card-fraud-detection-resample.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 导入官方库训练"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# preprocess\n",
    "import time\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.utils import resample\n",
    "\n",
    "# model\n",
    "import xgboost as xgb\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "\n",
    "# socre\n",
    "from sklearn.metrics import average_precision_score\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from sklearn.metrics import f1_score, precision_score, recall_score, auc\n",
    "\n",
    "# figure\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 定义辅助函数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 训练的辅助函数\n",
    "def trainHelper(model, x_train, y_train):\n",
    "    print(\"=== train ===\")\n",
    "    start = time.time()\n",
    "    model.fit(x_train, y_train)\n",
    "    trainingTime = time.time() - start\n",
    "    print(\"Training time: \", trainingTime, \"sec\\n\")\n",
    "    return model, trainingTime\n",
    "\n",
    "\n",
    "# 推理的辅助函数\n",
    "def inferenceHelper(model, x_test, y_test):\n",
    "    print(\"=== inference ===\")\n",
    "    start_time_inference = time.time()\n",
    "    y_pred = model.predict(x_test)\n",
    "    inference_time = time.time() - start_time_inference\n",
    "    # 计算F1, precision和recall\n",
    "    f1 = f1_score(y_test, y_pred)\n",
    "    precision = precision_score(y_test, y_pred)\n",
    "    recall = recall_score(y_test, y_pred)\n",
    "\n",
    "    print(\"F1 Score: \", f1)\n",
    "    print(\"precision: \", precision)\n",
    "    print(\"recall: \", recall)\n",
    "    print(f\"Inference Time: {inference_time} sec\\n\")\n",
    "\n",
    "    return f1,precision,recall,inference_time\n",
    "\n",
    "\n",
    "# gridSearch的辅助函数\n",
    "def gridSearchHelper(model, p_grid, x_train, y_train):\n",
    "    print(\"=== grid search ===\")\n",
    "    grid_search = GridSearchCV(estimator=model, param_grid=p_grid, scoring=\"f1\", cv=5)\n",
    "    start = time.time()\n",
    "    grid_search.fit(x_train, y_train)\n",
    "    searchTime = time.time() - start\n",
    "    print(\"grid search time: \", searchTime, \"sec\\n\")\n",
    "    return grid_search.best_params_, searchTime\n",
    "\n",
    "\n",
    "def plot_PRC(model, x_test, y_test):\n",
    "    y_score = model.predict_proba(x_test)\n",
    "\n",
    "    # 计算精确率和召回率和AUPRC\n",
    "    precision, recall, _ = precision_recall_curve(y_test, y_score[:, 1])\n",
    "    average_precision = average_precision_score(y_test, y_score[:, 1])\n",
    "    print(\"AUPRC:\", average_precision)\n",
    "\n",
    "    # 绘制PRC曲线\n",
    "    plt.figure()\n",
    "    plt.step(recall, precision, color='b', alpha=0.2, where='post')\n",
    "    plt.fill_between(recall, precision, step='post', alpha=0.2, color='b')\n",
    "    plt.xlabel('Recall')\n",
    "    plt.ylabel('Precision')\n",
    "    plt.ylim([0.0, 1.05])\n",
    "    plt.xlim([0.0, 1.0])\n",
    "    plt.title('Precision-Recall curve: AP={0:0.2f}'.format(average_precision))\n",
    "    plt.show()\n",
    "\n",
    "    return average_precision\n",
    "\n",
    "def printAsTable(origin_data, accelerate_data):\n",
    "    print(\"{:<10} {:<15} {:<15}\".format('Model', 'origin', 'oneAPI'))\n",
    "    print(\"-\" * 40)\n",
    "\n",
    "    models = ['LR', 'DT', 'RF', 'SVM', 'KNN']\n",
    "    for i in range(len(models)):\n",
    "        print(\"{:<10} {:<15.4f} {:<15.4f}\".format(models[i], origin_data[i], accelerate_data[i]))\n",
    "\n",
    "def plotCompareTime(title, origin, accelerated):\n",
    "    save_path = \"resample/\" + title + \".png\"\n",
    "    unit_1 = [1,1,1,1,1]\n",
    "    bar_width = 0.35\n",
    "    opacity = 0.8\n",
    "    algorithms = ['LR', 'DT', 'RF', 'SVM', 'KNN']\n",
    "    speedup = [o/a for o, a in zip(origin, accelerated)]\n",
    "    fig, ax = plt.subplots(figsize=(8,8))\n",
    "\n",
    "    rects1 = ax.bar(np.arange(len(unit_1)), unit_1, bar_width,\n",
    "                    alpha=opacity, color='#1893aa', align='center')\n",
    "    rects2 = ax.bar(np.arange(len(speedup)) + bar_width, speedup, bar_width,\n",
    "                    alpha=opacity, color='#225fe2', align='center')\n",
    "\n",
    "    ax.set_xlabel('Algorithm')\n",
    "    ax.set_ylabel('Speedup Ratio')\n",
    "    ax.set_title(title + ' Comparison: origin/oneAPI')\n",
    "    # ax.set_xticks(x)\n",
    "    ax.set_xticks(np.arange(len(unit_1)) + bar_width / 2)\n",
    "    ax.set_xticklabels(algorithms)\n",
    "    ax.legend((rects1[0], rects2[0]), ('origin', 'oneAPI'))\n",
    "\n",
    "    for rect in rects2:\n",
    "        height = rect.get_height()\n",
    "        ax.annotate(f'{height:.2f}', xy=(rect.get_x() + rect.get_width() / 2, height), xytext=(0, 3), textcoords=\"offset points\", ha='center', va='bottom')\n",
    "    plt.savefig(save_path, dpi=300, bbox_inches='tight')\n",
    "    plt.show()\n",
    "\n",
    "\n",
    "def plotCompareScore(title, origin, accelerated):\n",
    "    save_path = \"resample/\" + title + \".png\" \n",
    "    unit_1 = [1,1,1,1,1]\n",
    "    bar_width = 0.35\n",
    "    opacity = 0.8\n",
    "    algorithms = ['LR', 'DT', 'RF', 'SVM', 'KNN']\n",
    "    speedup = [a/o for o, a in zip(origin, accelerated)]\n",
    "    fig, ax = plt.subplots(figsize=(8,8))\n",
    "\n",
    "    rects1 = ax.bar(np.arange(len(unit_1)), unit_1, bar_width,\n",
    "                    alpha=opacity, color='#1893aa', align='center')\n",
    "    rects2 = ax.bar(np.arange(len(speedup)) + bar_width, speedup, bar_width,\n",
    "                    alpha=opacity, color='#225fe2', align='center')\n",
    "\n",
    "    ax.set_xlabel('Algorithm')\n",
    "    ax.set_ylabel('Speedup Ratio')\n",
    "    ax.set_title(title + ' Comparison: origin/oneAPI')\n",
    "    # ax.set_xticks(x)\n",
    "    ax.set_xticks(np.arange(len(unit_1)) + bar_width / 2)\n",
    "    ax.set_xticklabels(algorithms)\n",
    "    ax.legend((rects1[0], rects2[0]), ('origin', 'oneAPI'))\n",
    "\n",
    "    for rect in rects2:\n",
    "        height = rect.get_height()\n",
    "        ax.annotate(f'{height:.2f}', xy=(rect.get_x() + rect.get_width() / 2, height), xytext=(0, 3), textcoords=\"offset points\", ha='center', va='bottom')\n",
    "    plt.savefig(save_path, dpi=300, bbox_inches='tight')\n",
    "    plt.show()\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 训练过程\n",
    "1. 读入csv数据，去除无关列Time\n",
    "2. 使用**重采样的数据集进行训练**\n",
    "3. 预测时，使用原始数据集  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loading data...\n",
      "number of majority_class data:  284315\n",
      "number of minority_class data:  492\n",
      "fraction of minority class:  0.001727485630620034\n",
      "number of downSample data =  4920\n",
      "loading complete!\n"
     ]
    }
   ],
   "source": [
    "print(\"loading data...\")\n",
    "data = pd.read_csv('creditcard.csv')\n",
    "data = data.drop(columns=['Time'], axis=1)\n",
    "# Amount列值缩放到[-1,1]区间\n",
    "scaler = StandardScaler()\n",
    "data['Amount'] = scaler.fit_transform(data['Amount'].values.reshape(-1, 1))\n",
    "x = data.drop(columns=['Class'], axis=1) \n",
    "y = data['Class']\n",
    "\n",
    "x_test = x\n",
    "y_test = y\n",
    "# x_train, x_test, y_train, y_test = train_test_split(x, y,test_size=0.3, random_state=42)\n",
    "\n",
    "# 对数据下采样，比列为1:9\n",
    "majority_class = data[data.Class == 0]\n",
    "minority_class = data[data.Class == 1]\n",
    "print(\"number of majority_class data: \", len(majority_class))\n",
    "print(\"number of minority_class data: \", len(minority_class))\n",
    "print(\"fraction of minority class: \", len(minority_class)/len(data))\n",
    "majority_downsampled = resample(majority_class, replace=False, n_samples=len(minority_class)*9, random_state=123)\n",
    "majority_downsampled = pd.DataFrame(majority_downsampled, columns=data.columns) \n",
    "downsampled_data = pd.concat([majority_downsampled, minority_class])\n",
    "\n",
    "x_train_down = downsampled_data.drop(columns=['Class'], axis=1)\n",
    "y_train_down = downsampled_data[\"Class\"]\n",
    "\n",
    "print(\"number of downSample data = \", len(downsampled_data))\n",
    "print(\"loading complete!\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. 使用不同算法训练并预测\n",
    "   \n",
    "   2.1. 逻辑回归"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "logistic regression:\n",
      "=== grid search ===\n",
      "grid search time:  7.273085355758667 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.024288654327392578 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.5846579129232896\n",
      "precision:  0.44293193717277485\n",
      "recall:  0.8597560975609756\n",
      "Inference Time: 0.05008053779602051 sec\n",
      "\n",
      "AUPRC: 0.7471883996228341\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"logistic regression:\")\n",
    "# 参数搜索优化\n",
    "param_grid_LR = {\n",
    "    'penalty': ['l1', 'l2'],\n",
    "    'C': [0.1, 1.0, 10.0],\n",
    "    'solver': ['liblinear', 'saga']\n",
    "}\n",
    "best_para_LR, search_time_LR = gridSearchHelper(LogisticRegression(random_state=42), param_grid_LR,x_train_down, y_train_down)\n",
    "\n",
    "# 使用参数搜索的结果\n",
    "LR_model = LogisticRegression(**best_para_LR)\n",
    "LR_model, training_time_LR = trainHelper(LR_model,x_train_down, y_train_down)\n",
