.351   31        0  \n",
       "2                     0.672   32        1  "
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score, precision_score, recall_score, roc_auc_score\n",
    "from sklearn.metrics import f1_score, confusion_matrix, precision_recall_curve, roc_curve\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "diabetes_data = pd.read_csv('diabetes.csv')\n",
    "print(diabetes_data['Outcome'].value_counts())\n",
    "diabetes_data.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "638f51b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import Binarizer\n",
    "\n",
    "def get_eval_by_threshold(y_test, pred_proba_c1, thresholds):\n",
    "    for custom_threshold in thresholds:\n",
    "        binarizer = Binarizer(threshold=custom_threshold).fit(pred_proba_c1)\n",
    "        custom_predict = binarizer.transform(pred_proba_c1)\n",
    "        print('임곗값: ', custom_threshold)\n",
    "        get_clf_eval(y_test, custom_predict)\n",
    "\n",
    "def precision_recall_curve_plot(y_test, pred_proba_c1):\n",
    "    precisions, recalls, thresholds = precision_recall_curve(y_test, pred_proba_c1)\n",
    "    \n",
    "    plt.figure(figsize=(8, 6))\n",
    "    threshold_boundary = thresholds.shape[0]\n",
    "    plt.plot(thresholds, precisions[0:threshold_boundary], linestyle='--', label='precision')\n",
    "    plt.plot(thresholds, recalls[0:threshold_boundary], label='recall')\n",
    "    \n",
    "    start, end = plt.xlim()\n",
    "    plt.xticks(np.round(np.arange(start, end, 0.1), 2))\n",
    "    \n",
    "    plt.xlabel('Threshold value'); plt.ylabel('Precision and Recall value')\n",
    "    plt.legend(); plt.grid()\n",
    "    plt.show()\n",
    "    \n",
    "def get_clf_eval(y_test, pred):\n",
    "    confusion = confusion_matrix( y_test, pred)\n",
    "    accuracy = accuracy_score(y_test, pred)\n",
    "    precision = precision_score(y_test, pred)\n",
    "    recall = recall_score(y_test, pred)\n",
    "    f1 = f1_score(y_test, pred)\n",
    "    \n",
    "    print('오차 행렬')\n",
    "    print(confusion)\n",
    "    \n",
    "    print('정확도: {0:.4f}, 정밀도: {1:.4f}, 재현율: {2:.4f}, F1: {3:.4f}'.format(accuracy, precision, recall, f1))\n",
    "    \n",
    "def get_clf_eval(y_test, pred=None, pred_proba=None):\n",
    "    confusion = confusion_matrix( y_test, pred)\n",
    "    accuracy = accuracy_score(y_test, pred)\n",
    "    precision = precision_score(y_test, pred)\n",
    "    recall = recall_score(y_test, pred)\n",
    "    f1 = f1_score(y_test, pred)\n",
    "    \n",
    "    roc_auc = roc_auc_score(y_test, pred_proba)\n",
    "    print('오차 행렬')\n",
    "    print(confusion)\n",
    "    \n",
    "    print('정확도: {0:.4f}, 정밀도: {1:.4f}, 재현율: {2:.4f}, F1: {3:.4f}, AUC: {4:.4f}'.format(accuracy, precision, recall, f1, roc_auc))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "7059459b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "오차 행렬\n",
      "[[88 12]\n",
      " [23 31]]\n",
      "정확도: 0.7727, 정밀도: 0.7209, 재현율: 0.5741, F1: 0.6392, AUC: 0.7919\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\82108\\anaconda3\\lib\\site-packages\\sklearn\\linear_model\\_logistic.py:763: ConvergenceWarning: lbfgs failed to converge (status=1):\n",
      "STOP: TOTAL NO. of ITERATIONS REACHED LIMIT.\n",
      "\n",
      "Increase the number of iterations (max_iter) or scale the data as shown in:\n",
      "    https://scikit-learn.org/stable/modules/preprocessing.html\n",
      "Please also refer to the documentation for alternative solver options:\n",
      "    https://scikit-learn.org/stable/modules/linear_model.html#logistic-regression\n",
      "  n_iter_i = _check_optimize_result(\n"
     ]
    }
   ],
   "source": [
    "#feature 데이터 세트 X, label 데이터 세트 y를 추출\n",
    "#맨 끝이 Outcome column으로, label 값임. column 위치 -1을 이용해 추출\n",
    "X = diabetes_data.iloc[:, :-1]\n",
    "y = diabetes_data.iloc[:, -1]\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=156, stratify=y)\n",
    "\n",
    "#LogisticRegression으로 학습/예측/평가 수행\n",
    "lr_clf = LogisticRegression()\n",
    "lr_clf.fit(X_train, y_train)\n",
    "pred = lr_clf.predict(X_test)\n",
    "pred_proba = lr_clf.predict_proba(X_test)[:, 1]\n",
    "\n",
    "get_clf_eval(y_test, pred, pred_proba)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "89dba687",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pred_proba_c1 = lr_clf.predict_proba(X_test)[:, 1]\n",
    "precision_recall_curve_plot(y_test, pred_proba_c1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "fcd2d053",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  5.,   0.,   4.,  32., 156., 211., 163.,  95.,  56.,  46.]),\n",
       " array([  0. ,  19.9,  39.8,  59.7,  79.6,  99.5, 119.4, 139.3, 159.2,\n",
       "        179.1, 199. ]),\n",
       " <BarContainer object of 10 artists>)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist(diabetes_data['Glucose'], bins=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "5c298fdc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Glucose 0 건수는 5, 퍼센트는 0.65%\n",
      "BloodPressure 0 건수는 35, 퍼센트는 4.56%\n",
      "SkinThickness 0 건수는 227, 퍼센트는 29.56%\n",
      "Insulin 0 건수는 374, 퍼센트는 48.70%\n",
      "BMI 0 건수는 11, 퍼센트는 1.43%\n"
     ]
    }
   ],
