titanic/kaggle_titanic.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2D3940F0AF89478480A8B8B999B95629",
    "mdEditEnable": false
   },
   "source": [
    "# Kaggle入门：逻辑回归应用之Kaggle泰坦尼克之灾\n",
    "\n",
    "**原文链接：http://blog.csdn.net/han_xiaoyang/article/details/49797143 在K-lab上进行实现，对机器学习和kaggle感兴趣的小伙伴，欢迎点击右上角「Fork」按钮自己进行运行调试！**\n",
    "\n",
    "> 作者： 寒小阳 \n",
    "> 出处：http://blog.csdn.net/han_xiaoyang/article/details/49797143 \n",
    "> 声明：版权所有，转载请注明出处，谢谢。\n",
    "\n",
    "Github链接：https://github.com/HanXiaoyang/Kaggle_Titanic\n",
    "\n",
    "## 背景\n",
    "### 关于Kaggle\n",
    "\n",
    "*   [我是Kaggle地址，翻我牌子](https://www.kaggle.com/)\n",
    "*   这就是那个无数『数据挖掘先驱』们，在回答”枪我有了，哪能找到靶子练练手啊？”时候的答案！\n",
    "*   这是一个要数据有数据，要实际应用场景有场景，要一起在数据挖掘领域high得不要不要的小伙伴就有小伙伴的地方啊！！！\n",
    "\n",
    "Kaggle是一个数据分析建模的应用竞赛平台，有点类似KDD-CUP（国际知识发现和数据挖掘竞赛），企业或者研究者可以将问题背景、数据、期望指标等发布到Kaggle上，以竞赛的形式向广大的数据科学家征集解决方案。而热爱数(dong)据(shou)挖(zhe)掘(teng)的小伙伴们可以下载/分析数据，使用统计/机器学习/数据挖掘等知识，建立算法模型，得出结果并提交，排名top的可能会有奖金哦！\n",
    "\n",
    "### 关于泰坦尼克号之灾\n",
    "\n",
    "- 带大家去该问题页面溜达一圈吧\n",
    "\n",
    " - 下面是问题背景页 \n",
    "\t\n",
    "![问题背景页](https://www.z4a.net/images/2018/11/28/100.jpg)\n",
    "\t\n",
    " - 下面是可下载Data的页面 \n",
    "![可下载Data的页面](https://www.z4a.net/images/2018/11/28/3c77dc7ee0957a2f3.jpg)\n",
    " - 下面是小伙伴们最爱的forum页面，你会看到各种神级人物厉(qi)害(pa)的数据处理/建模想法，你会直视『世界真奇妙』。 \n",
    "![forum页面](https://www.z4a.net/images/2018/11/28/10047af6da58b79a1f2.jpg)\n",
    "- 泰坦尼克号问题之背景\n",
    "\n",
    "\t- 就是那个大家都熟悉的『Jack and Rose』的故事，豪华游艇倒了，大家都惊恐逃生，可是救生艇的数量有限，无法人人都有，副船长发话了『lady and kid first！』，所以是否获救其实并非随机，而是基于一些背景有rank先后的。\n",
    "\n",
    "\t- 训练和测试数据是一些乘客的个人信息以及存活状况，要尝试根据它生成合适的模型并预测其他人的存活状况。\n",
    "\n",
    "\t- 对，这是一个二分类问题，是我们之前讨论的logistic regression所能处理的范畴。\n",
    "\t\n",
    "## 说明\n",
    "接触过Kaggle的同学们可能知道这个问题，也可能知道RandomForest和SVM等等算法，甚至还对多个模型做过融合，取得过非常好的结果，那maybe这篇文章并不是针对你的，你可以自行略过。\n",
    "\n",
    "我们因为之前只介绍了Logistic Regression这一种分类算法。所以本次的问题解决过程和优化思路，都集中在这种算法上。其余的方法可能我们之后的文章里会提到。\n",
    "\n",
    "说点个人的观点。不一定正确。 \n",
    "『解决一个问题的方法和思路不止一种』 \n",
    "『没有所谓的机器学习算法优劣，也没有绝对高性能的机器学习算法，只有在特定的场景、数据和特征下更合适的机器学习算法。』\n",
    "\n",
    "## 怎么做？\n",
    "手把手教程马上就来，先来两条我看到的，觉得很重要的经验。\n",
    "\n",
    "印象中Andrew Ng老师似乎在coursera上说过，应用机器学习，千万不要一上来就试图做到完美，先撸一个baseline的model出来，再进行后续的分析步骤，一步步提高，所谓后续步骤可能包括『分析model现在的状态(欠/过拟合)，分析我们使用的feature的作用大小，进行feature selection，以及我们模型下的bad case和产生的原因』等等。\n",
    "\n",
    "Kaggle上的大神们，也分享过一些experience，说几条我记得的哈：\n",
    "\n",
    "『对数据的认识太重要了！』\n",
    "『数据中的特殊点/离群点的分析和处理太重要了！』\n",
    "『特征工程(feature engineering)太重要了！在很多Kaggle的场景下，甚至比model本身还要重要』\n",
    "『要做模型融合(model ensemble)啊啊啊！』\n",
    "\n",
    "## 初探数据\n",
    "先看看我们的数据，长什么样吧。在Data下我们train.csv和test.csv两个文件，分别存着官方给的训练和测试数据。\n",
    "\n",
    "pandas是常用的python数据处理包，把csv文件读入成dataframe各式，我们在ipython notebook中，看到data_train如下所示："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false,
    "id": "DFC95DE04483465083C0140B0D735524",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</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>1</td>\n",
       "      <td>0</td>\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>2</td>\n",
       "      <td>1</td>\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>1</td>\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>4</td>\n",
       "      <td>1</td>\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>5</td>\n",
       "      <td>0</td>\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": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "\n",
       "                                                Name     Sex   Age  SibSp  \\\n",
