bagging.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- voting 1. hard: 다수결 2. soft: 각각의 확률의 평균\n",
    "- bagging soft인데 각각의 알고리즘에 bootstrapping 독립 추출 한 후 aggregation\n",
    "- boosting week learner"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## sklearn.ensemble.VotingClassifier\n",
    "* class sklearn.ensemble.VotingClassifier(estimators, *, voting='hard', weights=None, n_jobs=None, flatten_transform=True, verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import VotingClassifier\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.datasets import load_breast_cancer\n",
    "import pandas as pd\n",
    "cancer = load_breast_cancer()\n",
    "\n",
    "data_df = pd.DataFrame(cancer.data, columns = cancer.feature_names)\n",
    "data_df.head(3)\n",
    "import sys\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9473684210526315\n",
      "LogisticRegression 0.9385964912280702\n",
      "KNeighborsClassifier 0.9385964912280702\n"
     ]
    }
   ],
   "source": [
    "lr_clf = LogisticRegression()\n",
    "knn_clf = KNeighborsClassifier(n_neighbors=8)\n",
    "\n",
    "vo_clf = VotingClassifier(estimators=[('LR',lr_clf), ('KNN',knn_clf)], voting = 'soft')\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(cancer.data, \\\n",
    "    cancer.target, test_size=0.2, random_state=156)\n",
    "\n",
    "vo_clf.fit(X_train, y_train)\n",
    "pred=vo_clf.predict(X_test)\n",
    "print(accuracy_score(y_test, pred))\n",
    "\n",
    "classifiers = [lr_clf, knn_clf]\n",
    "for classifier in classifiers:\n",
    "    classifier.fit(X_train, y_train)\n",
    "    pred = classifier.predict(X_test)\n",
    "    class_name = classifier.__class__.__name__\n",
    "    print(class_name, accuracy_score(y_test, pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## sklearn.ensemble.RandomForestClassifier\n",
    "* class sklearn.ensemble.RandomForestClassifier(n_estimators=100, *, criterion='gini', max_depth=None, min_samples_split=2, min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_features='sqrt', max_leaf_nodes=None, min_impurity_decrease=0.0, bootstrap=True, oob_score=False, n_jobs=None, random_state=None, verbose=0, warm_start=False, class_weight=None, ccp_alpha=0.0, max_samples=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "feature_name_df = pd.read_csv(\"./datasets/UCI HAR Dataset/features.txt\", \\\n",
    "    sep='\\s+', header=None, names=['column_index', 'column_name'])\n",
    "\n",
    "feature_name = feature_name_df.iloc[:,1].values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_new_feature_name_df(old_feature_name_df):\n",
    "    feature_dup_df = pd.DataFrame(data=old_feature_name_df.groupby('column_name').cumcount(),\n",
    "                                  columns=['dup_cnt'])\n",
    "    feature_dup_df = feature_dup_df.reset_index()\n",
    "    new_feature_name_df = pd.merge(old_feature_name_df.reset_index(), feature_dup_df, how='outer')\n",
    "    new_feature_name_df['column_name'] = new_feature_name_df[['column_name', 'dup_cnt']].apply(lambda x : x[0]+'-'+str(x[1])\n",
    "                                                                                                            if x[1] > 0 else x[0], axis=1)\n",
    "    \n",
    "    new_feature_name_df = new_feature_name_df.drop(['index'], axis=1)\n",
    "    return new_feature_name_df\n",
    "\n",
    "def get_human_dataset():\n",
    "    \n",
    "    # 각 데이터 파일들은 공백으로 분리되어 있으므로 read_csv에서 공백문자를 sep으로 할당\n",
    "    feature_name_df = pd.read_csv('./datasets/UCI HAR Dataset/features.txt', sep='\\s+',\n",
    "                                                     header=None, names=['column_index', 'column_name'])\n",
    "    # 중복된 피처명을 수정하는 get_new_feature_name_df()를 이용, 신규 피처명 DataFrame 생성\n",
