machine-learning/report_01_accident_claims/data_tutorial.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Exercise - 交通事故理赔审核预测\n",
    "\n",
    "\n",
    "这个比赛的链接：http://sofasofa.io/competition.php?id=2\n",
    "\n",
    "\n",
    "* 任务类型：二元分类\n",
    "\n",
    "* 背景介绍：在交通摩擦（事故）发生后，理赔员会前往现场勘察、采集信息，这些信息往往影响着车主是否能够得到保险公司的理赔。训练集数据包括理赔人员在现场对该事故方采集的36条信息，信息已经被编码，以及该事故方最终是否获得理赔。我们的任务是根据这36条信息预测该事故方没有被理赔的概率。\n",
    "\n",
    "* 数据介绍：训练集中共有200000条样本，预测集中有80000条样本。 \n",
    "![data_description](images/data_description.png)\n",
    "\n",
    "* 评价方法：Precision-Recall AUC\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Demo code\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "# read data\n",
    "homePath = \"data\"\n",
    "trainPath = os.path.join(homePath, \"train.csv\")\n",
    "testPath = os.path.join(homePath, \"test.csv\")\n",
    "submitPath = os.path.join(homePath, \"sample_submit.csv\")\n",
    "trainData = pd.read_csv(trainPath)\n",
    "testData = pd.read_csv(testPath)\n",
    "submitData = pd.read_csv(submitPath)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "参照数据说明，CaseID这列是没有意义的编号，因此这里将他丢弃。\n",
    "\n",
    "~drop()函数：axis指沿着哪个轴，0为行，1为列；inplace指是否在原数据上直接操作\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 去掉没有意义的一列\n",
    "trainData.drop(\"CaseId\", axis=1, inplace=True)\n",
    "testData.drop(\"CaseId\", axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 快速了解数据\n",
    "\n",
    "~head()：默认显示前5行数据，可指定显示多行，例如.head(15)显示前15行\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "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>Q1</th>\n",
       "      <th>Q2</th>\n",
       "      <th>Q3</th>\n",
       "      <th>Q4</th>\n",
       "      <th>Q5</th>\n",
       "      <th>Q6</th>\n",
       "      <th>Q7</th>\n",
       "      <th>Q8</th>\n",
       "      <th>Q9</th>\n",
       "      <th>Q10</th>\n",
       "      <th>...</th>\n",
       "      <th>Q28</th>\n",
       "      <th>Q29</th>\n",
       "      <th>Q30</th>\n",
       "      <th>Q31</th>\n",
       "      <th>Q32</th>\n",
       "      <th>Q33</th>\n",
       "      <th>Q34</th>\n",
       "      <th>Q35</th>\n",
       "      <th>Q36</th>\n",
       "      <th>Evaluation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>15 rows × 37 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    Q1  Q2  Q3  Q4  Q5  Q6  Q7  Q8  Q9  Q10     ...      Q28  Q29  Q30  Q31  \\\n",
       "0    0   0   0   0   0   0   0   0   0    0     ...        0    0    0    0   \n",
       "1    0   0   0   0   0   0   0   0   0    0     ...        0    1    1    1   \n",
       "2    0   0   0   0   0   0   0   1   0    0     ...        1    2    2    2   \n",
       "3    0   0   0   0   0   0   0   0   0    0     ...        1    3    2    3   \n",
       "4    0   0   0   0   0   0   0   0   0    0     ...        1    4    2    4   \n",
       "5    0   0   0   0   0   0   0   0   0    0     ...        1    2    3    5   \n",
       "6    0   0   0   0   0   0   0   0   0    1     ...        0    3    1    6   \n",
       "7    0   0   0   0   0   0   0   0   0    0     ...        1    3    1    3   \n",
       "8    0   0   0   0   0   0   0   2   0    0     ...        0    2    1    2   \n",
       "9    0   0   0   0   0   0   0   0   0    0     ...        0    2    1    7   \n",
       "10   0   0   0   0   0   0   0   0   0    0     ...        2    5    0    8   \n",
       "11   0   0   0   0   0   0   0   0   0    0     ...        0    2    1    1   \n",
       "12   1   0   0   0   0   0   0   0   0    0     ...        3    3    3    9   \n",
       "13   0   0   0   0   0   0   0   0   0    0     ...        0    1    1   10   \n",
       "14   0   0   0   0   0   0   0   3   0    0     ...        1    6    1    2   \n",
       "\n",
       "    Q32  Q33  Q34  Q35  Q36  Evaluation  \n",
       "0     0    0    0    0    0           0  \n",
       "1     1    0    0    0    0           0  \n",
       "2     1    0    0    0    0           0  \n",
       "3     1    0    0    1    1           0  \n",
       "4     1    0    0    1    1           0  \n",
       "5     1    0    0    0    0           0  \n",
       "6     1    0    0    1    1           1  \n",
       "7     1    0    0    1    1           1  \n",
       "8     1    0    0    0    0           0  \n",
       "9     1    0    0    0    0           0  \n",
       "10    1    0    0    1    1           0  \n",
       "11    1    0    0    0    0           0  \n",
       "12    1    0    0    1    1           0  \n",
       "13    1    0    0    0    0           0  \n",
       "14    1    0    0    1    1           0  \n",
       "\n",
       "[15 rows x 37 columns]"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trainData.head(15)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "显示数据简略信息，可以每列有多少非空的值，以及每列数据对应的数据类型。\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 200000 entries, 0 to 199999\n",