       "0                            Braund, Mr. Owen Harris    male  22.0      1   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.0      1   \n",
       "2                             Heikkinen, Miss. Laina  female  26.0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.0      1   \n",
       "4                           Allen, Mr. William Henry    male  35.0      0   \n",
       "\n",
       "   Parch            Ticket     Fare Cabin Embarked  \n",
       "0      0         A/5 21171   7.2500   NaN        S  \n",
       "1      0          PC 17599  71.2833   C85        C  \n",
       "2      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3      0            113803  53.1000  C123        S  \n",
       "4      0            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\", message=\"numpy.dtype size changed\")\n",
    "warnings.filterwarnings(\"ignore\", message=\"numpy.ufunc size changed\")\n",
    "#https://stackoverflow.com/q/40845304/10704205\n",
    "#Ignore RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility\n",
    "\n",
    "import pandas as pd #数据分析\n",
    "import numpy as np #科学计算\n",
    "from pandas import Series,DataFrame\n",
    "\n",
    "data_train = pd.read_csv('train.csv',engine = 'python',encoding='UTF-8')\n",
    "data_train.head() #dataframe格式"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0C9407007EEC48D78DEA889D00BFFFB5",
    "mdEditEnable": false
   },
   "source": [
    "这就是典型的dataframe格式，如果你没接触过这种格式，完全没有关系，你就把它想象成Excel里面的列好了。 \n",
    "我们看到，总共有12列，其中Survived字段表示的是该乘客是否获救，其余都是乘客的个人信息，包括：\n",
    "\n",
    "- `PassengerId` => 乘客ID\n",
    "- `Pclass` => 乘客等级(1/2/3等舱位)\n",
    "- `Name` => 乘客姓名\n",
    "- `Sex` => 性别\n",
    "- `Age` => 年龄\n",
    "- `SibSp` => 堂兄弟/妹个数\n",
    "- `Parch` => 父母与小孩个数\n",
    "- `Ticket` => 船票信息\n",
    "- `Fare` => 票价\n",
    "- `Cabin` => 客舱\n",
    "- `Embarked` => 登船港口\n",
    "\n",
    "逐条往下看，要看完这么多条，眼睛都有一种要瞎的赶脚。好吧，我们让dataframe自己告诉我们一些信息，如下所示："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false,
    "id": "878F32156AC54EFD832D9C6E816038C8",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 891 entries, 0 to 890\n",
      "Data columns (total 12 columns):\n",
      " #   Column       Non-Null Count  Dtype  \n",
      "---  ------       --------------  -----  \n",
      " 0   PassengerId  891 non-null    int64  \n",
      " 1   Survived     891 non-null    int64  \n",
      " 2   Pclass       891 non-null    int64  \n",
      " 3   Name         891 non-null    object \n",
      " 4   Sex          891 non-null    object \n",
      " 5   Age          714 non-null    float64\n",
      " 6   SibSp        891 non-null    int64  \n",
      " 7   Parch        891 non-null    int64  \n",
      " 8   Ticket       891 non-null    object \n",
      " 9   Fare         891 non-null    float64\n",
      " 10  Cabin        204 non-null    object \n",
      " 11  Embarked     889 non-null    object \n",
      "dtypes: float64(2), int64(5), object(5)\n",
      "memory usage: 83.7+ KB\n"
     ]
    }
   ],
   "source": [
    "data_train.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2E77A7E73003473DBE95365B98118688",
    "mdEditEnable": false
   },
   "source": [
    "上面的数据说啥了？它告诉我们，训练数据中总共有891名乘客，但是很不幸，我们有些属性的数据不全，比如说：\n",
    "\n",
    "- Age（年龄）属性只有714名乘客有记录\n",
    "- Cabin（客舱）更是只有204名乘客是已知的\n",
    "\n",
    "似乎信息略少啊，想再瞄一眼具体数据数值情况呢？恩，我们用下列的方法，得到数值型数据的一些分布(因为有些属性，比如姓名，是文本型；而另外一些属性，比如登船港口，是类目型。这些我们用下面的函数是看不到的)：\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false,
    "id": "5DDB3F0E16054E048EA1C5EDC1BB8C6D",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>714.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "      <td>891.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.383838</td>\n",