    "f1_LR,precision_LR,recall_LR,inference_time_LR  = inferenceHelper(LR_model, x_test, y_test)\n",
    "auprc_LR = plot_PRC(LR_model, x_test, y_test)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.2 决策树"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "decision tree:\n",
      "=== grid search ===\n",
      "grid search time:  160.09675478935242 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.09593653678894043 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.49461145774248433\n",
      "precision:  0.3430369787568843\n",
      "recall:  0.8861788617886179\n",
      "Inference Time: 0.05513930320739746 sec\n",
      "\n",
      "AUPRC: 0.3488049569191313\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"decision tree:\")\n",
    "param_grid_DT = {\n",
    "    'criterion': ['gini', 'entropy', \"log_loss\"],\n",
    "    'max_depth': [3, 5, 7, 9, None],\n",
    "    'min_samples_split': [2, 3, 5, 7, 10],\n",
    "    'min_samples_leaf': [1, 2, 3, 4],\n",
    "}\n",
    "\n",
    "best_para_DT, search_time_DT = gridSearchHelper(DecisionTreeClassifier(random_state=42),param_grid_DT,x_train_down, y_train_down)\n",
    "\n",
    "DT_model = DecisionTreeClassifier(**best_para_DT)\n",
    "DT_model, training_time_DT = trainHelper(DT_model, x_train_down, y_train_down)\n",
    "f1_DT, precision_DT, recall_DT,inference_time_DT  = inferenceHelper(DT_model, x_test, y_test)\n",
    "auprc_DT = plot_PRC(DT_model, x_test, y_test)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.3 随机森林"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Classifier\n",
      "=== grid search ===\n",
      "grid search time:  721.3495788574219 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  2.079598903656006 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.7102502017756255\n",
      "precision:  0.5890227576974565\n",
      "recall:  0.8943089430894309\n",
      "Inference Time: 1.432697057723999 sec\n",
      "\n",
      "AUPRC: 0.7978843542860654\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Random Forest Classifier\")\n",
    "param_grid_RF = {\n",
    "    'n_estimators': [50, 100, 200],\n",
    "    'max_depth': [5, 10, 15],\n",
    "    'min_samples_split': [2, 4, 6],\n",
    "    'min_samples_leaf': [1, 2, 4]\n",
    "}\n",
    "best_para_RF, search_time_RF = gridSearchHelper(RandomForestClassifier(random_state=42),param_grid_RF,x_train_down, y_train_down)\n",
    "\n",
    "RF_model = RandomForestClassifier(**best_para_RF)\n",
    "RF_model, training_time_RF = trainHelper(RF_model, x_train_down, y_train_down)\n",
    "f1_RF, precision_RF, recall_RF,inference_time_RF  = inferenceHelper(RF_model, x_test, y_test)\n",
    "auprc_RF = plot_PRC(RF_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.4 SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Support vector machine\n",
      "=== grid search ===\n",
      "grid search time:  90.71455121040344 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.42334675788879395 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.7102502017756255\n",
      "precision:  0.5890227576974565\n",
      "recall:  0.8943089430894309\n",
      "Inference Time: 1.4400532245635986 sec\n",
      "\n",
      "AUPRC: 0.7452344934950936\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Support vector machine\")\n",
    "param_grid_SVM = {\n",
    "    'C': [0.1, 1, 10],\n",
    "    'kernel': ['linear', 'rbf'],\n",
    "    'gamma': [0.1, 1, 10]\n",
    "}\n",
    "best_para_SVM, search_time_SVM = gridSearchHelper(SVC(random_state=42),param_grid_SVM,x_train_down, y_train_down)\n",
    "\n",
    "SVM_model = SVC(**best_para_SVM, probability=True)\n",
    "SVM_model, training_time_SVM = trainHelper(SVM_model, x_train_down, y_train_down)\n",
    "f1_SVM, precision_SVM, recall_SVM, inference_time_SVM  = inferenceHelper(RF_model, x_test, y_test)\n",
    "auprc_SVM = plot_PRC(SVM_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.5 KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== grid search ===\n",
      "grid search time:  1.0624923706054688 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.001859903335571289 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.6944248412138321\n",
      "precision:  0.5318918918918919\n",
      "recall:  1.0\n",
      "Inference Time: 1.1087913513183594 sec\n",
      "\n",
      "AUPRC: 0.8241273960248506\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid_KNN = {\n",
    "    'n_neighbors': [5, 10, 15], \n",
    "    'weights': ['uniform', 'distance']\n",
    "}\n",
    "\n",
    "best_para_KNN, search_time_KNN = gridSearchHelper(KNeighborsClassifier(), param_grid_KNN, x_train_down, y_train_down)\n",
    "\n",
    "KNN_model = KNeighborsClassifier(**best_para_KNN)\n",
    "KNN_model, training_time_KNN = trainHelper(KNN_model, x_train_down, y_train_down)\n",
    "f1_KNN, precision_KNN, recall_KNN,inference_time_KNN  = inferenceHelper(KNN_model, x_test, y_test)\n",
    "auprc_KNN = plot_PRC(KNN_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 以下使用oneAPI库加速"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Intel(R) Extension for Scikit-learn* enabled (https://github.com/intel/scikit-learn-intelex)\n"
     ]
    }
   ],
   "source": [
    "import modin.pandas as pd\n",
    "import modin.config as cfg\n",
    "cfg.StorageFormat.put('hdk')\n",
    "from sklearnex import patch_sklearn\n",
    "patch_sklearn()\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "# preprocess\n",
    "import time\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split, GridSearchCV\n",
    "from sklearn.utils import resample\n",
    "\n",
    "# model\n",
    "import xgboost as xgb\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "\n",
    "\n",
    "# socre\n",
    "from sklearn.metrics import average_precision_score\n",
    "from sklearn.metrics import precision_recall_curve\n",
    "from sklearn.metrics import f1_score, precision_score, recall_score, auc\n",
    "\n",
    "# figure\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 读取和预处理数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loading data...\n",
      "number of majority_class data:  284315\n",
      "number of minority_class data:  492\n",
      "fraction of minority class:  0.001727485630620034\n",
      "number of downSample data =  4920\n",
      "loading complete!\n"
     ]
    }
   ],
   "source": [
    "print(\"loading data...\")\n",
    "data = pd.read_csv('creditcard.csv')\n",
    "data = data.drop(columns=['Time'], axis=1)\n",
    "# Amount列值缩放到[-1,1]区间\n",
    "scaler = StandardScaler()\n",
    "data['Amount'] = scaler.fit_transform(data['Amount'].values.reshape(-1, 1))\n",
    "x = data.drop(columns=['Class'], axis=1) \n",
    "y = data['Class']\n",
    "\n",
    "x_test = x\n",
    "y_test = y\n",
    "# x_train, x_test, y_train, y_test = train_test_split(x, y,test_size=0.3, random_state=42)\n",
    "\n",
    "# 对数据下采样，比列为1:9\n",
    "majority_class = data[data.Class == 0]\n",
    "minority_class = data[data.Class == 1]\n",
    "print(\"number of majority_class data: \", len(majority_class))\n",
    "print(\"number of minority_class data: \", len(minority_class))\n",
    "print(\"fraction of minority class: \", len(minority_class)/len(data))\n",
    "majority_downsampled = resample(majority_class, replace=False, n_samples=len(minority_class)*9, random_state=123)\n",
    "majority_downsampled = pd.DataFrame(majority_downsampled, columns=data.columns) \n",
    "downsampled_data = pd.concat([majority_downsampled, minority_class])\n",
    "\n",
    "x_train_down = downsampled_data.drop(columns=['Class'], axis=1)\n",
    "y_train_down = downsampled_data[\"Class\"]\n",
    "\n",
    "print(\"number of downSample data = \", len(downsampled_data))\n",
    "print(\"loading complete!\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.1 逻辑回归"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "logistic regression:\n",
      "=== grid search ===\n",
      "grid search time:  5.367825984954834 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.012550115585327148 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.5846579129232896\n",
      "precision:  0.44293193717277485\n",
      "recall:  0.8597560975609756\n",
      "Inference Time: 0.038216590881347656 sec\n",
      "\n",
      "AUPRC: 0.7471654231422259\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"logistic regression:\")\n",
    "# 参数搜索优化\n",
    "param_grid_LR = {\n",
    "    'penalty': ['l1', 'l2'],\n",
    "    'C': [0.1, 1.0, 10.0],\n",
    "    'solver': ['liblinear', 'saga']\n",
    "}\n",
    "best_para_LR, search_time_LR_a = gridSearchHelper(LogisticRegression(random_state=42), param_grid_LR,x_train_down, y_train_down)\n",
    "\n",
    "# 使用参数搜索的结果\n",
    "LR_model = LogisticRegression(**best_para_LR)\n",
    "LR_model, training_time_LR_a = trainHelper(LR_model,x_train_down, y_train_down)\n",
    "f1_LR_a, precision_LR_a, recall_LR_a, inference_time_LR_a  = inferenceHelper(LR_model, x_test, y_test)\n",
    "auprc_LR_a = plot_PRC(LR_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.2 决策树"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "decision tree:\n",
      "=== grid search ===\n",
      "grid search time:  162.67543506622314 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.09676599502563477 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.49573621375781696\n",
      "precision:  0.3441199684293607\n",