      "Data columns (total 37 columns):\n",
      "Q1            200000 non-null int64\n",
      "Q2            200000 non-null int64\n",
      "Q3            200000 non-null int64\n",
      "Q4            200000 non-null int64\n",
      "Q5            200000 non-null int64\n",
      "Q6            200000 non-null int64\n",
      "Q7            200000 non-null int64\n",
      "Q8            200000 non-null int64\n",
      "Q9            200000 non-null int64\n",
      "Q10           200000 non-null int64\n",
      "Q11           200000 non-null int64\n",
      "Q12           200000 non-null int64\n",
      "Q13           200000 non-null int64\n",
      "Q14           200000 non-null int64\n",
      "Q15           200000 non-null int64\n",
      "Q16           200000 non-null int64\n",
      "Q17           200000 non-null int64\n",
      "Q18           200000 non-null int64\n",
      "Q19           200000 non-null int64\n",
      "Q20           200000 non-null int64\n",
      "Q21           200000 non-null int64\n",
      "Q22           200000 non-null int64\n",
      "Q23           200000 non-null int64\n",
      "Q24           200000 non-null int64\n",
      "Q25           200000 non-null int64\n",
      "Q26           200000 non-null int64\n",
      "Q27           200000 non-null int64\n",
      "Q28           200000 non-null int64\n",
      "Q29           200000 non-null int64\n",
      "Q30           200000 non-null int64\n",
      "Q31           200000 non-null int64\n",
      "Q32           200000 non-null int64\n",
      "Q33           200000 non-null int64\n",
      "Q34           200000 non-null int64\n",
      "Q35           200000 non-null int64\n",
      "Q36           200000 non-null int64\n",
      "Evaluation    200000 non-null int64\n",
      "dtypes: int64(37)\n",
      "memory usage: 56.5 MB\n"
     ]
    }
   ],
   "source": [
    "trainData.info()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "~hist():绘制直方图，参数figsize可指定输出图片的尺寸。\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce92a6f28>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9247518>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce925f860>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9276ef0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9215588>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce922ba90>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce91c8160>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce91e17f0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce91f8e80>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9194550>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce91acbe0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce91492b0>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9160940>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9178fd0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9115630>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce912fcc0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce90cf390>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce90e6a20>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce90810f0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9099780>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce90b2e10>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce904d4e0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9065b70>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9004240>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce901c8d0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce9033f60>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8fd0630>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8fe8cc0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f86390>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f9ca20>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8fba0f0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f53780>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f6ae10>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f064e0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f1eb70>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8f3b240>],\n",
       "       [<matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8ed28d0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8eebf60>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8e89630>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8ea1cc0>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8ec0390>,\n",
       "        <matplotlib.axes._subplots.AxesSubplot object at 0x7fcce8e56a20>]],\n",