       "      <td>2.308642</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>0.523008</td>\n",
       "      <td>0.381594</td>\n",
       "      <td>32.204208</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>257.353842</td>\n",
       "      <td>0.486592</td>\n",
       "      <td>0.836071</td>\n",
       "      <td>14.526497</td>\n",
       "      <td>1.102743</td>\n",
       "      <td>0.806057</td>\n",
       "      <td>49.693429</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.420000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>223.500000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>20.125000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>7.910400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>446.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>28.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>14.454200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>668.500000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>31.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>891.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>80.000000</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>6.000000</td>\n",
       "      <td>512.329200</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       PassengerId    Survived      Pclass         Age       SibSp  \\\n",
       "count   891.000000  891.000000  891.000000  714.000000  891.000000   \n",
       "mean    446.000000    0.383838    2.308642   29.699118    0.523008   \n",
       "std     257.353842    0.486592    0.836071   14.526497    1.102743   \n",
       "min       1.000000    0.000000    1.000000    0.420000    0.000000   \n",
       "25%     223.500000    0.000000    2.000000   20.125000    0.000000   \n",
       "50%     446.000000    0.000000    3.000000   28.000000    0.000000   \n",
       "75%     668.500000    1.000000    3.000000   38.000000    1.000000   \n",
       "max     891.000000    1.000000    3.000000   80.000000    8.000000   \n",
       "\n",
       "            Parch        Fare  \n",
       "count  891.000000  891.000000  \n",
       "mean     0.381594   32.204208  \n",
       "std      0.806057   49.693429  \n",
       "min      0.000000    0.000000  \n",
       "25%      0.000000    7.910400  \n",
       "50%      0.000000   14.454200  \n",
       "75%      0.000000   31.000000  \n",
       "max      6.000000  512.329200  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B6D0B24DA7CD46538040DB87EFDEA1BC",
    "mdEditEnable": false
   },
   "source": [
    "我们从上面看到更进一步的什么信息呢？ \n",
    "mean字段告诉我们，大概0.383838的人最后获救了，2/3等舱的人数比1等舱要多，平均乘客年龄大概是29.7岁(计算这个时候会略掉无记录的)等等…"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7AAF8DDD496A4E429B47B60927BC95D5",
    "mdEditEnable": false
   },
   "source": [
    "## 数据初步分析\n",
    "\n",
    "每个乘客都这么多属性，那我们咋知道哪些属性更有用，而又应该怎么用它们啊？说实话这会儿我也不知道，但我们记得前面提到过\n",
    "\n",
    "**- 『对数据的认识太重要了！』**\n",
    "\n",
    "**- 『对数据的认识太重要了！』**\n",
    "\n",
    "**- 『对数据的认识太重要了！』**\n",
    "\n",
    "重要的事情说三遍，恩，说完了。仅仅最上面的对数据了解，依旧无法给我们提供想法和思路。我们再深入一点来看看我们的数据，看看每个/多个 属性和最后的Survived之间有着什么样的关系呢。\n",
    "\n",
    "### 乘客各属性分布"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false,
    "id": "7D84F87CCD8D4DECA515BA51E3931BDF",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline \n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import rcParams\n",
    "\n",
    "#import matplotlib.font_manager as fm\n",
    "\n",
    "#print(plt.rcParams['font.sans-serif'])\n",
    "\n",
    "# 获取字体列表并输出字体路径\n",
    "#for font in fm.fontManager.ttflist:\n",
    "#    if 'Heiti' in font.name:\n",