      "recall:  0.8861788617886179\n",
      "Inference Time: 0.049069881439208984 sec\n",
      "\n",
      "AUPRC: 0.35006200482849215\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkIAAAHHCAYAAABTMjf2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAAA7Y0lEQVR4nO3deVyVZf7/8TcgHFABTQQVT+KapaaGSmhGNigu2ThTamqKtmipMyVtaiWZk2ipaWkxWS4zWW5ZY7mNklYWfSu3qcZ9SctAzQQFZTvX7w9/nOkEKiBwgPv1fDzO49G5znXd9+c+t8x5z33di4cxxggAAMCCPN1dAAAAgLsQhAAAgGURhAAAgGURhAAAgGURhAAAgGURhAAAgGURhAAAgGURhAAAgGURhAAAgGURhIAyMHz4cIWFhRVrzJYtW+Th4aEtW7aUSU2V3W233abbbrvN+f7IkSPy8PDQokWL3FYTgMqPIIQqYdGiRfLw8HC+fH191aJFC40dO1apqanuLq/Cyw8V+S9PT09dc8016tWrl5KTk91dHoppwIAB8vDw0FNPPVXo5/mhO//l7e2tJk2aaNiwYTp06FCp1OBwOPTiiy+qcePG8vX11Y033qh33323SGM//fRT3XnnnbLb7fL19VW9evXUs2dPff755wX63nbbbS7bkv/q2bNnqWwHqr5q7i4AKE3PP/+8GjdurAsXLmjr1q16/fXXtXbtWn333XeqXr16udUxf/58ORyOYo259dZbdf78efn4+JRRVVc2aNAg9e7dW3l5edq3b59ee+01devWTV9//bXatGnjtrpQdOnp6frwww8VFhamd999V9OmTZOHh0ehff/617+qY8eOysnJ0fbt2/XGG29ozZo1+vbbb9WgQYOrquPpp5/WtGnT9OCDD6pjx47617/+pcGDB8vDw0P33HPPZcfu27dPnp6eeuihh1SvXj39+uuvevvtt3XrrbdqzZo1BUJOw4YNlZCQ4NJ2tfXDQgxQBSxcuNBIMl9//bVLe1xcnJFk3nnnnUuOPXfuXFmXV+EdPnzYSDIvvfSSS/u6deuMJPPwww+7qbL/iYqKMlFRUc73+TUvXLjQbTXly8jIcHcJTgsWLDDe3t7m448/NpLMli1bCvTZvHmzkWRWrFjh0v7KK68YSWbq1KlXVcOPP/5ovL29zZgxY5xtDofDdO3a1TRs2NDk5uYWe5kZGRkmJCTExMTEuLRHRUWZVq1aXVW9sDamxlCl3X777ZKkw4cPS7p47k7NmjV18OBB9e7dW/7+/hoyZIiki4fyZ8+erVatWsnX11chISEaNWqUfv311wLLXbdunaKiouTv76+AgAB17NhR77zzjvPzws4RWrp0qcLDw51j2rRpozlz5jg/v9Q5QitWrFB4eLj8/PwUFBSke++9Vz/99JNLn/zt+umnn9SvXz/VrFlTdevW1eOPP668vLwSf39du3aVJB08eNCl/cyZM3r00Udlt9tls9nUrFkzTZ8+vcBRMIfDoTlz5qhNmzby9fVV3bp11bNnT33zzTfOPgsXLtTtt9+u4OBg2Ww23XDDDXr99ddLXHNhzpw5o3HjxiksLEw2m00NGzbUsGHDdOrUKUn/m1o9cuSIy7jC9sltt92m1q1ba9u2bbr11ltVvXp1TZw4UXfccYeaNGlS6PojIyPVoUMHl7a3337buV+vueYa3XPPPTp27JhLn8zMTO3Zs8dZZ1EsWbJE3bt3V7du3XT99ddryZIlRR77+7+XkvrXv/6lnJwcjR492tnm4eGhhx9+WD/++GOJplurV6+uunXr6syZM4V+npubq3PnzpW0ZFgYQQhVWv4PeJ06dZxtubm5iomJUXBwsGbMmKG77rpLkjRq1Cg98cQT6tKli+bMmaMRI0ZoyZIliomJUU5OjnP8okWL1KdPH50+fVoTJkzQtGnT1K5dO61fv/6SdWzcuFGDBg1S7dq1NX36dE2bNk233XZboec8/NaiRYs0YMAAeXl5KSEhQQ8++KBWrVqlW265pcAPQl5enmJiYlSnTh3NmDFDUVFRmjlzpt54443ifm1O+cGgdu3azrbMzExFRUXp7bff1rBhw/TKK6+oS5cumjBhguLi4lzG33///c7ANH36dI0fP16+vr768ssvnX1ef/11NWrUSBMnTtTMmTNlt9s1evRozZs3r8R1/9a5c+fUtWtXvfrqq+rRo4fmzJmjhx56SHv27NGPP/5YomX+8ssv6tWrl9q1a6fZs2erW7duGjhwoA4fPqyvv/7ape8PP/ygL7/80mU66IUXXtCwYcPUvHlzzZo1S48++qiSkpJ06623uuzXr776Stdff73mzp1bpLqOHz+uzZs3a9CgQZIuTnWuXLlS2dnZRRpf2N/LqVOnivTKyspyjtmxY4dq1Kih66+/3mX5nTp1cn5eFOnp6Tp16pT27NmjiRMn6rvvvtMf/vCHAv327dunGjVqyN/fX/Xq1dOzzz7r8jcLXJa7D0kBpSF/amzTpk3m5MmT5tixY2bp0qWmTp06xs/Pz/z444/GGGNiY2ONJDN+/HiX8Z999pmRZJYsWeLSvn79epf2M2fOGH9/fxMREWHOnz/v0tfhcDj/OzY21jRq1Mj5/pFHHjEBAQGXnRLIn67YvHmzMcaY7OxsExwcbFq3bu2yro8++shIMpMmTXJZnyTz/PPPuyyzffv2Jjw8/JLrzJc/zTR58mRz8uRJk5KSYj777DPTsWPHAlMoU6ZMMTVq1DD79u1zWcb48eONl5eXOXr0qDHGOKdm/vrXvxZY32+/q8zMzAKfx8TEmCZNmri0lXRqbNKkSUaSWbVq1SXryP/3c/jwYZfPf79P8uuQZBITE136pqWlGZvNZh577DGX9hdffNF4eHiYH374wRhjzJEjR4yXl5d54YUXXPp9++23plq1ai7t+euPj4+/7DbmmzFjhvHz8zPp6enGGGP27dtnJJn333+/0O1asGCBOXnypDl+/LhZs2aNCQsLMx4eHi5TzJKK9PrtfujTp0+B/WfMxemtwv7+LiUmJsa5fB8fHzNq1KgCf3f33Xefee6558x7771n/vGPf5g777zTSDIDBgwo0joATpZGlRIdHe3yvlGjRlqyZIlCQ0Nd2h9++GGX9ytWrFBgYKC6d+/uMg0RHh6umjVravPmzRo8eLA2btyos2fPOo9s/NalTkiVpFq1aikjI0MbN24s8tUs33zzjU6cOKHnnnvOZV19+vRRy5YttWbNGk2ePNllzEMPPeTyvmvXrvrnP/9ZpPVJUnx8vOLj453va9asqZkzZ+ruu+92tq1YsUJdu3ZV7dq1Xb6r6OhoTZs2TZ9++qmGDBmi9957Tx4eHi7Ly/fb78rPz8/532lpacrJyVFUVJQ2bNigtLQ0BQYGFrn+wrz33ntq27at/vSnP122juKw2WwaMWKES1tAQIB69eql5cuX66WXXnIue9myZbr55pt17bXXSpJWrVolh8OhAQMGuHx/9erVU/PmzbV582ZNnDhR0sVpOGNMketasmSJ+vTpI39/f0lS8+bNFR4eriVLlqhfv34F+t93330u7+vWravFixe7TONt3LixSOtu1aqV87/Pnz8vm81WoE/+v+Pz588XaZnTpk3TY489pmPHjmnx4sXKzs5Wbm6uS5+33nrL5f3QoUM1cuRIzZ8/X+PGjdPNN99cpHXBughCqFLmzZunFi1aqFq1agoJCdF1110nT0/XGeBq1aqpYcOGLm379+9XWlqagoODC13uiRMnJP1v6qB169bFqmv06NFavny5evXqpdDQUPXo0UMDBgy4bCj64YcfJEnXXXddgc9atmyprVu3urTln4PzW7Vr13Y5x+nkyZMu5wzVrFlTNWvWdL4fOXKk+vfvrwsXLujjjz/WK6+8UuAco/379+s///lPgXXl++131aBBA11zzTWX3EZJ+vzzzxUfH6/k5GRlZma6fFYaQejgwYPO6c/SEhoaWujVfQMHDtQHH3yg5ORkde7cWQcPHtS2bds0e/ZsZ5/9+/fLGKPmzZsXumxvb+8S1bR7927t2LFDw4YN04EDB5ztt912m+bNm6f09HQFBAS4jJk0aZK6du0qLy8vBQUF6frrr1e1aq4/C7//PxdF4efn5zJVlu/ChQvOz4uiXbt2zv++9957ddNNN2n48OFauXLlZcc99thjmj9/vjZt2kQQwhURhFCldOrUqcBJqb9ns9kKhCOHw6Hg4OBLnlh6qR/9ogoODtbOnTu1YcMGrVu3TuvWrdPChQs1bNgwLV68+KqWnc/Ly+uKfTp27OgMWNLFI0DPPfec833z5s2dP3x33HGHvLy8NH78eHXr1s35vTocDnXv3l1PPvlkoeto0aJFkWs+ePCg/vCHP6hly5aaNWuW7Ha7fHx8tHbtWr388svFvgVBSV3qyNClTjS/1A953759Vb16dS1fvlydO3fW8uXL5enpqf79+zv7OBwOeXh4aN26dYXus98G0+J4++23JUnjxo3TuHHjCnz+3nvvFTiK1aZNmysGnZSUlCKtPzAw0Pm91K9fX5s3b5YxxuW7/fnnnyWV7NJ2Hx8f3XnnnZo2bZrOnz9/2TBlt9slSadPny72emA9BCFAUtOmTbVp0yZ16dLlsv8D27RpU0nSd999p2bNmhVrHT4+Purbt6/69u0rh8Oh0aNH6+9//7ueffbZQpfVqFEjSdLevXudV/Pk27t3r/Pz4liyZInLtMSlrnLK9/TTT2v+/Pl65plnnCeDN23aVOfOnbviD2jTpk21YcMGnT59+pJHhT788ENlZWVp9erVzqkjSdq8eXNRN+mKmjZtqu++++6yffJPBv/9Cei/DY1FUaNGDd1xxx1asWKFZs2apWXLlqlr164uP/xNmzaVMUaNGzcuVmi8HGOM3nnnHXXr1s3lSq18U6ZM0ZIlSwoEoaKoX79+kfotXLhQw4cPl3TxSM6bb76p3bt364YbbnD2+b//+z/n5yVx/vx5GWN09uzZy/6d5t8U8mr/DwysgavGAF28E29eXp6mTJlS4LPc3FznD2SPHj3k7++vhIQE52H+fJc7l+OXX35xee/p6akbb7xRkgqdQpCkDh06KDg4WImJiS591q1bp927d6tPnz5F2rbf6tKli6Kjo52vKwWhWrVqadSoUdqwYYN27twp6eJ3lZycrA0bNhTof+bMGec5HHfddZeMMQXOY5L+913lHxH57XeXlpamhQsXFnvbLuWuu+7Srl279P7771+yjvyA++mnnzo/y8vLK9EVdwMHDtTx48f15ptvateuXRo4cKDL53/+85/l5eWlyZMnF/g3Y4xx+bdS1MvnP//8cx05ckQjRozQ3XffXeA1cOBAbd68WcePHy/29mzcuLFIr5iYGOeYP/7xj/L29tZrr73msm2JiYkKDQ1V586dne0///yz9uzZ43KVV/706m+dOXNG7733nux2u3MKOz09vcDfjzFGf/vb3yTJpSbgUjgiBEiKiorSqFGjlJCQoJ07d6pHjx7y9vbW/v37tWLFCs2ZM0d33323AgIC9PLLL+uBBx5Qx44dNXjwYNWuXVu7du1SZmbmJae5HnjgAZ0+fVq33367GjZsqB9++EGvvvqq2rVrV+AS43ze3t6aPn26RowYoaioKA0aNEipqamaM2eOwsLCCp3+KAuPPPKIZs+erWnTpmnp0qV64okntHr1at1xxx0aPny4wsPDlZGRoW+//VYrV67UkSNHFBQUpG7dumno0KF65ZVXtH//fvXs2VMOh0OfffaZunXrprFjx6pHjx7OI2WjRo3SuXPnNH/+fAUHBzunUa7WE088oZUrV6p///667777FB4ertOnT2v16tVKTExU27Zt1apVK918882aMGGC8wjW0qVLC5yYWxT596d6/PHH5eXlVeD8pKZNm+pvf/ubJkyYoCNHjqhfv37y9/fX4cOH9f7772vkyJF6/PHHJV28fL5bt24FpjB/b8mSJfLy8rpkOL7zzjv19NNPa+nSpQVucXAlJTlHqGHDhnr00Uf10ksvKScnRx07dtQHH3ygzz77zFlrvgkTJmjx4sU6fPiw895bvXr1UsOGDRUREaHg4GAdPXpUCxcu1PHjx7Vs2TLn2O3bt2vQoEEaNGiQmjVrpvPnz+v999/X559/rpEjR+qmm24qdu2wIDdcqQaUukvdWfr3YmNjTY0aNS75+RtvvGHCw8ONn5+f8ff3N23atDFPPvmkOX78uEu/1atXm86dOxs/Pz8TEBBgOnXqZN59912X9fz28vmVK1eaHj16mODgYOPj42OuvfZaM2rUKPPzzz87+xR2qbYxxixbtsy0b9/e2Gw2c80115ghQ4Y4bwdwpe2Kj483Rfkzv9SdpfMNHz7ceHl5mQMHDhhjjDl79qyZMGGCadasmfHx8TFBQUGmc+fOZsaMGSY7O9s5Ljc317z00kumZcuWxsfHx9StW9f06tXLbNu2zeW7vPHGG42vr68JCwsz06dPNwsWLChwOfvV3Fn6l19+MWPHjjWhoaHGx8fHNGzY0MTGxppTp045+xw8eNBER0cbm81mQkJCzMSJE83GjRsLvXz+SncyHjJkiJFkoqOjL9nnvffeM7fccoupUaOGqVGjhmnZsqUZM2aM2bt3r7NPUS6fz87ONnXq1DFdu3a9bE2NGzc27du3d1nu7+8sXZry8vLM1KlTTaNGjYyPj49p1aqVefvttwv0y7/1w2/39dy5c80tt9xigoKCTLVq1UzdunVN3759zaeffuoy9tChQ6Z///4mLCzM+Pr6murVq5vw8HCTmJjocosG4HI8jCnGtZkAAABVCOcIAQAAyyIIAQAAyyIIAQAAyyIIAQAAyyIIAQAAyyIIAQAAy7LcDRUdDoeOHz8uf3//Ej95GgAAlC/z/x+v0qBBgwLPi7walgtCx48fdz6QDwAAVC7Hjh1Tw4YNS215lgtC/v7+ki5+kQEBAW6uBgAAFEV6errsdrvzd7y0WC4I5U+HBQQEEIQAAKhkSvu0Fk6WBgAAlkUQAgAAlkUQAgAAlkUQAgAAlkUQAgAAlkUQAgAAlkUQAgAAlkUQAgAAlkUQAgAAlkUQAgAAluXWIPTpp5+qb9++atCggTw8PPTBBx9cccyWLVt00003yWazqVmzZlq0aFGZ1wkAAKomtwahjIwMtW3bVvPmzStS/8OHD6tPnz7q1q2bdu7cqUcffVQPPPCANmzYUMaVAgCAqsitD13t1auXevXqVeT+iYmJaty4sWbOnClJuv7667V161a9/PLLiomJKasyAQBAFVWpzhFKTk5WdHS0S1tMTIySk5PdVBEAAKjM3HpEqLhSUlIUEhLi0hYSEqL09HSdP39efn5+BcZkZWUpKyvL+T49PV2S5HCUba0AAKDiq1RHhEoiISFBgYGBzpfdbpcknT3r5sIAAIDbVaogVK9ePaWmprq0paamKiAgoNCjQZI0YcIEpaWlOV/Hjh2TJGVnl3m5AACggqtUU2ORkZFau3atS9vGjRsVGRl5yTE2m002m62sSwMAAJWQW48InTt3Tjt37tTOnTslXbw8fufOnTp69Kiki0dzhg0b5uz/0EMP6dChQ3ryySe1Z88evfbaa1q+fLnGjRvnjvIBAEAl59Yg9M0336h9+/Zq3769JCkuLk7t27fXpEmTJEk///yzMxRJUuPGjbVmzRpt3LhRbdu21cyZM/Xmm29y6TwAACgRD2OMcXcR5Sk9PV2BgYE6cCBNTZsGuLscAABQBPm/32lpaQoIKL3f70p1sjQAAEBpIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLIggBAADLcnsQmjdvnsLCwuTr66uIiAh99dVXl+0/e/ZsXXfddfLz85Pdbte4ceN04cKFcqoWAABUJW4NQsuWLVNcXJzi4+O1fft2tW3bVjExMTpx4kSh/d955x2NHz9e8fHx2r17t9566y0tW7ZMEydOLOfKAQBAVeDWIDRr1iw9+OCDGjFihG644QYlJiaqevXqWrBgQaH9v/jiC3Xp0kWDBw9WWFiYevTooUGDBl3xKBIAAEBh3BaEsrOztW3bNkVHR/+vGE9PRUdHKzk5udAxnTt31rZt25zB59ChQ1q7dq169+59yfVkZWUpPT3d5QUAACBJ1dy14lOnTikvL08hISEu7SEhIdqzZ0+hYwYPHqxTp07plltukTFGubm5euihhy47NZaQkKDJkyeXau0AAKBqcPvJ0sWxZcsWTZ06Va+99pq2b9+uVatWac2aNZoyZcolx0yYMEFpaWnO17Fjx8qxYgAAUJG57YhQUFCQvLy8lJqa6tKempqqevXqFTrm2Wef1dChQ/XAAw9Iktq0aaOMjAyNHDlSTz/9tDw9C+Y6m80mm81W+hsAAAAqPbcdEfLx8VF4eLiSkpKcbQ6HQ0lJSYqMjCx0TGZmZoGw4+XlJUkyxpRdsQAAoEpy2xEhSYqLi1NsbKw6dOigTp06afbs2crIyNCIESMkScOGDVNoaKgSEhIkSX379tWsWbPUvn17RURE6MCBA3r22WfVt29fZyACAAAoKrcGoYEDB+rkyZOaNGmSUlJS1K5dO61fv955AvXRo0ddjgA988wz8vDw0DPPPKOffvpJdevWVd++ffXCCy+4axMAAEAl5mEsNqeUnp6uwMBAHTiQpqZNA9xdDgAAKIL83++0tDQFBJTe73elumoMAACgNBGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZbk9CM2bN09hYWHy9fVVRESEvvrqq8v2P3PmjMaMGaP69evLZrOpRYsWWrt2bTlVCwAAqpJq7lz5smXLFBcXp8TEREVERGj27NmKiYnR3r17FRwcXKB/dna2unfvruDgYK1cuVKhoaH64YcfVKtWrfIvHgAAVHoexhjjrpVHRESoY8eOmjt3riTJ4XDIbrfrL3/5i8aPH1+gf2Jiol566SXt2bNH3t7eJVpnenq6AgMDdeBAmpo2Dbiq+gEAQPnI//1OS0tTQEDp/X67bWosOztb27ZtU3R09P+K8fRUdHS0kpOTCx2zevVqRUZGasyYMQoJCVHr1q01depU5eXllVfZAACgCinR1FheXp4WLVqkpKQknThxQg6Hw+Xzjz/++IrLOHXqlPLy8hQSEuLSHhISoj179hQ65tChQ/r44481ZMgQrV27VgcOHNDo0aOVk5Oj+Pj4QsdkZWUpKyvL+T49Pf2KtQEAAGsoURB65JFHtGjRIvXp00etW7eWh4dHaddVKIfDoeDgYL3xxhvy8vJSeHi4fvrpJ7300kuXDEIJCQmaPHlyudQHAAAqlxIFoaVLl2r58uXq3bt3iVccFBQkLy8vpaamurSnpqaqXr16hY6pX7++vL295eXl5Wy7/vrrlZKSouzsbPn4+BQYM2HCBMXFxTnfp6eny263l7huAABQdZToHCEfHx81a9bsqlbs4+Oj8PBwJSUlOdscDoeSkpIUGRlZ6JguXbrowIEDLlNx+/btU/369QsNQZJks9kUEBDg8gIAAJBKGIQee+wxzZkzR1d7wVlcXJzmz5+vxYsXa/fu3Xr44YeVkZGhESNGSJKGDRumCRMmOPs//PDDOn36tB555BHt27dPa9as0dSpUzVmzJirqgMAAFhTiabGtm7dqs2bN2vdunVq1apVgUvZV61aVaTlDBw4UCdPntSkSZOUkpKidu3aaf369c4TqI8ePSpPz/9lNbvdrg0bNmjcuHG68cYbFRoaqkceeURPPfVUSTYDAABYXInuI5R/xOZSFi5cWOKCyhr3EQIAoPIpq/sIleiIUEUOOgAAAEV1VY/YOHnypPbu3StJuu6661S3bt1SKQoAAKA8lOhk6YyMDN13332qX7++br31Vt16661q0KCB7r//fmVmZpZ2jQAAAGWiREEoLi5On3zyiT788EOdOXNGZ86c0b/+9S998skneuyxx0q7RgAAgDJRopOlg4KCtHLlSt12220u7Zs3b9aAAQN08uTJ0qqv1HGyNAAAlU+FeuhqZmZmgWeESVJwcDBTYwAAoNIoURCKjIxUfHy8Lly44Gw7f/68Jk+efMm7QgMAAFQ0JbpqbM6cOYqJiVHDhg3Vtm1bSdKuXbvk6+urDRs2lGqBAAAAZaVEQah169bav3+/lixZoj179kiSBg0apCFDhsjPz69UCwQAACgrJb6PUPXq1fXggw+WZi0AAADlqshBaPXq1erVq5e8vb21evXqy/a98847r7owAACAslbky+c9PT2VkpKi4OBglwehFligh4fy8vJKrcDSxuXzAABUPm5/1pjD4Sj0vwEAACqrEl0+X5gzZ86U1qIAAADKRYmC0PTp07Vs2TLn+/79++uaa65RaGiodu3aVWrFAQAAlKUSBaHExETZ7XZJ0saNG7Vp0yatX79evXr10hNPPFGqBQIAAJSVEl0+n5KS4gxCH330kQYMGKAePXooLCxMERERpVogAABAWSnREaHatWvr2LFjkqT169crOjpakmSMqdBXjAEAAPxWiY4I/fnPf9bgwYPVvHlz/fLLL+rVq5ckaceOHWrWrFmpFggAAFBWShSEXn75ZYWFhenYsWN68cUXVbNmTUnSzz//rNGjR5dqgQAAAGWlyDdUrCq4oSIAAJWP22+oyCM2AABAVcMjNgAAQIXn9iNCPGIDAABUNaX2iA0AAIDKpkRB6K9//ateeeWVAu1z587Vo48+erU1AQAAlIsSBaH33ntPXbp0KdDeuXNnrVy58qqLAgAAKA8lCkK//PKLAgMDC7QHBATo1KlTV10UAABAeShREGrWrJnWr19foH3dunVq0qTJVRcFAABQHkp0Z+m4uDiNHTtWJ0+e1O233y5JSkpK0syZMzV79uzSrA8AAKDMlCgI3XfffcrKytILL7ygKVOmSJLCwsL0+uuva9iwYaVaIAAAQFm56kdsnDx5Un5+fs7njVV03FARAIDKp6xuqFji+wjl5uZq06ZNWrVqlfKz1PHjx3Xu3LlSKw4AAKAslWhq7IcfflDPnj119OhRZWVlqXv37vL399f06dOVlZWlxMTE0q4TAACg1JXoiNAjjzyiDh066Ndff5Wfn5+z/U9/+pOSkpJKrTgAAICyVKIjQp999pm++OIL+fj4uLSHhYXpp59+KpXCAAAAylqJjgg5HI5CnzD/448/yt/f/6qLAgAAKA8lCkI9evRwuV+Qh4eHzp07p/j4ePXu3bu0agMAAChTJZoamzFjhnr27KkbbrhBFy5c0ODBg7V//34FBQXp3XffLe0aAQAAykSJgpDdbteuXbu0bNky7dq1S+fOndP999+vIUOGuJw8DQAAUJEVOwjl5OSoZcuW+uijjzRkyBANGTKkLOoCAAAoc8U+R8jb21sXLlwoi1oAAADKVYlOlh4zZoymT5+u3Nzc0q4HAACg3JToHKGvv/5aSUlJ+ve//602bdqoRo0aLp+vWrWqVIoDAAAoSyUKQrVq1dJdd91V2rUAAACUq2IFIYfDoZdeekn79u1Tdna2br/9dj333HNcKQYAACqlYp0j9MILL2jixImqWbOmQkND9corr2jMmDFlVRsAAECZKlYQ+sc//qHXXntNGzZs0AcffKAPP/xQS5YskcPhKKv6AAAAykyxgtDRo0ddHqERHR0tDw8PHT9+vNQLAwAAKGvFCkK5ubny9fV1afP29lZOTk6pFgUAAFAeinWytDFGw4cPl81mc7ZduHBBDz30kMsl9Fw+DwAAKoNiBaHY2NgCbffee2+pFQMAAFCeihWEFi5cWFZ1AAAAlLsSPWIDAACgKiAIAQAAyyIIAQAAyyIIAQAAyyIIAQAAyyIIAQAAyyIIAQAAy6oQQWjevHkKCwuTr6+vIiIi9NVXXxVp3NKlS+Xh4aF+/fqVbYEAAKBKcnsQWrZsmeLi4hQfH6/t27erbdu2iomJ0YkTJy477siRI3r88cfVtWvXcqoUAABUNW4PQrNmzdKDDz6oESNG6IYbblBiYqKqV6+uBQsWXHJMXl6ehgwZosmTJ6tJkyblWC0AAKhK3BqEsrOztW3bNkVHRzvbPD09FR0dreTk5EuOe/755xUcHKz777//iuvIyspSenq6ywsAAEBycxA6deqU8vLyFBIS4tIeEhKilJSUQsds3bpVb731lubPn1+kdSQkJCgwMND5stvtV103AACoGor10FV3O3v2rIYOHar58+crKCioSGMmTJiguLg45/v09HTZ7XY5HJLDUVaVAgBQPjzdfpJL5ebWIBQUFCQvLy+lpqa6tKempqpevXoF+h88eFBHjhxR3759nW2O/59mqlWrpr1796pp06YuY2w2m2w2W4Fl7dolXeF8bAAAKryAAKlNG3dXUXm5NQj5+PgoPDxcSUlJzkvgHQ6HkpKSNHbs2AL9W7ZsqW+//dal7ZlnntHZs2c1Z86cYk17HTok/fzzVZUPAIDb+fpKzZpJfn7urqRycvvUWFxcnGJjY9WhQwd16tRJs2fPVkZGhkaMGCFJGjZsmEJDQ5WQkCBfX1+1bt3aZXytWrUkqUD7lbRoIdWpUyqbAABAuXM4pN27pQsXpLw8d1dTebk9CA0cOFAnT57UpEmTlJKSonbt2mn9+vXOE6iPHj0qzzKYAK1e/eILAIDKyOGQvL2lnBx3V1K5eRhjjLuLKE/p6ekKDAzUhg1pqlMnwN3lAABQIg6H9J//SJmZ0ogRUs2a7q6obOX/fqelpSkgoPR+vznXHAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCAAAWBZBCACASs7huPhC8bn96fMAAODqfPONZLNJAQFSmzburqZyIQgBAFAJeXpKNWpIp09L339/sc3XV2rWTPLzc29tlQlBCACASqp5c8lul3Jzpd27pQsXpLw8d1dVuRCEAACoxHx9L54f5O0t5eS4u5rKh5OlAQCAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZVWIIDRv3jyFhYXJ19dXERER+uqrry7Zd/78+eratatq166t2rVrKzo6+rL9AQAALsXtQWjZsmWKi4tTfHy8tm/frrZt2yomJkYnTpwotP+WLVs0aNAgbd68WcnJybLb7erRo4d++umncq4cAABUdh7GGOPOAiIiItSxY0fNnTtXkuRwOGS32/WXv/xF48ePv+L4vLw81a5dW3PnztWwYcOu2D89PV2BgYHasCFNdeoEXHX9AAC4m8Mh/ec/UmamNGKEVLOmuysqffm/32lpaQoIKL3fb7ceEcrOzta2bdsUHR3tbPP09FR0dLSSk5OLtIzMzEzl5OTommuuKfTzrKwspaenu7wAAAAkNwehU6dOKS8vTyEhIS7tISEhSklJKdIynnrqKTVo0MAlTP1WQkKCAgMDnS+73X7VdQMAgKrB7ecIXY1p06Zp6dKlev/99+Xr61tonwkTJigtLc35OnbsWDlXCQAAKqpq7lx5UFCQvLy8lJqa6tKempqqevXqXXbsjBkzNG3aNG3atEk33njjJfvZbDbZbLZSqRcAAFQtbj0i5OPjo/DwcCUlJTnbHA6HkpKSFBkZeclxL774oqZMmaL169erQ4cO5VEqAACogtx6REiS4uLiFBsbqw4dOqhTp06aPXu2MjIyNGLECEnSsGHDFBoaqoSEBEnS9OnTNWnSJL3zzjsKCwtznktUs2ZN1ayKp8kDAIAy4/YgNHDgQJ08eVKTJk1SSkqK2rVrp/Xr1ztPoD569Kg8Pf934Or1119Xdna27r77bpflxMfH67nnnivP0gEAQCXn9vsIlTfuIwQAqGq4j1DJVeqrxgAAAK4GQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFgWQQgAAFhWNXcXAAAASo/DcfH1W54c9rgkghAAAFXIN99INptrW0CA1KaNe+qp6AhCAABUcp6eUo0a0unT0vffF/zc11dq1kzy8yv/2io6ghAAAFVA8+aS3e46LeZwSLt3SxcuSHl57qutIiMIAQBQRfj6ur53OCRvbyknxz31VAacPgUAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyLIAQAACyrmrsLAAAAZc/huPgqDk8LHC4hCAEAYAHffCPZbMUbExAgtWlTNvVUFAQhAACqKE9PqUYN6fRp6fvviz/e11dq1kzy8yv92ioKghAAAFVY8+aS3V68aTGHQ9q9W7pwQcrLK7vaKgKCEAAAVZyvb/H6OxySt7eUk1M29VQkFjgNCgAAoHAEIQAAYFkVIgjNmzdPYWFh8vX1VUREhL766qvL9l+xYoVatmwpX19ftWnTRmvXri2nSgEAQFXi9iC0bNkyxcXFKT4+Xtu3b1fbtm0VExOjEydOFNr/iy++0KBBg3T//fdrx44d6tevn/r166fvvvuunCsHAACVnYcxxrizgIiICHXs2FFz586VJDkcDtntdv3lL3/R+PHjC/QfOHCgMjIy9NFHHznbbr75ZrVr106JiYlXXF96eroCAwO1YUOa6tQJKL0NAQCginA4pP/8R8rMlGJjpZo13V3Rxd/v2rUDlZaWpoCA0vv9dutVY9nZ2dq2bZsmTJjgbPP09FR0dLSSk5MLHZOcnKy4uDiXtpiYGH3wwQdlWSoAAJZUkhsxloWMjLJZrluD0KlTp5SXl6eQkBCX9pCQEO3Zs6fQMSkpKYX2T0lJKbR/VlaWsrKynO/T0tIkSb/+mn41pQMAUKXl5EhnzlwMQhXBhQsXf7dLeyKryt9HKCEhQZMnTy7Qfs89djdUAwAArsYvv/yiwMDAUlueW4NQUFCQvLy8lJqa6tKempqqevXqFTqmXr16xeo/YcIEl6m0M2fOqFGjRjp69GipfpEovvT0dNntdh07dqxU53tRMuyPioN9UXGwLyqOtLQ0XXvttbrmmmtKdbluDUI+Pj4KDw9XUlKS+vXrJ+niydJJSUkaO3ZsoWMiIyOVlJSkRx991Nm2ceNGRUZGFtrfZrPJVsjkZmBgIP+oK4iAgAD2RQXC/qg42BcVB/ui4vD0LN0L3t0+NRYXF6fY2Fh16NBBnTp10uzZs5WRkaERI0ZIkoYNG6bQ0FAlJCRIkh555BFFRUVp5syZ6tOnj5YuXapvvvlGb7zxhjs3AwAAVEJuD0IDBw7UyZMnNWnSJKWkpKhdu3Zav36984Too0ePuqS/zp0765133tEzzzyjiRMnqnnz5vrggw/UunVrd20CAACopNwehCRp7Nixl5wK27JlS4G2/v37q3///iVal81mU3x8fKHTZShf7IuKhf1RcbAvKg72RcVRVvvC7TdUBAAAcBe3P2IDAADAXQhCAADAsghCAADAsghCAADAsqpkEJo3b57CwsLk6+uriIgIffXVV5ftv2LFCrVs2VK+vr5q06aN1q5dW06VVn3F2Rfz589X165dVbt2bdWuXVvR0dFX3HconuL+beRbunSpPDw8nDc+xdUr7r44c+aMxowZo/r168tms6lFixb8b1UpKe6+mD17tq677jr5+fnJbrdr3LhxunDhQjlVW3V9+umn6tu3rxo0aCAPD48iPUx9y5Ytuummm2Sz2dSsWTMtWrSo+Cs2VczSpUuNj4+PWbBggfn+++/Ngw8+aGrVqmVSU1ML7f/5558bLy8v8+KLL5r//ve/5plnnjHe3t7m22+/LefKq57i7ovBgwebefPmmR07dpjdu3eb4cOHm8DAQPPjjz+Wc+VVU3H3R77Dhw+b0NBQ07VrV/PHP/6xfIqt4oq7L7KyskyHDh1M7969zdatW83hw4fNli1bzM6dO8u58qqnuPtiyZIlxmazmSVLlpjDhw+bDRs2mPr165tx48aVc+VVz9q1a83TTz9tVq1aZSSZ999//7L9Dx06ZKpXr27i4uLMf//7X/Pqq68aLy8vs379+mKtt8oFoU6dOpkxY8Y43+fl5ZkGDRqYhISEQvsPGDDA9OnTx6UtIiLCjBo1qkzrtILi7ovfy83NNf7+/mbx4sVlVaKllGR/5Obmms6dO5s333zTxMbGEoRKSXH3xeuvv26aNGlisrOzy6tEyyjuvhgzZoy5/fbbXdri4uJMly5dyrROqylKEHryySdNq1atXNoGDhxoYmJiirWuKjU1lp2drW3btik6OtrZ5unpqejoaCUnJxc6Jjk52aW/JMXExFyyP4qmJPvi9zIzM5WTk1PqD9izopLuj+eff17BwcG6//77y6NMSyjJvli9erUiIyM1ZswYhYSEqHXr1po6dary8vLKq+wqqST7onPnztq2bZtz+uzQoUNau3atevfuXS41439K6/e7QtxZurScOnVKeXl5zsdz5AsJCdGePXsKHZOSklJo/5SUlDKr0wpKsi9+76mnnlKDBg0K/ENH8ZVkf2zdulVvvfWWdu7cWQ4VWkdJ9sWhQ4f08ccfa8iQIVq7dq0OHDig0aNHKycnR/Hx8eVRdpVUkn0xePBgnTp1SrfccouMMcrNzdVDDz2kiRMnlkfJ+I1L/X6np6fr/Pnz8vPzK9JyqtQRIVQd06ZN09KlS/X+++/L19fX3eVYztmzZzV06FDNnz9fQUFB7i7H8hwOh4KDg/XGG28oPDxcAwcO1NNPP63ExER3l2Y5W7Zs0dSpU/Xaa69p+/btWrVqldasWaMpU6a4uzSUUJU6IhQUFCQvLy+lpqa6tKempqpevXqFjqlXr16x+qNoSrIv8s2YMUPTpk3Tpk2bdOONN5ZlmZZR3P1x8OBBHTlyRH379nW2ORwOSVK1atW0d+9eNW3atGyLrqJK8rdRv359eXt7y8vLy9l2/fXXKyUlRdnZ2fLx8SnTmquqkuyLZ599VkOHDtUDDzwgSWrTpo0yMjI0cuRIPf300y4PCUfZutTvd0BAQJGPBklV7IiQj4+PwsPDlZSU5GxzOBxKSkpSZGRkoWMiIyNd+kvSxo0bL9kfRVOSfSFJL774oqZMmaL169erQ4cO5VGqJRR3f7Rs2VLffvutdu7c6Xzdeeed6tatm3bu3Cm73V6e5VcpJfnb6NKliw4cOOAMo5K0b98+1a9fnxB0FUqyLzIzMwuEnfyAanh0Z7kqtd/v4p3HXfEtXbrU2Gw2s2jRIvPf//7XjBw50tSqVcukpKQYY4wZOnSoGT9+vLP/559/bqpVq2ZmzJhhdu/ebeLj47l8vpQUd19MmzbN+Pj4mJUrV5qff/7Z+Tp79qy7NqFKKe7++D2uGis9xd0XR48eNf7+/mbs2LFm79695qOPPjLBwcHmb3/7m7s2ocoo7r6Ij483/v7+5t133zWHDh0y//73v03Tpk3NgAED3LUJVcbZs2fNjh07zI4dO4wkM2vWLLNjxw7zww8/GGOMGT9+vBk6dKizf/7l80888YTZvXu3mTdvHpfP53v11VfNtddea3x8fEynTp3Ml19+6fwsKirKxMbGuvRfvny5adGihfHx8TGtWrUya9asKeeKq67i7ItGjRoZSQVe8fHx5V94FVXcv43fIgiVruLuiy+++MJEREQYm81mmjRpYl544QWTm5tbzlVXTcXZFzk5Oea5554zTZs2Nb6+vsZut5vRo0ebX3/9tfwLr2I2b95c6G9A/vcfGxtroqKiCoxp166d8fHxMU2aNDELFy4s9no9jOFYHgAAsKYqdY4QAABAcRCEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAACAZRGEAECSh4eHPvjgA0nSkSNH5OHhoZ07d7q1JgBljyAEwO2GDx8uDw8PeXh4yNvbW40bN9aTTz6pCxcuuLs0AFVclXr6PIDKq2fPnlq4cKFycnK0bds2xcbGysPDQ9OnT3d3aQCqMI4IAagQbDab6tWrJ7vdrn79+ik6OlobN26UdPGJ4AkJCWrcuLH8/PzUtm1brVy50mX8999/rzvuuEMBAQHy9/dX165ddfDgQUnS119/re7duysoKEiBgYGKiorS9u3by30bAVQ8BCEAFc53332nL774Qj4+PpKkhIQE/eMf/1BiYqK+//57jRs3Tvfee68++eQTSdJPP/2kW2+9VTabTR9//LG2bdum++67T7m5uZKks2fPKjY2Vlu3btWXX36p5s2bq3fv3jp79qzbthFAxcDUGIAK4aOPPlLNmjWVm5urrKwseXp6au7cucrKytLUqVO1adMmRUZGSpKaNGmirVu36u9//7uioqI0b948BQYGaunSpfL29pYktWjRwrns22+/3WVdb7zxhmrVqqVPPvlEd9xxR/ltJIAKhyAEoELo1q2bXn/9dWVkZOjll19WtWrVdNddd+n7779XZmamunfv7tI/Oztb7du3lyTt3LlTXbt2dYag30tNTdUzzzyjLVu26MSJE8rLy1NmZqaOHj1a5tsFoGIjCAGoEGrUqKFmzZpJkhYsWKC2bdvqrbfeUuvWrSVJa9asUWhoqMsYm80mSfLz87vssmNjY/XLL79ozpw5atSokWw2myIjI5WdnV0GWwKgMiEIAahwPD09NXHiRMXFxWnfvn2y2Ww6evSooqKiCu1/4403avHixcrJySn0qNDnn3+u1157Tb1795YkHTt2TKdOnSrTbQBQOXCyNIAKqX///vLy8tLf//53Pf744xo3bpwWL16sgwcPavv27Xr11Ve1ePFiSdLYsWOVnp6ue+65R998843279+vf/7zn9q7d68kqXnz5vrnP/+p3bt36//+7/80ZMiQKx5FAmANHBECUCFVq1ZNY8eO1YsvvqjDhw+rbt26SkhI0KFDh1SrVi3ddNNNmjhxoiSpTp06+vjjj/XEE08oKipKXl5eateunbp06SJJeuuttzRy5EjddNNNstvtmjp1qh5//HF3bh6ACsLDGGPcXQQAAIA7MDUGAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAsiyAEAAAs6/8BbRS4iFJSK8oAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"decision tree:\")\n",