   "source": [
    "#0값을 검사할 feature명 리스트\n",
    "zero_features = ['Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI']\n",
    "\n",
    "#전체 데이터 건수\n",
    "total_count = diabetes_data['Glucose'].count()\n",
    "\n",
    "for feature in zero_features:\n",
    "    zero_count = diabetes_data[diabetes_data[feature] == 0][feature].count()\n",
    "    print('{0} 0 건수는 {1}, 퍼센트는 {2:.2f}%'.format(feature, zero_count, 100*zero_count/total_count))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "5ddc54fc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#SkinThickness와 Insulin의 0값의 비율이 높음"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "2eaad795",
   "metadata": {},
   "outputs": [],
   "source": [
    "mean_zero_features = diabetes_data[zero_features].mean()\n",
    "diabetes_data[zero_features] = diabetes_data[zero_features].replace(0, mean_zero_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "086bced2",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "오차 행렬\n",
      "[[90 10]\n",
      " [21 33]]\n",
      "정확도: 0.7987, 정밀도: 0.7674, 재현율: 0.6111, F1: 0.6804, AUC: 0.8433\n"
     ]
    }
   ],
   "source": [
    "X = diabetes_data.iloc[:, :-1]\n",
    "y = diabetes_data.iloc[:, -1]\n",
    "\n",
    "scaler = StandardScaler()\n",
    "X_scaled = scaler.fit_transform(X)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_scaled, y, test_size=0.2, random_state=156, stratify=y)\n",
    "\n",
    "#LogisticRegression으로 학습/예측/평가 수행\n",
    "lr_clf = LogisticRegression()\n",
    "lr_clf.fit(X_train, y_train)\n",
    "pred = lr_clf.predict(X_test)\n",
    "pred_proba = lr_clf.predict_proba(X_test)[:, 1]\n",
    "\n",
    "get_clf_eval(y_test, pred, pred_proba)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "764ddee5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "임곗값:  0.3\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\82108\\anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py:63: FutureWarning: Arrays of bytes/strings is being converted to decimal numbers if dtype='numeric'. This behavior is deprecated in 0.24 and will be removed in 1.1 (renaming of 0.26). Please convert your data to numeric values explicitly instead.\n",
      "  return f(*args, **kwargs)\n"
     ]
    },
    {
     "ename": "ValueError",
     "evalue": "Input contains NaN, infinity or a value too large for dtype('float64').",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_19296/3071637485.py\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[0mthresholds\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[1;36m0.3\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.33\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.36\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.39\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.42\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.45\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.48\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m0.50\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      2\u001b[0m \u001b[0mpred_proba\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlr_clf\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mpredict_proba\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mX_test\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 3\u001b[1;33m \u001b[0mget_eval_by_threshold\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpred_proba\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mreshape\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m-\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;36m1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mthresholds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_19296/2017515250.py\u001b[0m in \u001b[0;36mget_eval_by_threshold\u001b[1;34m(y_test, pred_proba_c1, thresholds)\u001b[0m\n\u001b[0;32m      6\u001b[0m         \u001b[0mcustom_predict\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mbinarizer\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtransform\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mpred_proba_c1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      7\u001b[0m         \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'임곗값: '\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcustom_threshold\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 8\u001b[1;33m         \u001b[0mget_clf_eval\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcustom_predict\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      9\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     10\u001b[0m \u001b[1;32mdef\u001b[0m \u001b[0mprecision_recall_curve_plot\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpred_proba_c1\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\AppData\\Local\\Temp/ipykernel_19296/2017515250.py\u001b[0m in \u001b[0;36mget_clf_eval\u001b[1;34m(y_test, pred, pred_proba)\u001b[0m\n\u001b[0;32m     42\u001b[0m     \u001b[0mf1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mf1_score\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpred\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     43\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 44\u001b[1;33m     \u001b[0mroc_auc\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mroc_auc_score\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mpred_proba\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     45\u001b[0m     \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;34m'오차 행렬'\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     46\u001b[0m     \u001b[0mprint\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mconfusion\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py\u001b[0m in \u001b[0;36minner_f\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m     61\u001b[0m             \u001b[0mextra_args\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mall_args\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     62\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mextra_args\u001b[0m \u001b[1;33m<=\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 63\u001b[1;33m                 \u001b[1;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     64\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     65\u001b[0m             \u001b[1;31m# extra_args > 0\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\metrics\\_ranking.py\u001b[0m in \u001b[0;36mroc_auc_score\u001b[1;34m(y_true, y_score, average, sample_weight, max_fpr, multi_class, labels)\u001b[0m\n\u001b[0;32m    522\u001b[0m     \u001b[0my_type\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtype_of_target\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    523\u001b[0m     \u001b[0my_true\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcheck_array\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mensure_2d\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mNone\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 524\u001b[1;33m     \u001b[0my_score\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mcheck_array\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0my_score\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mensure_2d\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;32mFalse\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    525\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    526\u001b[0m     if y_type == \"multiclass\" or (y_type == \"binary\" and\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py\u001b[0m in \u001b[0;36minner_f\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m     61\u001b[0m             \u001b[0mextra_args\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;33m-\u001b[0m \u001b[0mlen\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mall_args\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     62\u001b[0m             \u001b[1;32mif\u001b[0m \u001b[0mextra_args\u001b[0m \u001b[1;33m<=\u001b[0m \u001b[1;36m0\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 63\u001b[1;33m                 \u001b[1;32mreturn\u001b[0m \u001b[0mf\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m     64\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     65\u001b[0m             \u001b[1;31m# extra_args > 0\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py\u001b[0m in \u001b[0;36mcheck_array\u001b[1;34m(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, estimator)\u001b[0m\n\u001b[0;32m    718\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    719\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mforce_all_finite\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 720\u001b[1;33m             _assert_all_finite(array,\n\u001b[0m\u001b[0;32m    721\u001b[0m                                allow_nan=force_all_finite == 'allow-nan')\n\u001b[0;32m    722\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\sklearn\\utils\\validation.py\u001b[0m in \u001b[0;36m_assert_all_finite\u001b[1;34m(X, allow_nan, msg_dtype)\u001b[0m\n\u001b[0;32m    101\u001b[0m                 not allow_nan and not np.isfinite(X).all()):\n\u001b[0;32m    102\u001b[0m             \u001b[0mtype_err\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m'infinity'\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mallow_nan\u001b[0m \u001b[1;32melse\u001b[0m \u001b[1;34m'NaN, infinity'\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 103\u001b[1;33m             raise ValueError(\n\u001b[0m\u001b[0;32m    104\u001b[0m                     \u001b[0mmsg_err\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    105\u001b[0m                     (type_err,\n",
      "\u001b[1;31mValueError\u001b[0m: Input contains NaN, infinity or a value too large for dtype('float64')."
     ]
    }
   ],
   "source": [
    "thresholds = [0.3, 0.33, 0.36, 0.39, 0.42, 0.45, 0.48, 0.50]\n",
    "pred_proba = lr_clf.predict_proba(X_test)\n",
    "get_eval_by_threshold(y_test, pred_proba[:, 1].reshape(-1, 1), thresholds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "f73bdd33",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "오차 행렬\n",
      "[[88 12]\n",
      " [19 35]]\n",
      "정확도: 0.7987, 정밀도: 0.7447, 재현율: 0.6481, F1: 0.6931, AUC: 0.8433\n"
     ]
    }
   ],
   "source": [
    "binarizer = Binarizer(threshold=0.48)\n",
    "\n",
    "pred_th_048 = binarizer.fit_transform(pred_proba[:, 1].reshape(-1, 1))\n",
    "\n",
    "get_clf_eval(y_test, pred_th_048, pred_proba[:, 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1cedf7fe",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}