    "#        print(font.name, font.fname)\n",
    "\n",
    "# 配置 matplotlib 使用该字体\n",
    "rcParams['font.sans-serif'] = ['Heiti TC']\n",
    "rcParams['axes.unicode_minus'] = False  # 解决负号显示问题\n",
    "\n",
    "fig = plt.figure()\n",
    "fig.set(alpha=0.2)  # 设定图表颜色alpha参数\n",
    "\n",
    "plt.subplot2grid((2, 3), (0, 0))  # 在一张大图里分列几个小图\n",
    "data_train.Survived.value_counts().plot(kind='bar')  # plots a bar graph of those who survived vs those who did not.\n",
    "plt.title(u\"获救情况 (1为获救)\")  # puts a title on our graph\n",
    "plt.ylabel(u\"人数\")\n",
    "\n",
    "plt.subplot2grid((2, 3), (0, 1))\n",
    "data_train.Pclass.value_counts().plot(kind=\"bar\")\n",
    "plt.ylabel(u\"人数\")\n",
    "plt.title(u\"乘客等级分布\")\n",
    "\n",
    "plt.subplot2grid((2, 3), (0, 2))\n",
    "plt.scatter(data_train.Survived, data_train.Age)\n",
    "plt.ylabel(u\"年龄\")  # sets the y axis label\n",
    "plt.grid(visible=True, which='major', axis='y')  # Use visible=True instead of b=True\n",
    "plt.title(u\"按年龄看获救分布 (1为获救)\")\n",
    "\n",
    "plt.subplot2grid((2, 3), (1, 0), colspan=2)\n",
    "data_train.Age[data_train.Pclass == 1].plot(kind='kde')  # plots a kernel density estimate of the subset of the 1st class passengers' age\n",
    "data_train.Age[data_train.Pclass == 2].plot(kind='kde')\n",
    "data_train.Age[data_train.Pclass == 3].plot(kind='kde')\n",
    "plt.xlabel(u\"年龄\")  # plots an axis label\n",
    "plt.ylabel(u\"密度\")\n",
    "plt.title(u\"各等级的乘客年龄分布\")\n",
    "plt.legend((u'头等舱', u'2等舱', u'3等舱'), loc='best')  # sets our legend for our graph.\n",
    "\n",
    "plt.subplot2grid((2, 3), (1, 2))\n",
    "data_train.Embarked.value_counts().plot(kind='bar')\n",
    "plt.title(u\"各登船口岸上船人数\")\n",
    "plt.ylabel(u\"人数\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A329BAF0557149368662B041B9D39B74",
    "mdEditEnable": false
   },
   "source": [
    "我们在图上可以看出来，被救的人300多点，不到半数；3等舱乘客灰常多；遇难和获救的人年龄似乎跨度都很广；3个不同的舱年龄总体趋势似乎也一致，2/3等舱乘客20岁多点的人最多，1等舱40岁左右的最多(→_→似乎符合财富和年龄的分配哈，咳咳，别理我，我瞎扯的)；登船港口人数按照S、C、Q递减，而且S远多于另外俩港口。\n",
    "\n",
    "我们可能会有一些想法了：\n",
    "\n",
    "- **不同舱位/乘客等级可能和财富/地位有关系，最后获救概率可能会不一样**\n",
    "\n",
    "- **年龄对获救概率也一定是有影响的，毕竟前面说了，副船长还说『小孩和女士先走』呢**\n",
    "\n",
    "- **和登船港口是不是有关系呢？也许登船港口不同，人的出身地位不同？**\n",
    "\n",
    "口说无凭，空想无益。老老实实再来统计统计，看看这些属性值的统计分布吧。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A097A57249B140889E39306B2C44A70B",
    "mdEditEnable": false
   },
   "source": [
    "### 属性与获救结果的关联统计"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A8B155CD2A154348A15AB35464861C24",
    "mdEditEnable": false
   },
   "source": [
    "#### 看看各乘客等级的获救情况"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false,
    "id": "2CC0D1CFD0E54BCE8486ADB8FE0E2EAD",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAHACAYAAACvXBIfAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAAQPJJREFUeJzt3Qm8znX+//+X7PteZMmuQkLIaEFIkYSmSAslSlrIVqZQkzSqyaDIJDQmUQySsiSSktJmGbLEENnJmlz/2/P1+3+u73WOc4wm51zX+Xjcb7drzjnX51o+13XOdD29Xu8lUyQSiRgAAECInBPvEwAAADjTCDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDhAgvvss8+se/futnbt2jP6uPv377cff/zxtG//ww8/WMuWLe2111477fscOXLEOnfubAsWLEhy/c6dO+2KK66wf/7zn6e8/y+//GL333+//ec///GfT5w4keLttF7p0aNHLa1NmjTJFi9eHP35ySeftDfeeOO07vv5559b7969bePGjScdmzx5sv3lL3+xf//736nef/fu3Slef/z4cevUqdNJ7/GpHDp0yH7++Wd/35I7duyY7dmz57QfC0hUBBwgwb344ov24YcfWuHChU/7PvqgVCD67rvv7JNPPrEZM2bY3//+dxswYIB17NjRateubYUKFbJbbrklxft///33Hiq2b9/uYUQfePny5fPHeuedd2zv3r3+gbtjxw4PSRs2bPAPzeReffVVDwDZsmXzn/WhqtAzbtw4W79+vV1zzTUeYhS2Vq5caT/99FOS+8+ePdtefvllGzVqlP+s9yBXrlyWP39+/14XnVeWLFk8bKTkH//4h5UrV87Gjh2b5PqvvvrKKlSoYP3790/xfocPH7ZVq1bZ1q1b/bz0PvTt29cDjr7XZd68eR789D7ovdq8ebN9++23KT6ejivElC5d2v72t7/Z9ddfHw0YCjhPP/30KX/Hl156qbVv396/1+9j06ZN/rwHDx7013bgwAG/6Hn03uo9Dd7zBx980J8joO/z5s1r55xzjr93wSVz5syWPXt2a9WqVarnAWQY2qoBQGJavHhxJFOmTPoUTPVy1113nXS/li1b+rGcOXNGzj///Mgll1wSKV68eOSCCy6I9OjRI/LMM89EXnnllciYMWMie/fuPen+2bNnP+VzpnT58MMPkzzGxo0bI3nz5j3pdjNmzIgUK1Ysxcd47733kjzGjTfeGKlWrVrkl19+8Z+nTp0aad++feTiiy+OrFq1yi8TJ070++q9Sslf//rXyDnnnBPZvn17kuv1+nW/L7/8MsX7ffvtt7/5PQguKZk1a1b02NKlSyO5c+eOLF++PPLrr7/6+6Hfy6lUrFgxMmTIEP/+xRdf9MfS68qRI4d/r9+ZLsHfyz//+U+/7dGjR/39KlGiRGT//v1+3YQJEyLZsmWLvoexlxtuuCHSuHHjU54LkBFkiXfAApAy/cv77rvv9n9VqzrRtm3bJMd79OhhS5YssUGDBkWv07/aVV0ZMWKEV2zy5Mnj98+UKZN169bNtm3bZkOGDPHKgVobqp6ooqJ/8ZcoUcL/9R545ZVX7LbbbotWOypWrGi5c+c+6Tx37dplZcqUSXKdzkPtrGbNmtkLL7xgf/jDH7xVdeONN1qvXr2sSpUqtnz5cq+AqL2yevVqrwDpHAJq5cycOdPmzp3r1QW1qtSqu+CCC7zCdOGFF0arEeeee65dfvnlSc5Br0uvW5WsmjVr+m1+/fVXr8zofZk1a5ZddNFFljNnTn9+tb9UAdHrVHVLX/W+qEKk9yVr1qxWvXp1r4bce++9/hzXXnutV2J0bsF7uW/fvug5TJ061atoqjrpnGXYsGHeBlIlTV9VEVLVRY8R0PU6dz1nQNUW/c6mT5/u76Wet1ixYn5uOXLksGnTptmVV17pr09/AyVLlvT7qXqmv4eGDRvaY4895tUjPZYeP3gPY+n1pke7D0hz8U5YAE6mikWrVq28evH2229HChcuHNmyZUv0+LBhw/xf76poxNK/zP/XqoMqFoFKlSpFJk+eHK3E5MqVK3L77bdHduzYkeSydetW/6rK0JIlS6L379mzZ6R27dqRn3/+2X+uV69epGvXrpG1a9dG/va3v0U+/fRTv/7777/35164cOFJ78E999zj74EsW7bMKxNvvPFG5LHHHotceeWV0dv94Q9/iHTs2PGk++t9C16b3itVLPQY3bt3j+zevTta+Uh++eKLL5I8zn/+85/o988//3xk3bp10Z9VgdF7EDhx4kSS+w4cODBSp06dyNVXXx2pUaOGP36TJk38Z1XV9B7feeedKZ6HKl1yxx13RJ544olI5cqVvcqTJ0+eyAcffBB9XfrdBNW6LFmy+PeqfCXXtm1bP75hwwav7qjak5Lbbrstcu2116Z4DMhICDhAAlL7RC2Mr7/+2n/Wh2C5cuUi//73vyODBw+OZM2aNTJ27NiT7nfgwAH/wNUHuL7XB6E+sPbt2xe59957vXV17NixSMOGDSOvvfaa32bFihWRTZs2eSsjOYWXqlWrphqKrrnmmlRfw+HDhyNz5syJzJ492wOMAlD//v39fvPmzYt88skn/lUftjqH5OFEr1FtL70HOm/dVx566KFI8+bN/fs9e/ZEMmfOHHn22We91RT7OEEI0HV6nbqULl060qdPn8jTTz8dyZcvn1+XvIWk9zigtpYCxZ/+9KfIzJkz/bn0/geeeuqp6H0UCkaPHn3S+xC019R+C/5NGVyn0KrXqTabwp9C5OWXX+5hU69NLr30Um8pKuD85S9/8e/1/mzevDly/Phxv40eV++16Peov4HY4BWEySBIKiieKuAohAEZHS0qIAF16dLFrrvuOh+QKmPGjLEbbrjBWzv6PPvXv/5lzZs3P+l+ar2oDaXBp2qLaACsZteo7aCWRND2ULtErR7NCurZs6ctWrTISpUqleSxPvjgA2+jXHbZZdahQwfr2rWrD+5VK+fOO+/09pIGC6dGbRO1ozQAWeelVkowEFmPpZaO2ilqxWl20RNPPBG9r1osaq3pedT2KVu2bHQQsdpIBQsW9O/nzJnj7aunnnrKBwDr68UXX+zH9FpF74OeP7hOg6NHjhzpbabgetFgXYltw6mtpYHOeq/1vPfcc4+3qGJfox6/UqVKVrlyZXv99de9fRRQS0nvn1prse23xx9/3Ft0ej814Ltdu3Z+vd6f8uXLW9WqVf1n/a71Owz+DqRfv37ejtPsOgnaTWpZ6XejNpxowLIGQgf0uLoEj6uLWnPJqb2Y2mw1ICNhFhWQoPShpg93jRXRB73GotSoUcMDisbmaMzGW2+9ZV9//bXP8gk+oDVrSWNFNL4kdlxHLH0oKkRpRpSmagcffLHjav70pz/5uA19aM+fP9/q1Klj7777rgevFStW+Pif2A/tWAouCjD6EFdY0YysdevW+SwiWbNmjQcfhTEFHIUb3UdjT+T222/3D/svv/zSihQpYhMmTPAgI5rVpPfgz3/+s8/20Xuk2V3BmJjk9Lz6INdF56TxNXpPmzZt6sExEISv2ICjsSi6rej5FXA0XkZjkjQ+R2FJIUmvT2OJFJyWLl0avb/uo7FTCiXBuBa9J8OHD/fwp6njCnJvvvmmH9NjKizFzobTjKnk4VPjcPQ70mtXiFIg0Xun16fn0XmlNJtL750oBOn2CoPVqlXzEBZ8r99x8HsAMrR4l5AA/B/NaFK7pUOHDt6aUBtBbZHrrrvOWy7BmJBevXpFihYtmqRdpNaCqAWlmTkyf/58v53aE507d/YZMlK2bFkfC6MWh55TLark1OoIxpSsX78+Ur58eX8ejfl4//33T/k6dC66rc5dt8+fP3+kYMGC3u7R9YUKFYoUKFDAZ1lpLEww80djiAJqs6ldo3ZdcD5y0UUX+XukMSfBWB7NLipVqlSSMTBq2aTUVlOLKhjHpHE5GmMkI0eO9ON6/0RtPbWF6tat6++fZqu1aNEi2pbS+BeNadFrDF6Hrtf7FEuPo1lTQYvqyJEj0Rlnhw4ditxyyy3e9tL1Op/YmWQ6F52fjgUtqqB1Fry/wewpXYJxRmqppdQyLFOmTOSjjz7y91K/+/r160c6derkxzWOKmi/HTx48JS/XyAjoEUFJBC1LFQB+OKLL7x6Uq9ePWvTpo23aFR1CRaJU2VGFQO1aNRu0r/Mg5k9+te9/hWuf8F/88033orQWjpaD0b/wv/oo4+8EqS2iS6iasJzzz0XPQ/9C19rwGj9Hc3O0UwfzYDSbBxVjdQuK1q0qDVo0MDbZpqxo5lUmsUjaiep7aRqSDADSZegTSYff/yxP46qGKqqqIJTvHhxP6ZqlCoKqvDotWnGmB5bry2YFaXn1vo+devWtYkTJ3qbJ3jsWKoUBbO8Ymd7qU327LPP2qOPPuozsVTBCc5T1NZTZUXtIp2j2oLBGj46H7Wm9DtQFUwVNM1W0vumilQsPY6qQEG1Sy2g4PeoGVz6Xej3rZaffsf169eP3lfnolZicsG6Qu+//75dffXVJ62Xo8dJbvTo0V5p0syr4P56zVoPJzm9NiCjI+AACUYfqsGHrIKNFro7FX0wNmrUKPqzPkD1galxI/oA11gMjc9Q0NDYlZtvvtmnL2u6sFobarEUKFAgen+1uDSORB/m+iBs3bq1B6lgSnEwVXrKlCkeLDSOR20SBaQg4GiKtYKUPsB1TBcFmFhBEFA7RMd0GwUdUTBSKFN7SkFK1+urgpbOWQFHbawHHnjAg5CmYiuA/BZ6jxXwFAYUAhWqkk+DV1hQO07hTtO79Zx6f4Pz1PMq1Kh9qPdY7SaF1Njfhabna7q7xt2IxtTo/VObS6/x1ltvtcGDB3tI0vOkFDhS06JFC3/OWMnfZ9H4IbX0FAJjW2BaLDC1NiOQ4cW7hAQgdWpLPPnkkyke03Rf/V940aJFSa6PbdPomKaYawZVly5dfPrwoEGDvLWir8FMnVia4aMWkGY9qVV2qqnlaoloJtLcuXNPmrWjWWDBNGZ9n/wStHT0GPqqVk1q1KLRjKP77rvP21aiFovaeGrJPPLIIyfdJ2hR6X0KqA0TtKiS0/VaDC/WlClTvK3Wr18/f516vGBmmx5Xz61Wn6aABzOjYqm9pQX6NEMqdhaV2o163Zr+LmoV6b1avXp1qu9BbItKvzc9lt53PW/spUqVKif9zej90e2/+eab6HV6Ll2nWW7BexM7QwzI6BhkDISM2jQamKrqhNo4qhBo0b6ABg9rWwC1kfSvdy3/r8G/qkYEA2P79OnjM3VU9VELJJh1E3vR9gB6LlWAtOVC0F4KbNmyxasbajGpIqR2l6o2weW9996LDgJW9SZ2KwVVf3ReqoqoQqOqhhYoHD9+vLeFRK9RrRSdiwbcplS5CKpgQWtM1ZPUqMoRW8HRuWvfLQ0sfuaZZ6L7YamSJEOHDvXB2arQqNKl2WjJ93ZSVUuvRVtCxGrSpIm3EGvVquXvvQZJ633X70Xn8VsqOKq+xV7UWoyl2Vaq1mkAtgYRB1QZVLUpqLolp6qSfi9ARkXAARLcwIEDox/QsRd9cCenD2UFG8260ZgLfXhqxo5uH/vhqwCjsT6aPaNZVBovE8zCSk4tJI2FSX6JXbE3JUGrRi0Utbg04yl2E8dgpo7OTbdROyt2BpnGrShwaGyRxposW7bMzjvvPLvjjjt8CrumV2ssi8YFqa2ndpJmRQWzoYIPZ7WP9KGvy/nnn59kCrSCkR5bY3A0vic24OicNP5IbSRRC0zjaXQOGuOkgKB9rHT+uo1ChGZMBUFIrrrqKm9naQaXWomxdH5a6VkzrBRA9R7pOTTdXOEi9nFE5508QCkkxoZGXdRKjB2LpFlbeq6HHnooep1es/4udF1K4200m0urHivYARlWvEtIAFKnxf3UXtCibskvaiEFi+bF0iyfRo0a+Sq7sbRYYLBAXqyvvvoqusdRctr36VQtKi1SdzrUinnzzTf9e60E3Lp1a39tegzNMkqtLRXQ4nZq4bz77rs+o0z30wrBQftJrTjNDNMChsHid8FKxslbVGrVBTTbLHYhQ800S+6tt97y1p5mLKlF9o9//MPba8nfy+HDh/tj6FgwMyuYCRXMeNM5B4sVqr2lmV+xM9K0irJmiWlWVux5B+eu323sLKrY/b/0Pn7++ee+gOGf//znJPfVopDBzDrNFtOMK62mHLu4o34fev9ef/31yE033eSPH7s6NZDREHCABKap1amNwdGYF43xSB5wkm8XENAUcW1r8Fto00V9EKZEH5r6EAw2cPwt9AGqMTex08JPRa8p+DDXlHZtEhp8YAc0tTk2FGh14VdffTXJasX6wE6+arLGIml8i0JHSmFLY3kURLRdwrZt23y8jcJNSlOpp02b5uN1ktMqyQsWLIhOQdd7punvmiaeUrDTVO7kNLU+GGukafvJA86uXbs8XCnopbaB6KhRo/x+2gIj+fkrvAWbrCokpfZ3B2QUmfQ/8a4iAUhMQRsqdmbQ2U4zozTuKKWp2OlF/9nWgn4a+/Nbz0MrM2s8TkpT6oEwIeAAAIDQYZAxAAAIHQIOAAAIHQIOAAAInbN2qwatKaEFyLSAGIPtAADIGDR0WAtial2rUw2yP2sDjsKNFkMDAAAZz+bNm32j39SctQEn2NBOb5BWJwUAABljqQYVKP7bxrRnbcAJ2lIKNwQcAAAylv82vIRBxgAAIHQIOAAAIHQIOAAAIHTiPganU6dONnbs2OjPhQsXtp07d9qsWbOsV69etm7dOrvkkkts+PDhVqdOnejthgwZYsOGDbO9e/f6viovv/yynXfeeWf03H799Vf75Zdfzuhj4v9kzZrVMmfOHO/TAACEUNwDznfffWcvvfSSNW3a9P+dUJYstnLlSmvdurU98cQTdv3119vEiRM9xOi2JUqUsDFjxtizzz5rr7zyipUpU8YGDBhgLVu2tE8//fSMrGmjOfbbtm3z8IS0VaBAAStWrBhrEQEAwrPZpp5a07w+/vhju/TSS6PXd+nSxXbv3m2TJ0+OXtewYUOrXbu2Pffcc1a5cmV78MEHrVu3bn5MC/4o+Lz55pseiE53mpl2SNZuyclnUf34448ebs4991zLlSsXH75p9Ls/dOiQ/fTTTx5yihcvHu9TAgBkAKf6/E6YCs6GDRv8Q65SpUpJrl+0aJFXb2J16NDB21T6QFyzZo21adMmekwh6cYbb7Q5c+acdsA5VVsqCDdqlyHt5MyZ07/qd6r3m3YVACAUg4zVclJ1pFy5cl6BeeSRR+zIkSO2adMmq1ChQpLb6uf169f7sdy5c3tbI6Xjv1cw5kaVG6S94H1mrBMA4EyKawVH/2IfMWKEt6e2b99uvXv39oBz+PDhkwJGwYIF7eDBgykeiz2emqNHj/oltsR1KrSl0gfvMwAgdAGnefPmSX5WFadevXqWI0cODzqx1DZSsFFbI/mx2OOpGTx4sA0cOPAMnj0AAEhUCbUOzoUXXmjHjx/36d7J202aLq5WVunSpX1QsaaSp3Q8Nf369fMBScFFe1ABAIBwilvA0YynK664wmfTBObNm+cDhps0aWLTp09PcvtJkyZZ48aNfTCqZlHFHlfbSj/reGqyZ88e3XfqbNt/SqHxxIkTp3374Lb63WjQ9X+jgeIpTcb7b21AAABC16K65ppr7IEHHvAp4ffee6+tWrXKevToYY8//ri1aNHCLrvsMqtevbrf7u2337YlS5b4+jfy6KOPWp8+fXx6sXYUffrpp32Q8e+dQfXflOn7rqWnjc8mbeH9rzTFXoskPv/88ymGH7X88uTJ4z/fddddXgnTLDZVxRo0aODVLo2V0UBgXZK3Au+++24rW7as/y6XL19urVq18plR+t28/vrr1q5duzPyOgAASPiAU7RoUZs9e7Y9/PDDdtVVV/mUbH2vgcb6MFWoUZBR4NFKxh988IGVLFnS73vPPfd4i0rr4KhKoEUCZ8yYYeeck1Adt7hRC04hRCsFi2adxW4rr4qL6DbLli3zcU/BoG9VYvQ+PvPMM37dsWPHLFu2bB6ERCFGCyzG0nFV1nbt2uVhRtP/ly5d6lUyLdgIAMBZtdBfIi4UpGqGPqBVkdBg54xYwVE1S9UXBRaFHFVdFFp00fdqQT355JO+ArR+1uwzDd5WULnvvvt8LJTC5pYtWzw8rlixwu+j90YtK4Wlb775xhYsWOCVH221ccEFF3il6Msvv/Rg+tRTT3no