    "param_grid_DT = {\n",
    "    'criterion': ['gini', 'entropy', \"log_loss\"],\n",
    "    'max_depth': [3, 5, 7, 9, None],\n",
    "    'min_samples_split': [2, 3, 5, 7, 10],\n",
    "    'min_samples_leaf': [1, 2, 3, 4],\n",
    "}\n",
    "\n",
    "best_para_DT, search_time_DT_a = gridSearchHelper(DecisionTreeClassifier(random_state=42),param_grid_DT,x_train_down, y_train_down)\n",
    "\n",
    "DT_model_a = DecisionTreeClassifier(**best_para_DT)\n",
    "DT_model_a, training_time_DT_a = trainHelper(DT_model, x_train_down, y_train_down)\n",
    "f1_DT_a, precision_DT_a, recall_DT_a, inference_time_DT_a  = inferenceHelper(DT_model, x_test, y_test)\n",
    "auprc_DT_a = plot_PRC(DT_model, x_test, y_test)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.3 随机森林"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random Forest Classifier\n",
      "=== grid search ===\n",
      "grid search time:  127.08617734909058 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.2711164951324463 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.7069351230425055\n",
      "precision:  0.558303886925795\n",
      "recall:  0.9634146341463414\n",
      "Inference Time: 0.2921621799468994 sec\n",
      "\n",
      "AUPRC: 0.8467236761427657\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Random Forest Classifier\")\n",
    "param_grid_RF = {\n",
    "    'n_estimators': [50, 100, 200],\n",
    "    'max_depth': [5, 10, 15],\n",
    "    'min_samples_split': [2, 4, 6],\n",
    "    'min_samples_leaf': [1, 2, 4]\n",
    "}\n",
    "best_para_RF, search_time_RF_a = gridSearchHelper(RandomForestClassifier(random_state=42),param_grid_RF,x_train_down, y_train_down)\n",
    "\n",
    "RF_model = RandomForestClassifier(**best_para_RF)\n",
    "RF_model, training_time_RF_a = trainHelper(RF_model, x_train_down, y_train_down)\n",
    "f1_RF_a, precision_RF_a, recall_RF_a, inference_time_RF_a  = inferenceHelper(RF_model, x_test, y_test)\n",
    "auprc_RF_a = plot_PRC(RF_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.4 SVM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Support vector machine\n",
      "=== grid search ===\n",
      "grid search time:  40.798986196517944 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.4196341037750244 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.7069351230425055\n",
      "precision:  0.558303886925795\n",
      "recall:  0.9634146341463414\n",
      "Inference Time: 0.29553818702697754 sec\n",
      "\n",
      "AUPRC: 0.7459856729394014\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Support vector machine\")\n",
    "param_grid_SVM = {\n",
    "    'C': [0.1, 1, 10],\n",
    "    'kernel': ['linear', 'rbf'],\n",
    "    'gamma': [0.1, 1, 10]\n",
    "}\n",
    "best_para_SVM, search_time_SVM_a = gridSearchHelper(SVC(random_state=42),param_grid_SVM,x_train_down, y_train_down)\n",
    "\n",
    "SVM_model_a = SVC(**best_para_SVM, probability=True)\n",
    "SVM_model_a, training_time_SVM_a = trainHelper(SVM_model, x_train_down, y_train_down)\n",
    "f1_SVM_a, precision_SVM_a, recall_SVM_a,inference_time_SVM_a  = inferenceHelper(RF_model, x_test, y_test)\n",
    "auprc_SVM_a = plot_PRC(SVM_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2.5 KNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== grid search ===\n",
      "grid search time:  0.559220552444458 sec\n",
      "\n",
      "=== train ===\n",
      "Training time:  0.0038688182830810547 sec\n",
      "\n",
      "=== inference ===\n",
      "F1 Score:  0.5957152729785764\n",
      "precision:  0.4513089005235602\n",
      "recall:  0.8760162601626016\n",
      "Inference Time: 0.3825702667236328 sec\n",
      "\n",
      "AUPRC: 0.655023001572397\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "param_grid_KNN = {\n",
    "    'n_neighbors': [5, 10, 15], \n",
    "    'weights': ['uniform', 'distance']\n",
    "}\n",
    "best_para_KNN, search_time_KNN_a = gridSearchHelper(KNeighborsClassifier(), param_grid_KNN, x_train_down, y_train_down)\n",
    "\n",
    "KNN_model = KNeighborsClassifier(**best_para_KNN)\n",
    "KNN_model, training_time_KNN_a = trainHelper(KNN_model, x_train_down, y_train_down)\n",
    "f1_KNN_a, precision_KNN_a, recall_KNN_a,inference_time_KNN_a  = inferenceHelper(KNN_model, x_test, y_test)\n",
    "auprc_KNN_a = plot_PRC(KNN_model, x_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 比较加速后训练指标与未加速的训练指标"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "search_time = [search_time_LR, search_time_DT, search_time_RF, search_time_SVM, search_time_KNN]\n",
    "training_time = [training_time_LR, training_time_DT, training_time_RF, training_time_SVM, training_time_KNN]\n",
    "inference_time = [inference_time_LR, inference_time_DT, inference_time_RF, inference_time_SVM, inference_time_KNN]\n",
    "f1 = [f1_LR, f1_DT, f1_RF, f1_SVM, f1_KNN]\n",
    "precision = [precision_LR, precision_DT, precision_RF,precision_SVM, precision_KNN]\n",
    "recall = [recall_LR, recall_DT, recall_RF, recall_SVM, recall_KNN]\n",
    "auprc = [auprc_LR, auprc_DT, auprc_RF, auprc_SVM, auprc_KNN]\n",
    "\n",
    "search_time_a = [search_time_LR_a, search_time_DT_a, search_time_RF_a, search_time_SVM_a, search_time_KNN_a]\n",
    "training_time_a = [training_time_LR_a, training_time_DT_a, training_time_RF_a, training_time_SVM_a, training_time_KNN_a]\n",
    "inference_time_a = [inference_time_RF_a, inference_time_DT_a, inference_time_RF_a, inference_time_RF_a, inference_time_SVM_a, inference_time_KNN_a]\n",
    "f1_a = [f1_LR_a, f1_DT_a, f1_RF_a, f1_SVM_a, f1_KNN_a]\n",
    "precision_a = [precision_LR_a, precision_DT_a, precision_RF_a, precision_SVM_a, precision_KNN_a]\n",
    "recall_a = [recall_LR_a, recall_DT_a, recall_RF_a, recall_SVM_a, recall_KNN_a]\n",
    "auprc_a = [auprc_LR_a, auprc_DT_a, auprc_RF_a, auprc_SVM_a, auprc_KNN_a]\n",
    "\n",
    "unit_1 = [1,1,1,1,1]\n",
    "bar_width = 0.35\n",
    "opacity = 0.8\n",
    "algorithms = ['LR', 'DT', 'RF', 'SVM', 'KNN']\n",
    "x = np.arange(len(algorithms))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "训练时间比较"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "training Time (sec)\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         0.0243          0.0126         \n",
      "DT         0.0959          0.0968         \n",
      "RF         2.0796          0.2711         \n",
      "SVM        0.4233          0.4196         \n",
      "KNN        0.0019          0.0039         \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"training Time (sec)\")\n",
    "printAsTable(training_time, training_time_a)\n",
    "\n",
    "plotCompareTime(\"training time\", training_time, training_time_a)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "推理时间比较\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "inference Time (sec)\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         0.0501          0.2922         \n",
      "DT         0.0551          0.0491         \n",
      "RF         1.4327          0.2922         \n",
      "SVM        1.4401          0.2922         \n",
      "KNN        1.1088          0.2955         \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"inference Time (sec)\")\n",
    "printAsTable(inference_time, inference_time_a)\n",
    "\n",
    "plotCompareTime(\"inference time\", inference_time, inference_time_a)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "grid search时间比较"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "grid search time (sec)\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         7.2731          5.3678         \n",
      "DT         160.0968        162.6754       \n",
      "RF         721.3496        127.0862       \n",
      "SVM        90.7146         40.7990        \n",
      "KNN        1.0625          0.5592         \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"grid search time (sec)\")\n",
    "printAsTable(search_time, search_time_a)\n",
    "\n",
    "plotCompareTime(\"grid search time\", search_time, search_time_a)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "AUPRC比较"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "AUPRC\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         0.7472          0.7472         \n",
      "DT         0.3488          0.3501         \n",
      "RF         0.7979          0.8467         \n",
      "SVM        0.7452          0.7460         \n",
      "KNN        0.8241          0.6550         \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"AUPRC\")\n",
    "printAsTable(auprc, auprc_a)\n",
    "\n",
    "plotCompareScore(\"AUPRC\", auprc, auprc_a)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "f1 比较"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "f1 score\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         0.5847          0.5847         \n",
      "DT         0.4946          0.4957         \n",
      "RF         0.7103          0.7069         \n",
      "SVM        0.7103          0.7069         \n",
      "KNN        0.6944          0.5957         \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"f1 score\")\n",