       "      dtype=object)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1440 with 42 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainData.hist(figsize=(20, 20))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "想要了解特征之间的相关性，可计算相关系数矩阵。然后可对某个特征来排序。\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Evaluation    1.000000\n",
       "Q28           0.410700\n",
       "Q30           0.324421\n",
       "Q36           0.302709\n",
       "Q35           0.224996\n",
       "Q34           0.152743\n",
       "Q32           0.049397\n",
       "Q21           0.034897\n",
       "Q33           0.032248\n",
       "Q13           0.023603\n",
       "Q8            0.021922\n",
       "Q19           0.019694\n",
       "Q20           0.013903\n",
       "Q4            0.011626\n",
       "Q27           0.004262\n",
       "Q23           0.002898\n",
       "Q7            0.001143\n",
       "Q31          -0.000036\n",
       "Q14          -0.000669\n",
       "Q29          -0.002014\n",
       "Q10          -0.002711\n",
       "Q12          -0.005287\n",
       "Q1           -0.006511\n",
       "Q16          -0.007184\n",
       "Q18          -0.007643\n",
       "Q26          -0.008188\n",
       "Q11          -0.009252\n",
       "Q24          -0.010891\n",
       "Q22          -0.011821\n",
       "Q25          -0.012660\n",
       "Q6           -0.016072\n",
       "Q2           -0.018307\n",
       "Q15          -0.019570\n",
       "Q9           -0.021261\n",
       "Q5           -0.023893\n",
       "Q3           -0.026349\n",
       "Q17          -0.028461\n",
       "Name: Evaluation, dtype: float64"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix = trainData.corr()\n",
    "corr_matrix[\"Evaluation\"].sort_values(ascending=False) # ascending=False 降序排列"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "从训练集中分离标签"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = trainData['Evaluation']\n",
    "trainData.drop(\"Evaluation\", axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "使用K-Means训练模型\n",
    "\n",
    "KMeans()：\n",
    "* `n_clusters`指要预测的有几个类；\n",
    "* `init`指初始化中心的方法，默认使用的是`k-means++`方法，而非经典的K-means方法的随机采样初始化，当然你可以设置为random使用随机初始化；\n",
    "* `n_jobs`指定使用CPU核心数，-1为使用全部CPU。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# do k-means\n",
    "from sklearn.cluster import KMeans\n",
    "est = KMeans(n_clusters=2, init=\"k-means++\", n_jobs=-1)\n",
    "est.fit(trainData, y)\n",
    "\n",
    "y_train = est.predict(trainData)\n",
    "y_pred = est.predict(testData)\n",
    "\n",
    "# 保存预测的结果\n",
    "submitData['Evaluation'] = y_pred\n",
    "submitData.to_csv(\"submit_data.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "acc_train = 0.682140\n"
     ]
    }
   ],
   "source": [
    "# calculate accuracy\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "acc_train = accuracy_score(y, y_train)\n",
    "print(\"acc_train = %f\" % (acc_train))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 随机森林\n",
    "\n",
    "使用K-means可能得到的结果没那么理想。在官网上，举办方给出了两个标杆模型，效果最好的是随机森林。以下是代码，读者可以自己测试。\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "\n",
    "# 读取数据\n",
    "train = pd.read_csv(\"data/train.csv\")\n",
    "test = pd.read_csv(\"data/test.csv\")\n",
    "submit = pd.read_csv(\"data/sample_submit.csv\")\n",
    "\n",
    "# 删除id\n",
    "train.drop('CaseId', axis=1, inplace=True)\n",
    "test.drop('CaseId', axis=1, inplace=True)\n",
    "\n",
    "# 取出训练集的y\n",
    "y_train = train.pop('Evaluation')\n",
    "\n",
    "# 建立随机森林模型\n",
    "clf = RandomForestClassifier(n_estimators=100, random_state=0)\n",
    "clf.fit(train, y_train)\n",
    "y_pred = clf.predict_proba(test)[:, 1]\n",
    "\n",
    "# 输出预测结果至my_RF_prediction.csv\n",
    "submit['Evaluation'] = y_pred\n",
    "submit.to_csv('my_RF_prediction.csv', index=False)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.00177294 0.00207449 0.00187096 0.00471492 0.00443815 0.0029538\n",
      " 0.00364967 0.00652341 0.00235713 0.00739511 0.00245106 0.00106103\n",
      " 0.0007513  0.00090631 0.00150727 0.0037793  0.00183821 0.00196833\n",
      " 0.00209665 0.00726069 0.00816243 0.00107563 0.00559247 0.00766561\n",
      " 0.00760666 0.00028462 0.00025573 0.18472067 0.25559838 0.21436631\n",
      " 0.0425301  0.00662325 0.00297955 0.03148822 0.03907383 0.13060584]\n",
      "[0 0 0 ... 0 0 0]\n",
      "acc_train = 0.931525\n"
     ]
    }
   ],
   "source": [
    "# freature importances\n",
    "print(clf.feature_importances_)\n",
    "\n",
    "# Train accuracy\n",
    "from sklearn.metrics import accuracy_score\n",
    "y_train_pred = clf.predict(train)\n",
    "print(y_train_pred)\n",
    "\n",
    "acc_train = accuracy_score(y_train, y_train_pred)\n",
    "print(\"acc_train = %f\" % (acc_train))"
   ]
  }
 ],
 "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": "3.5.2"
  },
  "main_language": "python"
 },
 "nbformat": 4,
 "nbformat_minor": 2
}