    "    new_feature_name_df = get_new_feature_name_df(feature_name_df)\n",
    "\n",
    "    # DataFrame에 피처명을 컬럼으로 부여하기 위해 리스트 객체로 다시 반환\n",
    "    feature_name = new_feature_name_df.iloc[:, 1].values.tolist()\n",
    "    \n",
    "    # 학습 피처 데이터세트와 테스트 피처 데이터를 데이터프레임으로 로딩\n",
    "    # 컬럼명은 feature_name 적용\n",
    "    X_train = pd.read_csv('./datasets/UCI HAR Dataset/train/X_train.txt', sep='\\s+', names=feature_name)\n",
    "    X_test = pd.read_csv('./datasets/UCI HAR Dataset/test/X_test.txt', sep='\\s+', names=feature_name)\n",
    "    \n",
    "    # 학습 레이블과 테스트 레이블 데이터를 데이터 프레임으로 로딩, 컬럼명은 action으로 부여\n",
    "    y_train = pd.read_csv('./datasets/UCI HAR Dataset/train/y_train.txt', sep='\\s+', header=None, names=['action'])\n",
    "    y_test = pd.read_csv('./datasets/UCI HAR Dataset/test/y_test.txt', sep='\\s+', header=None, names=['action'])\n",
    "    \n",
    "    # 로드된 학습/테스트용 데이터프레임을 모두 반환\n",
    "    return X_train, X_test, y_train, y_test\n",
    "\n",
    "X_train, X_test, y_train, y_test = get_human_dataset()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9236511706820495"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "rf_clf = RandomForestClassifier(random_state=156)\n",
    "\n",
    "rf_clf.fit(X_train, y_train)\n",
    "\n",
    "pred = rf_clf.predict(X_test)\n",
    "accuracy = accuracy_score(y_test, pred)\n",
    "accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier\n",
    "dt_clf = DecisionTreeClassifier(random_state=156)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 12 candidates, totalling 60 fits\n",
      "DecisionTreeClassifier(max_depth=8, min_samples_split=16, random_state=156)\n",
      "{'max_depth': 8, 'min_samples_split': 16}\n",
      "0.8548794147162603\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "parameters = {'max_depth': [6,8,10,12], 'min_samples_split' : [8,16,20]}\n",
    "grid_cv = GridSearchCV(dt_clf, param_grid=parameters,\\\n",
    "     cv = 5, scoring='accuracy', verbose=1)\n",
    "grid_cv.fit(X_train, y_train)\n",
    "\n",
    "print(grid_cv.best_estimator_)\n",
    "print(grid_cv.best_params_)\n",
    "print(grid_cv.best_score_) #0.8548794147162603"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 5 folds for each of 12 candidates, totalling 60 fits\n",
      "RandomForestClassifier(max_depth=12, min_samples_split=8, random_state=156)\n",
      "{'max_depth': 12, 'min_samples_split': 8}\n",
      "0.9179851736751804\n"
     ]
    }
   ],
   "source": [
    "rf_clf = RandomForestClassifier(random_state=156)\n",
    "parameters = {'max_depth': [6,8,10,12], 'min_samples_split' : [8,16,20]}\n",
    "grid_cv = GridSearchCV(rf_clf, param_grid=parameters,\\\n",
    "     cv = 5, scoring='accuracy', verbose=1)\n",
    "grid_cv.fit(X_train, y_train)\n",
    "\n",
    "print(grid_cv.best_estimator_)\n",
    "print(grid_cv.best_params_)\n",
    "print(grid_cv.best_score_) #0.9179851736751804"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9212758737699356\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "best_df_clf = grid_cv.best_estimator_\n",
    "pred1 = best_df_clf.predict(X_test)\n",
    "print(accuracy_score(y_test,pred1))\n",
    "ftr_importances_values = best_df_clf.feature_importances_\n",
    "ftr_importances = pd.Series(ftr_importances_values, index=X_train.columns)\n",
    "\n",
    "ftr_top10 = ftr_importances.sort_values(ascending=False)[:10]\n",
    "plt.figure(figsize=(8,6))\n",
    "plt.title('Feature importances Top 10')\n",
    "sns.barplot(x=ftr_top10, y= ftr_top10.index)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.7.1 64-bit",
   "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": "3.7.1"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "31f8360559af503f7d1219287c80e0cc362e8a430fe9b8335f384415946613ec"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}