    "printAsTable(f1, f1_a)\n",
    "\n",
    "plotCompareScore(\"F1 score\", f1, f1_a)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "precision 比较"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "precision\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         0.4429          0.4429         \n",
      "DT         0.3430          0.3441         \n",
      "RF         0.5890          0.5583         \n",
      "SVM        0.5890          0.5583         \n",
      "KNN        0.5319          0.4513         \n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"precision\")\n",
    "printAsTable(precision, precision_a)\n",
    "\n",
    "plotCompareScore(\"precision\",precision, precision_a)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "recall比较"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "recall\n",
      "Model      origin          oneAPI         \n",
      "----------------------------------------\n",
      "LR         0.8598          0.8598         \n",
      "DT         0.8862          0.8862         \n",
      "RF         0.8943          0.9634         \n",
      "SVM        0.8943          0.9634         \n",
      "KNN        1.0000          0.8760         \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAArMAAAK9CAYAAAA37eRrAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/SrBM8AAAACXBIWXMAAA9hAAAPYQGoP6dpAABWSElEQVR4nO3de3zP9f//8ft7582cDxsam0NOOWXxcSiqsRxT+SRncyj1cah9cyrM4ZOhMCEkbH1Q6KScKsuUKDFKhZLJIbP5hDEattfvj37en97tPTa2vT25XS+X9+Xi/Xw/X6/X4/V+Wt0993y9XjbLsiwBAAAABnJzdQEAAADA9SLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCuCGxsbGy2Ww6dOiQva1Vq1Zq1aqVy2oyyfjx42Wz2VxdRqG7kfN29ncOwO2LMAvApTIzM7VkyRK1atVKpUqVkre3t4KDgxUREaEdO3a4ujzcwho1aqRnnnnG1WU4aNy4sWw2m+bNm+f08ytB/srLx8dHd955pwYPHqwTJ07Y+yUkJMhms+mdd94prNIBlyHMAnCZCxcuqEOHDurXr58sy9ILL7ygefPmqXfv3tq2bZsaN26so0ePurrMAjVmzBhduHDB1WUUuhs57169eunChQuqXLnydR//+PHj2rVrl9q3b3/d+8hvP//8s7755hsFBwdr2bJlV+07ceJE/ec//9GcOXPUrFkzzZs3T02bNtX58+cLqVrg5uHh6gIA3Lj09HQVKVLE1WXk2fDhw7VhwwbNnDlTzz77rMNnUVFRmjlzpmsKKwRXxszDw0MeHrfPf4rz47zd3d3l7u5+Q3WsX79ePj4+euCBB25oP/lp6dKlKleunKZPn64uXbro0KFDCg4Odtq3bdu2Cg0NlSQNGDBApUuX1owZM7R69Wp169atEKsGXI+ZWcAwV9Ya/vjjj+revbtKliypFi1a2D9funSpGjVqJF9fX5UqVUpPPPGEjhw5km0/X3/9tdq1a6eSJUuqSJEiqlevnmbNmmX//LvvvlPfvn1VpUoV+fj4KDAwUP369dN///vffDmPo0ePasGCBWrdunW2ICv9GVief/553XHHHfa2Xbt2qW3btipWrJj8/f314IMP6quvvnLY7sqvYbds2aKhQ4eqbNmyKlGihJ566ildvHhRp0+fVu/evVWyZEmVLFlSI0aMkGVZ9u0PHTokm82mV155RTNnzlTlypXl6+urli1b6vvvv3c4Vm6/o6uNmbO1o59++qlatGihEiVKyN/fXzVq1NALL7zg0CclJUX9+/dXQECAfHx8VL9+fcXFxTn0+eu5vP7666pataq8vb11zz336JtvvnHoe+nSJe3bt0/Hjx93NlzZfPbZZ7r33ntVpEgRlShRQg8//LD27t17Q+d94cIFDR06VGXKlFHRokXVqVMnHTt2TDabTePHj7f3c7ZmNjg4WB06dNCWLVvUuHFj+fj4qEqVKnrzzTed1r927Vrdf//98vX1tbetWrXK/rNTpkwZ9ezZU8eOHXPYrm/fvvL399exY8fUuXNn+fv7q2zZsnr++eeVmZnp0DcrK0sxMTGqU6eOfHx8FBAQoKeeekqnTp1yWtPy5cvVpUsXdejQQcWLF9fy5cudf/lOXAnlSUlJud4GuFXcPtMBwC3mn//8p6pXr67Jkyfbw9hLL72ksWPH6vHHH9eAAQOUmpqq2bNn67777tOuXbtUokQJSX+GpQ4dOqh8+fIaNmyYAgMDtXfvXq1Zs0bDhg2z9zl48KAiIiIUGBioH374Qa+//rp++OEHffXVVzd80dL69et1+fJl9erVK1f9f/jhB917770qVqyYRowYIU9PTy1YsECtWrXS5s2b1aRJE4f+Q4YMUWBgoCZMmKCvvvpKr7/+ukqUKKGtW7eqUqVKmjx5statW6eXX35Zd911l3r37u2w/ZtvvqmzZ8/qX//6l/744w/NmjVLDzzwgPbs2aOAgIDr+o6cjZmz8+zQoYPq1auniRMnytvbWwcOHNCXX35p73PhwgW1atVKBw4c0ODBgxUSEqJVq1apb9++On36tH0Mr1i+fLnOnj2rp556SjabTdOmTdOjjz6qgwcPytPTU5J07Ngx1apVS3369FFsbOxVx2Ljxo1q27atqlSpovHjx+vChQuaPXu2mjdvrsTExGyzibk5b+nPoLhy5Ur16tVL//jHP7R58+Y8LQM4cOCAunTpov79+6tPnz5avHix+vbtq0aNGqlOnTr2fpcuXdLGjRs1efJke1tsbKwiIiJ0zz33KDo6WidOnNCsWbP05ZdfOvzsSH+u8w4PD1eTJk30yiuvaOPGjZo+fbqqVq2qp59+2t7vqaeesu936NChSkpK0pw5c7Rr1y59+eWX9u9e+vMflwcOHNCSJUvk5eWlRx99VMuWLcv2j5ic/PLLL5Kk0qVL5/r7Am4ZFgCjREVFWZKsbt26ObQfOnTIcnd3t1566SWH9j179lgeHh729suXL1shISFW5cqVrVOnTjn0zcrKsv/5/Pnz2Y791ltvWZKszz//3N62ZMkSS5KVlJRkb2vZsqXVsmXLq57Hc889Z0mydu3addV+V3Tu3Nny8vKyfvnlF3vbb7/9ZhUtWtS67777stUTHh7ucD5Nmza1bDabNWjQIHvb5cuXrTvuuMOh1qSkJEuS5evrax09etTe/vXXX1uSrOeee87eltvvKKcx++tnV8ycOdOSZKWmpub4XcTExFiSrKVLl9rbLl68aDVt2tTy9/e30tLSHM6ldOnS1u+//27vu3r1akuS9dFHH2U77z59+uR43CsaNGhglStXzvrvf/9rb/v2228tNzc3q3fv3td13jt37rQkWc8++6xDv759+1qSrKioKHubs79zlStXzva9p6SkWN7e3tb//d//OewzPj7eYfuLFy9a5cqVs+666y7rwoUL9n5r1qyxJFnjxo2zt/Xp08eSZE2cONFhnw0bNrQaNWpkf//FF19Ykqxly5Y59NuwYYPT9sGDB1tBQUH2v7OffPKJ05+PK+e+ceNGKzU11Tpy5Ij19ttvW6VLl3b4O7tp0yZLkrVq1SoLuNWxzAAw1KBBgxzev/fee8rKytLjjz+ukydP2l+BgYGqXr26Nm3aJOnPX9UnJSXp2WefdZhtkuQwk/jXX7/+8ccfOnnypP7xj39IkhITE2+4/rS0NElS0aJFr9k3MzNTn3zyiTp37qwqVarY28uXL6/u3btry5Yt9v1d0b9/f4fzadKkiSzLUv/+/e1t7u7uCg0N1cGDB7Mds3PnzqpYsaL9fePGjdWkSROtW7fO3pbX7+jvY+bMlTFZvXq1srKynPZZt26dAgMDHdZGenp6aujQoTp37pw2b97s0L9r164qWbKk/f29994rSQ7nHRwcLMuyrjkre/z4ce3evVt9+/ZVqVKl7O316tVT69atHb6fK3Jz3hs2bJCkbHcXGDJkyDW3vaJ27dr2c5OksmXLqkaNGtnGd926dapdu7Z9BnnHjh1KSUnRM888Ix8fH3u/9u3bq2bNmlq7du01z+nee+91OM6qVatUvHhxtW7d2uHnsVGjRvL397f/PErS5cuXtWLFCnXt2tX+d/aBBx5QuXLlcrwQLCwsTGXLllVQUJCeeOIJ+fv76/3333f4OwvcLgizgKFCQkIc3v/888+yLEvVq1dX2bJlHV579+5VSkqKpP/9OvKuu+666v5///13DRs2TAEBAfL19VXZsmXtxzxz5swN11+sWDFJ0tmzZ6/ZNzU1VefPn1eNGjWyfVarVi1lZWVlWxdcqVIlh/fFixeXJAUFBWVrd7aGsXr16tna7rzzTod1mnn9jv4+Zs507dpVzZs314ABAxQQEKAnnnhCK1eudAi2v/76q6pXry43N8f/hNeqVcv++V/9/bu4EmxzWrt5NVf2ndNYnDx5Uunp6Q7tuTnvX3/9VW5ubtn6VqtWLde1/f08pT/P9e/nuXbtWoflC1c7p5o1a2b7Pn18fFS2bNmrHufnn3/WmTNnVK5cuWw/j+fOnbP/PErSJ598otTUVDVu3FgHDhzQgQMHlJSUpPvvv19vvfWW03/UzJ07V59++qk2bdqkH3/8UQcPHlR4ePjVvh7glsWaWcBQf50VlP682MRms2n9+vVOr/T29/fP0/4ff/xxbd26VcOHD1eDBg3k7++vrKwsPfTQQznOGOZFzZo1JUl79uxRgwYNbnh/f5fT1e7O2q2rrOO8mrx+R38fM2d8fX31+eefa9OmTVq7dq02bNigFStW6IEHHtAnn3xyXVfx57TN9Z53XuXmvPNDbs4zKSlJ+/bty/E+rjdynL/Kysq66szqX8PwlT6PP/64076bN2/W/fff79DWuHFj+90MgNsdYRa4RVStWlWWZSkkJER33nnnVftJ0vfff6+wsDCnfU6dOqX4+HhNmDBB48aNs7f//PPP+VZv27Zt5e7urqVLl17zIrCyZcvKz89P+/fvz/bZvn375Obmlm3G9UY5O9effvrJ/qvpgvyO3Nzc9OCDD+rBBx/UjBkzNHnyZL344ovatGmTwsLCVLlyZX333XfKyspymJ3dt2+fJN3Q/Vev5cq+cxqLMmXKXNdt4ipXrqysrCwlJSU5zIofOHDg+ot1Yu3atSpevLjDHUD+ek5/v1XX/v37r+v7rFq1qjZu3KjmzZtfNcynp6dr9erV6tq1q7p06ZLt86FDh2rZsmXZwiyA/2GZAXCLePTRR+Xu7q4JEyZkm3GzLMt+u6i7775bISEhiomJ0enTp7P1k/438/T3/cTExORbvUFBQRo4cKA++eQTzZ49O9vnWVlZmj59uo4ePSp3d3e1adNGq1evdvg1/4kTJ7R8+XK1aNHCvmwhv3zwwQcOt2Xavn27vv76a7Vt21ZSwX1Hv//+e7a2KzPXGRkZkqR27dopOTlZK1assPe5fPmyZs+eLX9/f7Vs2TLPx83trbnKly+vBg0aKC4uzuHvz/fff69PPvlE7dq1y/OxJdl/Rf7aa685tDv7u3Ej1q1bpzZt2jjc4zY0NFTlypXT/Pnz7d+x9OcdN/bu3XtdD1Z4/PHHlZmZqUmTJmX77PLly/bv7v3331d6err+9a9/qUuXLtleHTp00LvvvutQFwBHzMwCt4iqVavq3//+t0aPHq1Dhw6pc+fOKlq0qJKSkvT+++/rySef1PPPPy83NzfNmzdPHTt2VIMGDRQREaHy5ctr3759+uGHH/Txxx+rWLFiuu+++zRt2jRdunRJFStW1CeffJLv97CcPn26fvnlFw0dOlTvvfeeOnTooJIlS+rw4cNatWqV9u3bpyeeeEKS9O9//9t+/9VnnnlGHh4eWrBggTIyMjRt2rR8rUv6c61mixYt9PTTTysjI0MxMTEqXbq0RowYIUkF9h1NnDhRn3/+udq3b6/KlSsrJSVFr732mu644w77bOKTTz6pBQsWqG/fvtq5c6eCg4P1zjvv6Msvv1RMTEyuLqr7u7zcmuvll19W27Zt1bRpU/Xv399+a67ixYs73A82Lxo1aqTHHntMMTEx+u9//2u/NddPP/0kSTd8Kzjpz1uabdq0SfPnz3do9/T01NSpUxUREaGWLVuqW7du9ltzBQcH67nnnsvzsVq2bKmnnnpK0dHR2r17t9q0aSNPT0/9/PPPWrVqlWbNmqUuXbpo2bJlKl26tJo1a+Z0P506ddLChQu1du1aPfroo9d13sCtjjAL3EJGjRqlO++8UzNnztSECRMk/TkD2qZNG3Xq1MneLzw8XJs2bdKECRM0ffp0ZWVlqWrVqho4cKC9z/LlyzVkyBDNnTtXlmWpTZs2Wr9+vSpUqJBv9fr5+Wn9+vWKjY1VXFycJk2apPPnz6tChQp64IEHtGzZMvvV2XXq1NEXX3yh0aNHKzo6WllZWWrSpImWLl2a7R6z+aF3795yc3NTTEyMUlJS1LhxY82ZM0fly5e39ymI76hTp046dOiQFi9erJMnT6pMmTJq2bKlJkyYYL+IzdfXVwkJCRo1apTi4uKUlpamGjVqaMmSJerbt++Nnvo1hYWFacOGDYqKitK4cePk6empli1baurUqbm62Csnb775pgIDA/XWW2/p/fffV1hYmFasWKEaNWo43GXgen322WfKyMiwz67/Vd++feXn56cpU6Zo5MiRKlKkiB555BFNnTo1210/cmv+/Plq1KiRFixYoBdeeEEeHh4KDg5Wz5491bx5c6WkpGjjxo3q1q1bjutwH3zwQfn5+Wnp0qWEWSAHNquwrgAAAAMcOnRIISEhevnll/X888+7upzb3u7du9WwYUMtXbpUPXr0uKF9PfPMM9qxY4e2b9+eT9UBuBkwMwsAuClcuHAh28VSMTExcnNz03333XfD+2/QoIE6dux4w/sBcHMhzAIAbgrTpk3Tzp07df/998vDw0Pr16/X+vXr9eSTT+bL3SqefPLJfKgSwM2GMAsAuCk0a9ZMn376qSZNmqRz586pUqVKGj9+vF588UVXlwbgJsaaWQAAABiL+8wCAADAWIRZAAAAGOu2WzOblZWl3377TUWLFs2Xm3ADAAAgf1mWpbNnz6pChQoOj+125rYLs7/99lu+P8MdAAAA+e/IkSO64447rtrntguzVx7zeOTIkXx/ljsAAABuXFpamoKCgnL1eO7bLsxeWVpQrFgxwiwAAMBNLDdLQrkADAAAAMYizAIAAMBYhFkAAAAY67ZbMwsAAHBFZmamLl265Ooybkuenp5yd3e/4f0QZgEAwG3p3LlzOnr0qCzLcnUptyWbzaY77rhD/v7+N7QfwiwAALjtZGZm6ujRo/Lz81PZsmV5kFIhsyxLqampOnr0qKpXr35DM7SEWQAAcNu5dOmSLMtS2bJl5evr6+pybktly5bVoUOHdOnSpRsKs1wABgAAblvMyLpOfn33hFkAAAAYizALALe4zz//XB07dlSFChVks9n0wQcfXHObhIQE3X333fL29la1atUUGxvr8HlmZqbGjh2rkJAQ+fr6qmrVqpo0aRIX0rgA44vbHWtmAeAWl56ervr166tfv3569NFHr9k/KSlJ7du316BBg7Rs2TLFx8drwIABKl++vMLDwyVJU6dO1bx58xQXF6c6depox44dioiIUPHixTV06NCCPiX8BeObv8JXbCzU433cNaxQjjN+/Hh98MEH2r17d663adWqlRo0aKCYmJgCqys/2Kzb7J9ZaWlpKl68uM6cOaNixYq5uhwAKFQ2m03vv/++OnfunGOfkSNHau3atfr+++/tbU888YROnz6tDRs2SJI6dOiggIAALVq0yN7nsccek6+vr5YuXVpg9ePqGN/c++OPP5SUlKSQkBD5+PjY22/VMHvu3DllZGSodOnSud7m999/l6enp4oWLVogNeU0BlLe8hrLDAAADrZt26awMMf/wYaHh2vbtm32982aNVN8fLx++uknSdK3336rLVu2qG3btoVaK/KO8b29WJaly5cvy9/fP09BVpJKlSpVYEE2PxFmAQAOkpOTFRAQ4NAWEBCgtLQ0XbhwQZI0atQoPfHEE6pZs6Y8PT3VsGFDPfvss+rRo4crSkYeML7my8jI0NChQ1WuXDn5+PioRYsW+uabbyT9uR7aZrNp/fr1atSokby9vbVlyxaNHz9eDRo0sO/j8uXLGjp0qEqUKKHSpUtr5MiR6tOnj8OsfqtWrfTss8/a3wcHB2vy5Mnq16+fihYtqkqVKun1118vpLPOGWEWAJBnK1eu1LJly7R8+XIlJiYqLi5Or7zyiuLi4lxdGvIB43tzGzFihN59913FxcUpMTFR1apVU3h4uH7//Xd7n1GjRmnKlCnau3ev6tWrl20fU6dO1bJly7RkyRJ9+eWXSktLy9XFg9OnT1doaKh27dqlZ555Rk8//bT279+fn6eXZ1wABgBwEBgYqBMnTji0nThxQsWKFbPfXH748OH22TtJqlu3rn799VdFR0erT58+hV4zco/xNVt6errmzZun2NhY+7KPhQsX6tNPP9WiRYt0zz33SJImTpyo1q1b57if2bNna/To0XrkkUckSXPmzNG6deuuefx27drpmWeekfTn+uuZM2dq06ZNqlGjxo2e2nVjZhYA4KBp06aKj493aPv000/VtGlT+/vz58/Lzc3xfyHu7u7KysoqlBpx/Rhfs/3