homework/Day_030_HW.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# 作業 : (Kaggle)鐵達尼生存預測"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [作業目標]\n",
    "- 試著模仿範例寫法, 在鐵達尼生存預測中, 使用葉編碼並觀察預測效果"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# [作業重點]\n",
    "- 仿造範例, 完成葉編碼的寫作 : 使用隨機森林 (In[3], Out[3], In[4], Out[4])\n",
    "- 仿造範例, 觀察葉編碼搭配邏輯斯迴歸後的效果 (In[5], Out[5], In[6], Out[6]) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T01:56:38.372835Z",
     "start_time": "2019-05-22T01:56:36.874462Z"
    }
   },
   "outputs": [],
   "source": [
    "# 做完特徵工程前的所有準備\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import copy\n",
    "from sklearn.preprocessing import LabelEncoder, MinMaxScaler\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "# 因為擬合(fit)與編碼(transform)需要分開, 因此不使用.get_dummy, 而採用 sklearn 的 OneHotEncoder\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import roc_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T01:56:41.653820Z",
     "start_time": "2019-05-22T01:56:41.596400Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Pclass                                               Name     Sex   Age  \\\n",
       "0       3                            Braund, Mr. Owen Harris    male  22.0   \n",
       "1       1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.0   \n",
       "2       3                             Heikkinen, Miss. Laina  female  26.0   \n",
       "3       1       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.0   \n",
       "4       3                           Allen, Mr. William Henry    male  35.0   \n",
       "\n",
       "   SibSp  Parch            Ticket     Fare Cabin Embarked  \n",
       "0      1      0         A/5 21171   7.2500   NaN        S  \n",
       "1      1      0          PC 17599  71.2833   C85        C  \n",
       "2      0      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3      1      0            113803  53.1000  C123        S  \n",
       "4      0      0            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_path = 'data/'\n",
    "df = pd.read_csv(data_path + 'titanic_train.csv')\n",
    "\n",
    "train_Y = df['Survived']\n",
    "df = df.drop(['PassengerId', 'Survived'] , axis=1)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T01:56:52.920177Z",
     "start_time": "2019-05-22T01:56:52.870667Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n",
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/utils/validation.py:595: DataConversionWarning: Data with input dtype int64 was converted to float64 by MinMaxScaler.\n",
      "  warnings.warn(msg, DataConversionWarning)\n"
     ]
    }
   ],
   "source": [
    "# 因為需要把類別型與數值型特徵都加入, 故使用最簡版的特徵工程\n",
    "LEncoder = LabelEncoder()\n",
    "MMEncoder = MinMaxScaler()\n",
    "\n",
    "for c in df.columns:\n",
    "    df[c] = df[c].fillna(-1)\n",
    "    if df[c].dtype == 'object':\n",
    "        df[c] = LEncoder.fit_transform(list(df[c].values))\n",
    "    df[c] = MMEncoder.fit_transform(df[c].values.reshape(-1, 1))\n",
    "\n",
    "    df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T01:57:01.895098Z",
     "start_time": "2019-05-22T01:57:01.881413Z"
    }
   },
   "outputs": [],
   "source": [
    "train_X = df.values\n",
    "# 因為訓練邏輯斯迴歸時也要資料, 因此將訓練及切成三部分 train / val / test, 採用 test 驗證而非 k-fold 交叉驗證\n",
    "# train 用來訓練梯度提升樹, val 用來訓練邏輯斯迴歸, test 驗證效果\n",
    "train_X, test_X, train_Y, test_Y = train_test_split(train_X, train_Y, test_size = 0.5)\n",
    "train_X, val_X, train_Y, val_Y = train_test_split(train_X, train_Y, test_size = 0.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# 作業1\n",
    "* 請對照範例，完成隨機森林的鐵達尼生存率預測，以及對應的葉編碼+邏輯斯迴歸"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T02:00:20.557234Z",
     "start_time": "2019-05-22T02:00:20.466987Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/jasonch/anaconda3/lib/python3.6/site-packages/sklearn/preprocessing/_encoders.py:371: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.\n",
      "If you want the future behaviour and silence this warning, you can specify \"categories='auto'\".\n",
      "In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.\n",
      "  warnings.warn(msg, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=1000, multi_class='warn',\n",
       "          n_jobs=None, penalty='l2', random_state=None, solver='lbfgs',\n",
       "          tol=0.0001, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 隨機森林擬合後, 再將葉編碼 (*.apply) 結果做獨熱 / 邏輯斯迴歸\n",
    "rf = RandomForestClassifier(n_estimators = 20, min_samples_split = 10, min_samples_leaf = 5, max_features = 4, max_depth = 3, bootstrap = True)\n",
    "onehot = OneHotEncoder()\n",
    "lr = LogisticRegression(solver = 'lbfgs', max_iter = 1000)\n",
    "\n",
    "\"\"\"\n",
    "Your Code Here (Hint : 隨機森林的葉編碼(.apply)不需要加上[:, :, 0], 直接用rf.apply()調用即可, 本作業其餘寫法相同)\n",
    "\"\"\"\n",
    "rf.fit(train_X, train_Y)\n",
    "onehot.fit(rf.apply(train_X))\n",
    "lr.fit(onehot.transform(rf.apply(val_X)), val_Y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T02:05:36.900533Z",
     "start_time": "2019-05-22T02:05:36.881683Z"
    }
   },
   "outputs": [],
   "source": [
    "# 將隨機森林+葉編碼+邏輯斯迴歸結果輸出\n",
    "\"\"\"\n",
    "Your Code Here \n",
    "\"\"\"\n",
    "pred_rf_lr = lr.predict_proba(onehot.transform(rf.apply(test_X)))[:, 1]\n",
    "fpr_rf_lr, tpr_rf_lr, _ = roc_curve(test_Y, pred_rf_lr)\n",
    "\n",
    "# 將隨機森林結果輸出\n",
    "\"\"\"\n",
    "Your Code Here \n",
    "\"\"\"\n",
    "pred_rf = rf.predict_proba(test_X)[:, 1]\n",
    "fpr_rf, tpr_rf, _ = roc_curve(test_Y, pred_rf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 作業2\n",
    "* 上述的結果，葉編碼是否有提高預測的正確性呢?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-05-22T02:05:38.685217Z",
     "start_time": "2019-05-22T02:05:38.500184Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "# 將結果繪圖\n",
    "\"\"\"\n",
    "Your Code Here \n",
    "\"\"\"\n",
    "plt.plot([0, 1], [0, 1], 'k--')\n",
    "plt.plot(fpr_rf, tpr_rf, label = 'RF')\n",
    "plt.plot(fpr_rf_lr, tpr_rf_lr, label = 'RF + LR')\n",
    "\n",
    "\n",
    "plt.xlabel('False positive rate')\n",
    "plt.ylabel('True positive rate')\n",
    "plt.title('ROC curve')\n",
    "plt.legend(loc='best')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}