yyy+6dOmSmjdvbm/z9PRU48aNtXfvXntbaGhojvs4c+aMTpw4ocaNG9vb3N3d1ahRo2se/6+zvDabTYGBgUpJScnraeQrZmYB4BZ37tw5HThwwP4+KSlJu3fvVqlSpVSpUiWNHj1ax44d05tvvilJGjRokObMmaMRI0aoX79++uyzz7Ry5UqtXbvWvo+OHTvqpZdeUqVKlVSnTh3t2rVLM2bMUL9+/Qr9/G53jC+cKVKkSIHs19PT0+G9zWZz+T9ymJkFgFvcjh071LBhQzVs2FCSFBkZqYYNG2rcuHGSpOPHj+vw4cP2/iEhIVq7dq0+/fRT1a9fX9OnT9cbb7xhvwep9OevKLt06aJnnnlGtWrV0vPPP6+nnnpKkyZNKtyTA+N7m6lataq8vLz05Zdf2tsuXbqkb775RrVr187VPooXL66AgAD7RWPSnw/KSExMzPd6CwMzswBwi2vVqtVVn9z096c/Xdlm165dOW5TtGhRxcTE3PQ3U78dML63lyJFiujpp5/W8OHD7bPv06ZN0/nz59W/f399++23udrPkCFDFB0drWrVqqlmzZqaPXu2Tp06JZvNVsBnkP8IswAAAP9fYT3E4EZMmTJFWVlZ6tWrl86ePavQ0FB9/PHHKlmyZK73MXLkSCUnJ6t3795yd3fXk08+qfDwcLm7uxdg5QWDJ4ABAIDbztWePnU7ysrKUq1atfT4448X2nKS/HoCGDOzAAAAt5lff/1Vn3zyiVq2bKmMjAzNmTNHSUlJ6t69u6tLyzMuAAMAALjNuLm5KTY2Vvfcc4+aN2+uPXv2aOPGjapVq5arS8szZmYBAABuM0FBQQ53RDAZM7MAAAAwFjOzAHALeHh8sqtLyJXV4wNdXYKRGF8gZ8zMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLC4AAwAA+P8K+2I7Lpq7cczMAgAAwKno6Gi5u7vr5ZdfzvZZbGysbDabbDab3NzcdMcddygiIkIpKSn2PjabTR988EGB1kiYBQAAgFOLFy/WiBEjtHjxYqefFytWTMePH9fRo0e1cOFCrV+/Xr169SrUGgmzAAAABsnIyNDQoUNVrlw5+fj4qEWLFvrmm28kSQkJCbLZbIqPj1doaKj8/PzUrFkz7d+/32Efq1ev1t133y0fHx9VqVJFEyZM0OXLlx36bN68WRcuXNDEiROVlpamrVu3ZqvFZrMpMDBQFSpUUNu2bTV06FBt3LhRFy5cKLgv4G8IswAAAAYZMWKE3n33XcXFxSkxMVHVqlVTeHi4fv/9d3ufF198UdOnT9eOHTvk4eGhfv362T/74osv1Lt3bw0bNkw//vijFixYoNjYWL300ksOx1m0aJG6desmT09PdevWTYsWLbpmbb6+vsrKysoWjAsSYRYAAMAQ6enpmjdvnl5++WW1bdtWtWvX1sKFC+Xr6+sQNl966SW1bNlStWvX1qhRo7R161b98ccfkqQJEyZo1KhR6tOnj6pUqaLWrVtr0qRJWrBggX37tLQ0vfPOO+rZs6ckqWfPnlq5cqXOnTuXY20///yz5s+fr9DQUBUtWrSAvoHsCLMAAACG+OWXX3Tp0iU1b97c3ubp6anGjRtr79699rZ69erZ/1y+fHlJsl+Y9e2332rixIny9/e3vwYOHKjjx4/r/PnzkqS33npLVatWVf369SVJDRo0UOXKlbVixQqHes6cOSN/f3/5+fmpRo0aCggI0LJlywrm5HPArbkAAABuMZ6envY/22w2SVJWVpYk6dy5c5owYYIeffTRbNv5+PhI+nOJwQ8//CAPj/9FxaysLC1evFj9+/e3txUtWlSJiYlyc3NT+fLl5evrWyDnczWEWQAAAENUrVpVXl5e+vLLL1W5cmVJ0qVLl/TNN9/o2WefzdU+7r77bu3fv1/VqlVz+vmePXu0Y8cOJSQkqFSpUvb233//Xa1atdK+fftUs2ZNSZKbm1uO+ykshFkAAABDFClSRE8//bSGDx+uUqVKqVKlSpo2bZrOnz+v/v3769tvv73mPsaNG6cOHTqoUqVK6tKli9zc3PTtt9/q+++/17///W8tWrRIjRs31n333Zdt23vuuUeLFi1yet9ZVyHMAgAA/H8mPJFrypQpysrKUq9evXT27FmFhobq448/VsmSJXO1fXh4uNasWaOJEydq6tSp8vT0VM2aNTVgwABdvHhRS5cu1ciRI51u+9hjj2n69OmaPHlyfp7SDbFZlmW5uojClJaWpuLFi+vMmTMqVqyYq8sBgHxR2I/gvF4mBIWbEeOb//744w8lJSUpJCTEvk4UhetqY5CXvMbdDAAAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAA3LZus+vgbyr59d0TZgEAwG3H3d1dknTx4kUXV3L7uvLdXxmL68V9ZgEAwG3Hw8NDfn5+Sk1Nlaenp9zcmN8rTFlZWUpNTZWfn5/DI3OvB2EWAADcdmw2m8qXL6+kpCT9+uuvri7ntuTm5qZKlSrJZrPd0H4IswAA4Lbk5eWl6tWrs9TARby8vPJlRpw5deTa559/ro4dO6pChQqy2Wz64IMPrrlNQkKC7r77bnl7e6tatWqKjY3N1mfu3LkKDg6Wj4+PmjRpou3bt+d/8bgmxhfA7cjNzU0+Pj68XPDKr6UdhFnkWnp6uurXr6+5c+fmqn9SUpLat2+v+++/X7t379azzz6rAQMG6OOPP7b3WbFihSIjIxUVFaXExETVr19f4eHhSklJKajTQA4YXwCAiWzWbXZPirw86xc5s9lsev/999W5c+cc+4wcOVJr167V999/b2974okndPr0aW3YsEGS1KRJE91zzz2aM2eOpD8XhAcFBWnIkCEaNWpUgZ4Dcsb4mufh8cmuLiFXVo8PdHUJRmJ8cbvJS15jZhYFZtu2bQoLC3NoCw8P17Zt2yT9eUuOnTt3OvRxc3NTWFiYvQ9uXowvAOBmQJhFgUlOTlZAQIBDW0BAgNLS0nThwgWdPHlSmZmZTvskJ5sxC3E7Y3wBADcDwiwAAACMxa25UGACAwN14sQJh7YTJ06oWLFi8vX1lbu7u9zd3Z32CQxk3dXNjvEFANwMmJlFgWnatKni4+Md2j799FM1bdpU0p/3l2vUqJFDn6ysLMXHx9v74ObF+AIAbgaEWeTauXPntHv3bu3evVvSn7dm2r17tw4fPixJGj16tHr37m3vP2jQIB08eFAjRozQvn379Nprr2nlypV67rnn7H0iIyO1cOFCxcXFae/evXr66aeVnp6uiIiIQj03ML4AADOxzAC5tmPHDt1///3295GRkZKkPn36KDY2VsePH7cHH0kKCQnR2rVr9dxzz2nWrFm644479MYbbyg8PNzep2vXrkpNTdW4ceOUnJysBg0aaMOGDdkuGkLBY3wBACbiPrMAcAvgPqS3NsYXtxvuMwsAAIDbgkvDbEE9Cx4AAAC3B5eG2YJ4FjwAAABuHy69AKxt27Zq27ZtrvvPnz9fISEhmj59uiSpVq1a2rJli2bOnOlw0QkAAABuD0atmb3Ws+CdycjIUFpamsMLAAAAtwajbs11rWfB+/r6ZtsmOjpaEyZMKKwSnQpfsdGlx88tn713ubqEXLnZrpZlfPMX43t9fGTG+N5sGN9bmynj+3HXsGt3Qo6Mmpm9HqNHj9aZM2fsryNHjri6JAAAAOQTo2Zmr/UseGe8vb3l7e1dGOUBAACgkBk1M3utZ8EDAADg9uLSMFsQz4IHAADA7cOlYXbHjh1q2LChGjZsKOnPZ8E3bNhQ48aNk6QcnwX/6aefqn79+po+fXq2Z8EDAADg9uHSNbOtWrWSZVk5fu7s6V6tWrXSrl27CrAqAAAAmMKoNbMAAADAXxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAN7m5c+cqODhYPj4+atKkibZv337V/jExMapRo4Z8fX0VFBSk5557Tn/88Yf988zMTI0dO1YhISHy9fVV1apVNWnSJFmWVdCnku88XF0AAAAAcrZixQpFRkZq/vz5atKkiWJiYhQeHq79+/erXLly2fovX75co0aN0uLFi9WsWTP99NNP6tu3r2w2m2bMmCFJmjp1qubNm6e4uDjVqVNHO3bsUEREhIoXL66hQ4cW9ineEGZmAQAAbmIzZszQwIEDFRERodq1a2v+/Pny8/PT4sWLnfbfunWrmjdvru7duys4OFht2rRRt27dHGZzt27dqocffljt27dXcHCwunTpojZt2lxzxvdmRJgFAAC4SV28eFE7d+5UWFiYvc3NzU1hYWHatm2b022aNWumnTt32oPpwYMHtW7dOrVr186hT3x8vH766SdJ0rfffqstW7aobdu2BXg2BYNlBgAAADepkydPKjMzUwEBAQ7tAQEB2rdvn9NtunfvrpMnT6pFixayLEuXL1/WoEGD9MILL9j7jBo1SmlpaapZs6bc3d2VmZmpl156ST169CjQ8ykIzMwCAADcQhISEjR58mS99tprSkxM1Hvvvae1a9dq0qRJ9j4rV67UsmXLtHz5ciUmJiouLk6vvPKK4uLiXFj59WFmFgAA4CZVpkwZubu768SJEw7tJ06cUGBgoNNtxo4dq169emnAgAGSpLp16yo9PV1PPvmkXnzxRbm5uWn48OEaNWqUnnjiCXufX3/9VdHR0erTp0/BnlQ+Y2YWAADgJuXl5aVGjRopPj7e3paVlaX4+Hg1bdrU6Tbnz5+Xm5tjxHN3d5ck+623cuqTlZWVn+UXCmZmAQAAbmKRkZHq06ePQkND1bhxY8XExCg9PV0RERGSpN69e6tixYqKjo6WJHXs2FEzZsxQw4YN1aRJEx04cEBjx45Vx44d7aG2Y8eOeumll1SpUiXVqVNHu3bt0owZM9SvXz+Xnef1IswCAADcxLp27arU1FSNGzdOycnJatCggTZs2GC/KOzw4cMOs6xjxoyRzWbTmDFjdOzYMZUtW9YeXq+YPXu2xo4dq2eeeUYpKSmqUKGCnnrqKY0bN67Qz+9GEWYBAABucoMHD9bgwYOdfpaQkODw3sPDQ1FRUYqKispxf0WLFlVMTIxiYmLysUrXYM0sAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGItbcwEAALjQw+OTXV1Crqwe7/zxua7GzCwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxnJ5mJ07d66Cg4Pl4+OjJk2aaPv27VftHxMToxo1asjX11dBQUF67rnn9McffxRStQAAALiZuDTMrlixQpGRkYqKilJiYqLq16+v8PBwpaSkOO2/fPlyjRo1SlFRUdq7d68WLVqkFStW6IUXXijkygEAAHAzcGmYnTFjhgYOHKiIiAjVrl1b8+fPl5+fnxYvXuy0/9atW9W8eXN1795dwcHBatOmjbp163bN2VwAAADcmlwWZi9evKidO3cqLCzsf8W4uSksLEzbtm1zuk2zZs20c+dOe3g9ePCg1q1bp3bt2uV4nIyMDKWlpTm8AAAAcGvwcNWBT548qczMTAUEBDi0BwQEaN++fU636d69u06ePKkWLVrIsixdvnxZgwYNuuoyg+joaE2YMCFfawcAAMDNweUXgOVFQkKCJk+erNdee02JiYl67733tHbtWk2aNCnHbUaPHq0zZ87YX0eOHCnEigEAAFCQXDYzW6ZMGbm7u+vEiRMO7SdOnFBgYKDTbcaOHatevXppwIABkqS6desqPT1dTz75pF588UW5uWXP5t7e3vL29s7/EwAAAIDLuWxm1svLS40aNVJ8fLy9LSsrS/Hx8WratKnTbc6fP58tsLq7u0uSLMsquGIBAABwU3LZzKwkRUZGqk+fPgoNDVXjxo0VExOj9PR0RURESJJ69+6tihUrKjo6WpLUsWNHzZgxQw0bNlSTJk104MABjR07Vh07drSHWgAAANw+XBpmu3btqtTUVI0bN07Jyclq0KCBNmzYYL8o7PDhww4zsWPGjJHNZtOYMWN07NgxlS1bVh07dtRLL73kqlMAAACAC7k0zErS4MGDNXjwYKefJSQkOLz38PBQVFSUoqKiCqEyAAAA3OyMupsBAAAA8FeEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxvK4no0yMzP1wQcfaO/evZKkOnXqqFOnTnJ3d8/X4gAAAICryXOYPXDggNq3b6+jR4+qRo0akqTo6GgFBQVp7dq1qlq1ar4XCQAAADiT52UGQ4cOVZUqVXTkyBElJiYqMTFRhw8fVkhIiIYOHVoQNQIAAABO5XlmdvPmzfrqq69UqlQpe1vp0qU1ZcoUNW/ePF+LAwAAAK4mzzOz3t7eOnv2bLb2c+fOycvLK1+KAgAAAHIjz2G2Q4cOevLJJ/X111/LsixZlqWvvvpKgwYNUqdOnQqiRgAAAMCpPIfZV199VVWrVlXTpk3l4+MjHx8fNW/eXNWqVdOsWbMKokYAAADAqTyvmS1RooRWr16tn3/+Wfv27ZMk1apVS9WqVcv34gAAAICrua77zEpS9erVVb169fysBQAAAMiTXIXZyMhITZo0SUWKFFFkZORV+86YMSNfCgMAAACuJVdhdteuXbp06ZL9zwAAAMDNIFdhdtOmTU7/DAAAALhSnu9m0K9fP6f3mU1PT1e/fv3ypSgAAAAgN/IcZuPi4nThwoVs7RcuXNCbb76ZL0UBAAAAuZHruxmkpaXZH5Jw9uxZ+fj42D/LzMzUunXrVK5cuQIpEgAAAHAm12G2RIkSstlsstlsuvPOO7N9brPZNGHChHwtDgAAALiaXIfZTZs2ybIsPfDAA3r33XdVqlQp+2deXl6qXLmyKlSoUCBFAgAAAM7kOsy2bNlSkpSUlKSgoCC5ueV5uS0AAACQr/L8BLDKlStLks6fP6/Dhw/r4sWLDp/Xq1cvfyoDAAAAriHPYTY1NVURERFav369088zMzNvuCgAAAAgN/K8VuDZZ5/V6dOn9fXXX8vX11cbNmxQXFycqlevrg8//LAgagQAAACcyvPM7GeffabVq1crNDRUbm5uqly5slq3bq1ixYopOjpa7du3L4g6AQAAgGzyPDObnp5uv59syZIllZqaKkmqW7euEhMT87c6AAAA4CryHGZr1Kih/fv3S5Lq16+vBQsW6NixY5o/f77Kly+f7wUCAAAAOcnzMoNhw4bp+PHjkqSoqCg99NBDWrZsmby8vBQbG5vf9QEAAAA5ynOY7dmzp/3PjRo10q+//qp9+/apUqVKKlOmTL4WBwAAAFzNDT/5wM/PT3fffbf8/f31yiuv5Hn7uXPnKjg4WD4+PmrSpIm2b99+1f6nT5/Wv/71L5UvX17e3t668847tW7duustHwAAAAbLU5hNTU3VmjVr9Mknn9jvJ3vp0iXNmjVLwcHBmjJlSp4OvmLFCkVGRioqKkqJiYmqX7++wsPDlZKS4rT/xYsX1bp1ax06dEjvvPOO9u/fr4ULF6pixYp5Oi4AAABuDbleZrBlyxZ16NBBaWlpstlsCg0N1ZIlS9S5c2d5eHho/Pjx6tOnT54OPmPGDA0cOFARERGSpPnz52vt2rVavHixRo0ala3/4sWL9fvvv2vr1q3y9PSUJAUHB+fpmAAAALh15HpmdsyYMWrXrp2+++47RUZG6ptvvtEjjzyiyZMn68cff9SgQYPk6+ub6wNfvHhRO3fuVFhY2P+KcXNTWFiYtm3b5nSbDz/8UE2bNtW//vUvBQQE6K677tLkyZOv+tSxjIwMpaWlObwAAABwa8h1mN2zZ4/GjBmju+66SxMnTpTNZtO0adPUpUuX6zrwyZMnlZmZqYCAAIf2gIAAJScnO93m4MGDeuedd5SZmal169Zp7Nixmj59uv7973/neJzo6GgVL17c/goKCrquegEAAHDzyXWYPXXqlP1uBb6+vvLz89Ndd91VYIU5k5WVpXLlyun1119Xo0aN1LVrV7344ouaP39+jtuMHj1aZ86csb+OHDlSiBUDAACgIOXp1lw//vijfdbUsizt379f6enpDn3q1auXq32VKVNG7u7uOnHihEP7iRMnFBgY6HSb8uXLy9PTU+7u7va2WrVqKTk5WRcvXpSXl1e2bby9veXt7Z2rmgAAAGCWPIXZBx98UJZl2d936NBBkmSz2WRZlmw221XXr/6Vl5eXGjVqpPj4eHXu3FnSnzOv8fHxGjx4sNNtmjdvruXLlysrK0tubn9OKv/0008qX7680yALAACAW1uuw2xSUlK+HzwyMlJ9+vRRaGioGjdurJiYGKWnp9vvbtC7d29VrFhR0dHRkqSnn35ac+bM0bBhwzRkyBD9/PPPmjx5soYOHZrvtQEAAODml+swW7ly5Xw/eNeuXZWamqpx48YpOTlZDRo00IYNG+wXhR0+fNg+AytJQUFB+vjjj/Xcc8+pXr16qlixooYNG6aRI0fme20AAAC4+eX5cbb5bfDgwTkuK0hISMjW1rRpU3311VcFXBUAAABMcMOPswUAAABchTALAAAAYxFmAQAAYKzrXjObkpKi/fv3S5Jq1KihcuXK5VtRAAAAQG7keWb27Nmz6tWrlypWrKiWLVuqZcuWqlixonr27KkzZ84URI0AAACAU3kOswMGDNDXX3+tNWvW6PTp0zp9+rTWrFmjHTt26KmnniqIGgEAAACn8rzMYM2aNfr444/VokULe1t4eLgWLlyohx56KF+LAwAAAK4mzzOzpUuXVvHixbO1Fy9eXCVLlsyXogAAAIDcyHOYHTNmjCIjI5WcnGxvS05O1vDhwzV27Nh8LQ4AAAC4mjwvM5g3b54OHDigSpUqqVKlSpL+fOyst7e3UlNTtWDBAnvfxMTE/KsUAAAA+Js8h9nOnTsXQBkAAABA3uU5zEZFRRVEHQAAAECe8QQwAAAAGCvPM7Nubm6y2Ww5fp6ZmXlDBQEAAAC5lecw+/777zu8v3Tpknbt2qW4uDhNmDAh3woDAAAAriXPYfbhhx/O1talSxfVqVNHK1asUP/+/fOlMAAAAOBa8m3N7D/+8Q/Fx8fn1+4AAACAa8qXMHvhwgW9+uqrqlixYn7sDgAAAMiVPC8zKFmypMMFYJZl6ezZs/Lz89PSpUvztTgAAADgavIcZmfOnOkQZt3c3FS2bFk1adJEJUuWzNfiAAAAgKvJc5jt27dvAZQBAAAA5F2uwux3332X6x3Wq1fvuosBAAAA8iJXYbZBgway2WyyLEuSeGgCAAAAbgq5uptBUlKSDh48qKSkJL333nsKCQnRa6+9pl27dmnXrl167bXXVLVqVb377rsFXS8AAABgl6uZ2cqVK9v//M9//lOvvvqq2rVrZ2+rV6+egoKCNHbsWHXu3DnfiwQAAACcyfN9Zvfs2aOQkJBs7SEhIfrxxx/zpSgAAAAgN/IcZmvVqqXo6GhdvHjR3nbx4kVFR0erVq1a+VocAAAAcDV5vjXX/Pnz1bFjR91xxx32Oxd89913stls+uijj/K9QAAAACAneQ6zjRs31sGDB7Vs2TLt27dPktS1a1d1795dRYoUyfcCAQAAgJzkOcxKUpEiRfTkk0/mdy0AAABAnuR5zawk/ec//1GLFi1UoUIF/frrr5L+fMzt6tWr87U4AAAA4GryHGbnzZunyMhItW3bVqdOnbI/JKFkyZKKiYnJ7/oAAACAHOU5zM6ePVsLFy7Uiy++KA+P/61SCA0N1Z49e/K1OAAAAOBq8hxmk5KS1LBhw2zt3t7eSk9Pz5eiAAAAgNzIc5gNCQnR7t27s7Vv2LCB+8wCAACgUOX5bgaRkZH617/+pT/++EOWZWn79u166623FB0drTfeeKMgagQAAACcynOYHTBggHx9fTVmzBidP39e3bt3V4UKFTRr1iw98cQTBVEjAAAA4NR13We2R48e6tGjh86fP69z586pXLly+V0XAAAAcE3XdZ/Zy5cva+PGjfrPf/4jX19fSdJvv/2mc+fO5WtxAAAAwNXkeWb2119/1UMPPaTDhw8rIyNDrVu3VtGiRTV16lRlZGRo/vz5BVEnAAAAkE2eZ2aHDRum0NBQnTp1yj4rK0mPPPKI4uPj87U4AAAA4GryPDP7xRdfaOvWrfLy8nJoDw4O1rFjx/KtMAAAAOBa8jwzm5WVZX+E7V8dPXpURYsWzZeiAAAAgNzIc5ht06aNYmJi7O9tNpvOnTunqKgotWvXLj9rAwAAAK4qz8sMpk+frvDwcNWuXVt//PGHunfvrp9//lllypTRW2+9VRA1AgAAAE7lOczecccd+vbbb/X222/ru+++07lz59S/f3/16NHD4YIwAAAAoKBd10MTPDw81LNnz/yuBQAAAMiT6wqz+/fv1+zZs7V3715JUq1atTR48GDVrFkzX4sDAAAAribPF4C9++67uuuuu7Rz507Vr19f9evXV2JiourWrat33323IGoEAAAAnMrzzOyIESM0evRoTZw40aE9KipKI0aM0GOPPZZvxQEAAABXk+eZ2ePHj6t3797Z2nv27Knjx4/nS1EAAABAbuQ5zLZq1UpffPFFtvYtW7bo3nvvzZeiAAAAgNzI8zKDTp06aeTIkdq5c6f+8Y9/SJK++uorrVq1ShMmTNCHH37o0BcAAAAoKHkOs88884wk6bXXXtNrr73m9DPpzyeDOXvsLQAAAJBf8hxms7KyCqIOAAAAIM/yvGYWAAAAuFnkOsxu27ZNa9ascWh78803FRISonLlyunJJ59URkZGvhcIAAAA5CTXYXbixIn64Ycf7O/37Nmj/v37KywsTKNGjdJHH32k6OjoAikSAAAAcCbXYXb37t168MEH7e/ffvttNWnSRAsXLlRkZKReffVVrVy5skCKBAAAAJzJdZg9deqUAgIC7O83b96stm3b2t/fc889OnLkSP5WBwAAAFxFrsNsQECAkpKSJEkXL15UYmKi/T6zknT27Fl5enrmf4UAAABADnIdZtu1a6dRo0bpiy++0OjRo+Xn5+fwxK/vvvtOVatWLZAiAQAAAGdyfZ/ZSZMm6dFHH1XLli3l7++vuLg4eXl52T9fvHix2rRpUyBFAgAAAM7kOsyWKVNGn3/+uc6cOSN/f3+5u7s7fL5q1Sr5+/vne4EAAABATvL8BLDixYs7bS9VqtQNFwMAAADkBU8AAwAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsW6KMDt37lwFBwfLx8dHTZo00fbt23O13dtvvy2bzabOnTsXbIEAAAC4Kbk8zK5YsUKRkZGKiopSYmKi6tevr/DwcKWkpFx1u0OHDun555/XvffeW0iVAgAA4Gbj8jA7Y8YMDRw4UBEREapdu7bmz58vPz8/LV68OMdtMjMz1aNHD02YMEFVqlQpxGoBAABwM3FpmL148aJ27typsLAwe5ubm5vCwsK0bdu2HLebOHGiypUrp/79+1/zGBkZGUpLS3N4AQAA4Nbg0jB78uRJZWZmKiAgwKE9ICBAycnJTrfZsmWLFi1apIULF+bqGNHR0SpevLj9FRQUdMN1AwAA4Obg8mUGeXH27Fn16tVLCxcuVJkyZXK1zejRo3XmzBn768iRIwVcJQAAAAqLhysPXqZMGbm7u+vEiRMO7SdOnFBgYGC2/r/88osOHTqkjh072tuysrIkSR4eHtq/f7+qVq3qsI23t7e8vb0LoHoAAAC4mktnZr28vNSoUSPFx8fb27KyshQfH6+mTZtm61+zZk3t2bNHu3fvtr86deqk+++/X7t372YJAQAAwG3GpTOzkhQZGak+ffooNDRUjRs3VkxMjNLT0xURESFJ6t27typWrKjo6Gj5+Pjorrvucti+RIkSkpStHQAAALc+l4fZrl27KjU1VePGjVNycrIaNGigDRs22C8KO3z4sNzcjFraCwAAgELi8jArSYMHD9bgwYOdfpaQkHDVbWNjY/O/IAAAABiBKU8AAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAY90UYXbu3LkKDg6Wj4+PmjRpou3bt+fYd+HChbr33ntVsmRJlSxZUmFhYVftDwAAgFuXy8PsihUrFBkZqaioKCUmJqp+/foKDw9XSkqK0/4JCQnq1q2bNm3apG3btikoKEht2rTRsWPHCrlyAAAAuJrLw+yMGTM0cOBARUREqHbt2po/f778/Py0ePFip/2XLVumZ555Rg0aNFDNmjX1xhtvKCsrS/Hx8YVcOQAAAFzNpWH24sWL2rlzp8LCwuxtbm5uCgsL07Zt23K1j/Pnz+vSpUsqVaqU088zMjKUlpbm8AIAAMCtwaVh9uTJk8rMzFRAQIBDe0BAgJKTk3O1j5EjR6pChQoOgfivoqOjVbx4cfsrKCjohusGAADAzcHlywxuxJQpU/T222/r/fffl4+Pj9M+o0eP1pkzZ+yvI0eOFHKVAAAAKCgerjx4mTJl5O7urhMnTji0nzhxQoGBgVfd9pVXXtGUKVO0ceNG1atXL8d+3t7e8vb2zpd6AQAAcHNx6cysl5eXGjVq5HDx1pWLuZo2bZrjdtOmTdOkSZO0YcMGhYaGFkapAAAAuAm5dGZWkiIjI9WnTx+FhoaqcePGiomJUXp6uiIiIiRJvXv3VsWKFRUdHS1Jmjp1qsaNG6fly5crODjYvrbW399f/v7+LjsPAAAAFD6Xh9muXbsqNTVV48aNU3Jysho0aKANGzbYLwo7fPiw3Nz+N4E8b948Xbx4UV26dHHYT1RUlMaPH1+YpQMAAMDFXB5mJWnw4MEaPHiw088SEhIc3h86dKjgCwIAAIARjL6bAQAAAG5vhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxiLMAgAAwFiEWQAAABiLMAsAAABjEWYBAABgLMIsAAAAjEWYBQAAgLEIswAAADAWYRYAAADGIswCAADAWIRZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAAAAY66YIs3PnzlVwcLB8fHzUpEkTbd++/ar9V61apZo1a8rHx0d169bVunXrCqlSAAAA3ExcHmZXrFihyMhIRUVFKTExUfXr11d4eLhSUlKc9t+6dau6deum/v37a9euXercubM6d+6s77//vpArBwAAgKu5PMzOmDFDAwcOVEREhGrXrq358+fLz89Pixcvdtp/1qxZeuihhzR8+HDVqlVLkyZN0t133605c+YUcuUAAABwNQ9XHvzixYvauXOnRo8ebW9zc3NTWFiYtm3b5nSbbdu2KTIy0qEtPDxcH3zwgdP+GRkZysjIsL8/c+aMJCktLe0Gq8+9y+fTC+1YN+JSxllXl5AraWl+ri7BAeObvxjf68P4Xh/GN38xvteH8XV2rD9zmmVZ1+zr0jB78uRJZWZmKiAgwKE9ICBA+/btc7pNcnKy0/7JyclO+0dHR2vChAnZ2oOCgq6zarha8SmurgAFifG9tTG+tzbG99bmivE9e/asihcvftU+Lg2zhWH06NEOM7lZWVn6/fffVbp0adlsNhdWdnNJS0tTUFCQjhw5omLFirm6HOQzxvfWxvje2hjfWxvj65xlWTp79qwqVKhwzb4uDbNlypSRu7u7Tpw44dB+4sQJBQYGOt0mMDAwT/29vb3l7e3t0FaiRInrL/oWV6xYMX6YbmGM762N8b21Mb63NsY3u2vNyF7h0gvAvLy81KhRI8XHx9vbsrKyFB8fr6ZNmzrdpmnTpg79JenTTz/NsT8AAABuXS5fZhAZGak+ffooNDRUjRs3VkxMjNLT0xURESFJ6t27typWrKjo6GhJ0rBhw9SyZUtNnz5d7du319tvv60dO3bo9ddfd+VpAAAAwAVcHma7du2q1NRUjRs3TsnJyWrQoIE2bNhgv8jr8OHDcnP73wRys2bNtHz5co0ZM0YvvPCCqlevrg8++EB33XWXq07hluDt7a2oqKhsSzJwa2B8b22M762N8b21Mb43zmbl5p4HAAAAwE3I5Q9NAAAAAK4XYRYAAADGIswCAADAWIRZAAAAGIswe5vp27evOnfu7PSz4OBg2Ww22Ww2+fn5qW7dunrjjTcKt0Bct759+9rHz9PTUwEBAWrdurUWL16srKwsJSQk2D/P6ZWQkODq08BV/H2MQ0JCNGLECP3xxx/2Ps7GtUWLFi6sGn+Xmpqqp59+WpUqVZK3t7cCAwMVHh6uzZs3q0yZMpoyxfkzQydNmqSAgABdunRJsbGxstlsqlWrVrZ+q1atks1mU3BwcAGfCZxx9v/Zd955Rz4+Ppo+fbr95/jv4/zBBx84PJn0yn+z69Spo8zMTIe+JUqUUGxsbEGdgnEIs3AwceJEHT9+XN9//7169uypgQMHav369a4uC7n00EMP6fjx4zp06JDWr1+v+++/X8OGDVOHDh3UrFkzHT9+3P56/PHH7f2vvJo1a+bqU8A1XBmzgwcPaubMmVqwYIGioqIc+ixZssRhXD/88EMXVQtnHnvsMe3atUtxcXH66aef9OGHH6pVq1Y6c+aMevbsqSVLlmTbxrIsxcbGqnfv3vL09JQkFSlSRCkpKdq2bZtD30WLFqlSpUqFci64tjfeeEM9evTQvHnz9H//93+SJB8fH02dOlWnTp265vYHDx7Um2++WdBlGo0wCwdFixZVYGCgqlSpopEjR6pUqVL69NNPXV0WcunKLE/FihV1991364UXXtDq1au1fv16vfnmmwoMDLS/fH197f2vvLy8vFx9CriGK2MWFBSkzp07KywsLNvPaIkSJRzGtVSpUi6qFn93+vRpffHFF5o6daruv/9+Va5cWY0bN9bo0aPVqVMn9e/fXz/99JO2bNnisN3mzZt18OBB9e/f397m4eGh7t27a/Hixfa2o0ePKiEhQd27dy+0c0LOpk2bpiFDhujtt9+2PwxKksLCwhQYGGh/INTVDBkyRFFRUcrIyCjIUo1GmIVTWVlZevfdd3Xq1CkCjuEeeOAB1a9fX++9956rS0E++/7777V161Z+Rg3i7+8vf39/ffDBB07DSd26dXXPPfc4BFTpz9n2Zs2aqWbNmg7t/fr108qVK3X+/HlJUmxsrB566CH7g4fgOiNHjtSkSZO0Zs0aPfLIIw6fubu7a/LkyZo9e7aOHj161f08++yzunz5smbPnl2Q5RqNMAsHI0eOlL+/v7y9vdWlSxeVLFlSAwYMcHVZuEE1a9bUoUOHXF0G8sGaNWvk7+8vHx8f1a1bVykpKRo+fLhDn27dutlD05XghJuDh4eHYmNjFRcXpxIlSqh58+Z64YUX9N1339n79O/fX6tWrdK5c+ckSWfPntU777yjfv36Zdtfw4YNVaVKFb3zzjv2pQjO+qFwrV+/XtOmTdPq1av14IMPOu3zyCOPqEGDBtmWCf2dn5+foqKiFB0drTNnzhREucYjzMLB8OHDtXv3bn322Wdq0qSJZs6cqWrVqrm6LNwgy7IcLiyAue6//37t3r1bX3/9tfr06aOIiAg99thjDn1mzpyp3bt321+tW7d2UbVw5rHHHtNvv/2mDz/8UA899JASEhJ099132y/o6datmzIzM7Vy5UpJ0ooVK+Tm5qauXbs63V+/fv20ZMkSbd68Wenp6WrXrl1hnQpyUK9ePQUHBysqKsr+jxJnpk6dqri4OO3du/eq++vfv79Kly6tqVOn5neptwTCLByUKVNG1apV07333qtVq1Zp6NCh+vHHH11dFm7Q3r17FRIS4uoykA+KFCmiatWqqX79+lq8eLG+/vprLVq0yKFPYGCgqlWrZn8VKVLERdUiJz4+PmrdurXGjh2rrVu3qm/fvvYZumLFiqlLly72C8GWLFmixx9/XP7+/k731aNHD3311VcaP368evXqJQ8Pj0I7DzhXsWJFJSQk6NixY3rooYd09uxZp/3uu+8+hYeHa/To0Vfdn4eHh1566SXNmjVLv/32W0GUbDTCLHIUFBSkrl27XvOHDDe3zz77THv27Mk2ewfzubm56YUXXtCYMWN04cIFV5eDG1C7dm2lp6fb3/fv319btmzRmjVrtHXrVocLv/6uVKlS6tSpkzZv3swSg5tI5cqVtXnzZiUnJ1810E6ZMkUfffRRtrtS/N0///lP1alTRxMmTCiIco1GmL0NnTlzxuFXkLt379aRI0ec9h02bJg++ugj7dixo5CrxPXIyMhQcnKyjh07psTERE2ePFkPP/ywOnTooN69e7u6PBSAf/7zn3J3d9fcuXNdXQpy4b///a8eeOABLV26VN99952SkpK0atUqTZs2TQ8//LC933333adq1aqpd+/eqlmz5jVvmxcbG6uTJ09mu0AMrhUUFKSEhASlpKQoPDxcaWlp2frUrVtXPXr00KuvvnrN/U2ZMkWLFy92+IcPCLO3pYSEBDVs2NDhldO/9GrXrq02bdpo3LhxhVwlrseGDRtUvnx5BQcH66GHHtKmTZv06quvavXq1XJ3d3d1eSgAHh4eGjx4sKZNm8b/4Azg7+9vvx7hvvvu01133aWxY8dq4MCBmjNnjr2fzWZTv379dOrUqVzNtvr6+qp06dIFWTqu0x133KGEhASdPHkyx0A7ceJEZWVlXXNfDzzwgB544AFdvny5IEo1ls2yLMvVRQAAAADXg5lZAAAAGIswCwAAAGMRZgEAAGAswiwAAACMRZgFAACAsQizAAAAMBZhFgAAAMYizAIAAMBYhFkAKGAJCQmy2Ww6ffr0TXOs4OBgxcTEFHg9AFDQCLMAkE+2bdsmd3d3tW/f3mU1NGvWTMePH1fx4sUlSbGxsSpRooTL6gGAgkaYBYB8smjRIg0ZMkSff/65fvvtt0I//qVLl+Tl5aXAwEDZbLZCPz4AuAJhFgDywblz57RixQo9/fTTat++vWJjY6/af+HChQoKCpKfn58eeeQRzZgxI9sM6rx581S1alV5eXmpRo0a+s9//uPwuc1m07x589SpUycVKVJEL730ksMyg4SEBEVEROjMmTOy2Wyy2WwaP368ffvz58+rX79+Klq0qCpVqqTXX3/d/tmhQ4dks9m0cuVK3XvvvfL19dU999yjn376Sd98841CQ0Pl7++vtm3bKjU19Ua/PgC4fhYA4IYtWrTICg0NtSzLsj766COratWqVlZWlmVZlrVp0yZLknXq1CnLsixry5Ytlpubm/Xyyy9b+/fvt+bOnWuVKlXKKl68uH1/7733nuXp6WnNnTvX2r9/vzV9+nTL3d3d+uyzz+x9JFnlypWzFi9ebP3yyy/Wr7/+6nCsjIwMKyYmxipWrJh1/Phx6/jx49bZs2cty7KsypUrW6VKlbLmzp1r/fzzz1Z0dLTl5uZm7du3z7Isy0pKSrIkWTVr1rQ2bNhg/fjjj9Y//vEPq1GjRlarVq2sLVu2WImJiVa1atWsQYMGFcI3DADOEWYBIB80a9bMiomJsSzLsi5dumSVKVPG2rRpk2VZ2cNs165drfbt2zts36NHD4cw26xZM2vgwIEOff75z39a7dq1s7+XZD377LMOff5+rCVLljjs94rKlStbPXv2tL/PysqyypUrZ82bN8+yrP+F2TfeeMPe56233rIkWfHx8fa26Ohoq0aNGlf5ZgCgYLHMAABu0P79+7V9+3Z169ZNkuTh4aGuXbtq0aJFOfZv3LixQ9vf3+/du1fNmzd3aGvevLn27t3r0BYaGnrddderV8/+Z5vNpsDAQKWkpOTYJyAgQJJUt25dh7a/bwMAhcnD1QUAgOkWLVqky5cvq0KFCvY2y7Lk7e2tOXPmFOixixQpct3benp6Ory32WzKysrKsc+Vi8r+3vb3bQCgMDEzCwA34PLly3rzzTc1ffp07d692/769ttvVaFCBb311lvZtqlRo4a++eYbh7a/v69Vq5a+/PJLh7Yvv/xStWvXzlN9Xl5eyszMzNM2AGASZmYB4AasWbNGp06dUv/+/e33dr3iscce06JFi/Tyyy87tA8ZMkT33XefZsyYoY4dO+qzzz7T+vXrHW6nNXz4cD3++ONq2LChwsLC9NFHH+m9997Txo0b81RfcHCwzp07p/j4eNWvX19+fn7y8/O7/hMGgJsMM7MAcAMWLVqksLCwbEFW+jPM7tixQ999951De/PmzTV//nzNmDFD9evX14YNG/Tcc8/Jx8fH3qdz586aNWuWXnnlFdWpU0cLFizQkiVL1KpVqzzV16xZMw0aNEhdu3ZV2bJlNW3atOs6TwC4Wdksy7JcXQQA3O4GDhyoffv26YsvvnB1KQBgFJYZAIALvPLKK2rdurWKFCmi9evXKy4uTq+99pqrywIA4zAzCwAu8PjjjyshIUFnz55VlSpVNGTIEA0aNMjVZQGAcQizAAAAMBYXgAEAAMBYhFkAAAAYizALAAAAYxFmAQAAYCzCLAAAAIxFmAUAAICxCLMAAAAwFmEWAAAAxvp/gitc3kVn9bsAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"recall\")\n",
    "printAsTable(recall, recall_a)\n",
    "\n",
    "plotCompareScore(\"recall\", recall, recall_a)\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "modin-hdk",
   "language": "python",
   "name": "modin-hdk"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}