0YrTQRsxWPcm1qnebwAAMuRCf0gbWvBQYUEXhZabbrrJW1R9+/b1iowGaauqIwpAmj7//vvv2y233GIVK1a0Zs2aRWe1qRWoKo+CjH4OKkE7duzwrTEUWtSOUvDROKeff/7Z1ypSBU6VOO0XFlSBgsoRAJxxA/LH+wzCYcA+CwsCTsjs2bPHtz2IrYaocqM9voJxNoUKFfKwoTCi61TBUcVFlZjPPvvMx86oiiP9+/e3F154wfcD075f+l40Nkotrho1aljXrl19LSOtNr1161Z77733fO+wt956y/cZ02Pq/gAApBcCTsioTfT111979STYrVsVFLWTFE7UktLPajn17NnThg4d6mObrrzyShs1apRvh6G2lNYVUvjR42i8lAZy6366XoO7gx3dy5cvH31uVXu0San2FtO0fNEAZQUjAADSE6NyQ0bT77VXl8bP6FK1alXvUWprCw0AVuBQC0mz1jTORoFF1RcFmSlTpnjFRSFHK0OrjaXbalabWly1atXyYwFtfKoNUEUtKlV+1JqaP3++V4REYUuz4QAASE9UcEImaC2JWkRqQymcaBCw2kSPPfaYfx9LM9DU1lLbSVtmtG3b1iswwZTw1157zW6++Wav+sTuGaXqzbRp06JtMA0UVtVIQUjfa68xTe8fP358Or4DAABQwQmtlStXWufOnb3KIgowWq9G22NonE4sjcnRbuxaR0jVFk3BV2vr888/96CioBSEGA0kFlWCNM7mo48+igYfVYD0OJox1b59e9/xXS0zPTcAAOmJgBNCqpxoHM1tt93m7aWAxtFosLFaUrNmzfLrNK5G4280TkYDihV+gvWERo8ebZ06dfKp3KryxHrjjTesWrVqNnPmTL+9ZklptWmtZ6THVEtMrbKWLVum86sHAIAWVVxWFk5L48aNs4ceesjbTX/729+SbL2gQcPaDkPtJlVyVIHRjKfJkyd7a0trC6xdu9bDjqaNK9RoTM51113nizG+/PLLdvvtt0cHFAdBSNPORe0sLQKojU21arICkwYy6zG1vo6qQQAApAcqOCGi9WzUTlJ7SF/VZhJVVIKxM6qwzJ0712dU3XHHHT64+KKLLvLqTt26df2i41qB+K9//au3r2rWrOkL9un2quiIwo0WWdLA4nfeecc3QdVXje+ZMGGCr8WjgcbaI+wf//iHt790HgAApAcqOCGi8TGaxZR8RWBVWHbv3h39WeHkkUceSXIbLeCnlpb2B1N1RuFG+1IFdH2lSpV8deKAVpLUrCpdr5CkAKPKjQYpB6sWay8rjQfSmB21sQAASA9s1fAbtmrAmcf7DeCMYCXjs2Yl4/2nuVUDLSoAABA6BBwAABA6BBwAABA6BBwAABA6BBwAABA6BBwAABA6BJyQC1YxDnb8PhXtIRV7+9gpeQAAZCQs9JfI6yz8xvUIFE60e7gW7dMKxQMGDLBNmzb5buDHjx/3hQAPHjzo683oZ61Bo+0bAoMGDbIffvjBt2fQSsRaEVmPqVWKtQt579690+BFAgBw5hFwQuSnn36K7tytgBNcJk6c6NcprCj8qJKj9R21F9Xs2bOj99dKw1qpWCsS33PPPb46sbZ7OHTokG/TAABARkHACRGFk127dvmmlworzz33nG3bts23XZACBQrY3r17Pdxo3ypdVLGZOnWqB59vvvnG21TaaPPuu+/2cDRt2jTfl0q7g4vukzVr1uhGmwAAJCICTogodKgNpc0ttSnmeeed57uAB9Sq+uijj6xChQoeXs4//3xvYX366aeWO3du3y9KAWbx4sVetfnxxx/t9ddf932s1MpS+FG4WbdunZUoUSKurxUAgFPhn+Eho2Aybtw4a9y4sT300EP2xRdfRI+pkvP555/7hpjaPFOqVatm/fv3993CW7RoYddcc42NHDnSj2twcdGiRb3qo9CkncY1bodwAwBIdFRwQqZIkSLWtGlTe+KJJ+zRRx+1jh07ettKG5KpaqPA0q5dO9u4caNfX7hwYRszZoxXbxR2RK2tefPm2erVq+2Pf/yjj9nR8TJlysT75QEAcFoIOCHyyy+/eOVGIWbRokV2/fXXW+XKlX3AsFpLmjm1bNkyb2Wp3XTXXXd5uFEYUoVGgUd02ylTptiWLVvshhtusMcff9y+/vprq169erxfIgAAp4WAEyIKJmPHjrULLrjABwZrBlSpUqV8VpSOaYZVjx49rFevXtGBxlKoUCFvPSkAie6jgcoKNJdccomtXLnSBx537tw5zq8QAIDTwxickNHA4TvvvNNbVaq+aOq4wo0GEyvAaH0bhRu1q7Qezp49e7xlpYqNxu+Iqj8NGjTwio7Wwunatavlz5/f6tSpE++XBwDAaSHghEzLli09tGigsL4G07lfffVVa9Wqlbeshg8fHr291sHRGJuFCxf6OB3RfVW9UeDR9PA1a9b44wIAkFHQokrDlYXjQeFEKw8r2GjcTN++ff2rZlBp/I2me1988cXennrkkUe8enPLLbf47Q8cOOCPoe9ffvlle+WVV7ydpccYNWqUbd++3e9TpUqVeL9MAABOiQpOyGiNGw0crlevni/2pzVsNPW7Z8+eXr3RFG9NI+/Xr581adLE76OWlSo2L774ooejjz/+2MqXL28DBw60SZMm2eDBg+3DDz/01lXDhg2jg5EBAEhUVHBCSONlGjVq5NWWYE0b7UsVUKvqk08+iVZsNBh56NChPti4Vq1aHnieeeYZa968ua9+LJpCrpWOv/vuO59aDgBAIssU0afZWUiL2CkI7Nu3Lzr2RDSbaMOGDVa2bFkfhIu0xfsNIENuhhxWA/Zl2M/v5GhRAQCA0CHgpEI7byPt8T4DANICY3CS0cBczSLaunWrrwisn7VmDM4sdUa1AOGOHTv8/db7DADAmULASUYfthoPop20FXKQtrRicjCtHQCAM4WAkwJVE/Shq5V/tQge0oZmb2XJkoUKGQDgjCPgpEIfutriQBcAAJCx0BcAAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChk1ABp0+fPrZ69Wr/ftasWValShXLkSOH1alTx5YuXZrktkOGDLESJUpY7ty5rXXr1rZ9+/Y4nTUAAEg0CRNwFi5caEOHDvXvV65c6aHltttus08//dQaNGhg1157rW3ZssWPjxkzxp599ll74YUXbP78+Xb48GFr2bKlRSKROL8KAACQCDJFEiAV7N+/36pXr24bN260VatW2Ysvvmi7d++2yZMnR2/TsGFDq127tj333HNWuXJle/DBB61bt25+7MCBA17NefPNN+36668/7efMnz+/7du3z/Lly5dmrw0AkA4G5I/3GYTDgH2W6E738zshKjgKK0WKFIn+vGjRImvTpk2S23To0MHmzJljP/30k61ZsybJ8bx589qNN97oxwEAALLE+wSmTp3qlZrly5d7ZUY2bdpkFSpUSHI7/bx+/Xo/pnE3xYoVO+n4l19+merzHD161C+xCRAAAIRTXCs427Zts3vvvdeef/55q1SpUvR6janJlStXktsWLFjQDh48mOKx2OOpGTx4sJe0gkupUqXO8KsBAACJIq4Bp1OnTnb55Zdb165dk1yfM2dOO3LkSJLr9u7d68EmpWOxx1PTr18/79cFl82bN5/BVwIAABJJ3FpUL7/8sr333nt2zjnnWJYs/3caVatWtdKlS3s7qmbNmtHr161bZ+XKlfNjGlS8c+fOJON2guOpyZ49u18AAED4xa2C07RpU/v222/t66+/tq+++sovwZgcHZs+fXqS20+aNMkaN25s5557ro/ViT2utpV+1nEAAIC4VXDKly+f4vUVK1a07t2722WXXeZTx6+55hp7++23bcmSJb7+jTz66KO+KGCBAgV8LM3TTz/tg4xPd4o4AAAIt7jPokqJVjBWqFGQefzxx+2SSy6xDz74wEqWLOnH77nnHm9RaR0czYZSxWfGjBne7gIAAEiIhf7igYX+ACBEWOjvzBjAQn8AAAAJi4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCh4ADAABCJ0u8TwAAgN+rzJGJ8T6FUNho4UEFBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhE5cA862bdusbdu2lj9/fitbtqwNHz48emzp0qVWu3Zty5Ejh1WpUsVmzZqV5L7jxo2zcuXKWc6cOa1Ro0a2du3aOLwCAACQiOIacFq3bm0///yzzZkzx55//nkbMGCAvfHGG7Zjxw5r1qyZ1atXz5YsWWIdO3a0Nm3a2PLly/1+un2XLl2sT58+tnjxYitZsqQ1adLEDh48GM+XAwAAEkSWeFZvypcvby+99JIVKlTI6tSpYwsXLrQpU6bY5s2brUKFCjZs2DC/bY0aNWzFihU2ePBge+utt2zo0KHWvXt3DzkyduxYq1y5sld17r///ni9JAAAcLZXcIoVK2YTJkzwcHPs2DH76KOP7O2337aiRYvaokWLvLoTq0OHDl65iUQiXrVRRSeQOXNma9eunR8HAABIiEHGdevWtQYNGnhQUZtq06ZNXsGJpZ/37t1rO3fu9FZUSsfXr1+f6nMcPXrU9u/fn+QCAADCKSECzvjx4+3vf/+7ZcuWzaZNm2aHDx+2XLlyJblNwYIF/euRI0f8a0rHTzUGR+0tDWYOLqVKlUqT1wIAAM7iMTixqlWr5perr77aZ0ypGhMEmYCqN5I9e3b/quOxIUfHk4eeWP369bMePXpEf1YFh5ADAEA4xS3gfP311x5S1J4KaNBxnjx5fBxO8nbTunXrvPKiY7lz5/bjGr8Te1zTxlOjYBSEIwAAEG5xa1GtXr3aBwYfP348et2aNWts165dPuV7+vTpSW4/adIka9y4sWXKlMmuuOKKJMdPnDhhkydP9uMAAABxCzgtW7a0rFmz+ho3WtRv5syZduONN/rPnTt3tu+++8569+5tX331lU8X1zgdtZmkZ8+evm6OpodrbRxNF9d6OrovAABA3AKOViDW6sS7d+/2io3WtdHU8FdeecXbULNnz7b58+d7C2vUqFFeoalVq5bfV7cfMWKEDRo0yBcDVLtq7ty53roCAADIFNHCMmchDTLWmJ59+/ZZvnz54n06AIDfoUzfd+N9CqGw8dnmFpbP74SYJg4AAHAmEXAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDopEnAOXHiRFo8LAAAQNoFnEcffdSGDRuW6vHHH3/cevbs+b88NAAAQHwCTiQSsf79+1vVqlWtWbNm1rt3b5sxY4YdOXLEtm3bZqNHj7bChQv//rMDAABI64Dz1Vdf/b87nXOO9erVyyZNmmTdunWzPHny2DPPPGOlS5e2xo0bW/369a1fv37/y/kAAAD8bll+y41btmxpmTJlspw5c9p1111nF110ke3fv98yZ85s5513nq1bt87H3xQrVsxvBwAAkPAVHLWhnn/+ea/SvPPOO1a5cmWbOnWq5cqVy+6//3774YcfbMmSJbZgwQKbOHFi2p01AADAmQo4FStWtObNm3vIefjhh23o0KEecM4991ybO3euV3by589vffr0sb59+9rx48d/y8MDAACkf8CZN2+eVatWzZYtW2ajRo2y7Nmze7i55ZZbbMSIEdalSxd76aWXbMKECT4O59ChQ2fmLAEAANIq4FxzzTXeelKVJkeOHLZ582b75z//aVmyZLELLrjAatasaYMHD/avuj5fvny/5eEBAADSP+AUL17cWrVqZU2aNLFVq1bZn/70J6tSpYr9+c9/9nE4quBosPGUKVP8OAAAQEIHHK19s3DhQm9Tffzxx161qVOnjv98+PBhW7lypd17771e3XnkkUe8ZQUAAJDQAUfTvnfv3u1jcFasWGGlSpWydu3a+YDiJ554wkqWLGmXX365tWjRwltZmnEFAACQ8Ovg1KtXzzZt2uQL+2mtGwUcDTgeP368FSlSxDp16uS3O3DggOXNm9eOHj3qA5EBAADSU6aIek+/w86dO30wcbZs2aLXHTt2zLdtSORBxlqgUO20ffv2JfR5AgD+uzJ93433KYTCxmebW6I73c/v372buCo3seFGrrrqKpszZ87vfWgAAID/ye8OOMm9+eabtnTpUvvxxx/P9EMDAACkf8DR+JwHH3zQrr76anvggQfO5EMDAACkf8D5/vvvrWHDhr4OzrRp087UwwIAAKR/wNEg40GDBln16tV9802tlaPBPwAAABlimnjVqlWtbt26vubNiRMn7JNPPrE1a9bYlVde6ZtuNm3aNO3OFAAA4EwHHK1pc/DgQfvss89swYIFtnXrVr9Oi/s1aNDAQw4AAECGalFpwb4NGzbYd999Z+vWrfOdwrU9w80332x/+ctfvD31zjvvpO3ZAgAApOUYHG3dcOGFF/q+U9q64Y477vCVjbt16/a/PiQAAED6j8FJjRb6e/rpp61+/frWpk0bb12NGTPmTDw0AABAfNfBue666+ztt9+2sWPH2oQJE87kQwMAAMRvJWOFnC5dutjixYvP9EMDAACkX4sqOQ06zpo1a1o8NAAAQHwCTu7cudPiYQEAAOLTogIAAIg3Ag4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAidNNlNHGfQgPzxPoPwGLAv3mcAAEgnVHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoxDXg7N271+644w7Lly+flSxZ0nr16mVHjx71Y0uXLrXatWtbjhw5rEqVKjZr1qwk9x03bpyVK1fOcubMaY0aNbK1a9fG6VUAAIBEE9eAc/vtt9v3339vM2fOtNGjR9uMGTPs4Ycfth07dlizZs2sXr16tmTJEuvYsaO1adPGli9f7vebM2eOdenSxfr06WOLFy/2cNSkSRM7ePBgPF8OAAA429fBWbFihQeVjRs3WrFixfy64sWLW506dfznChUq2LBhw/z6GjVq+O0HDx5sb731lg0dOtS6d+/uIUfGjh1rlStX9qrO/fffH6+XBAAAzvYKzq5du+zWW2+NhhupVKmSHT9+3BYsWGCtW7dOcvsOHTp4IIpEIl61UUUnkDlzZmvXrp0fBwAAiFvAueqqq+z1119Pcp1aVaribNu2zSs4sfSzxuzs3LnTW1EpHV+/fn26nDsAAEhsCTOLasuWLfbggw9a//797dixY5YrV64kxwsWLOhfjxw54l9TOn6qMTgavLx///4kFwAAEE4JEXB+/vlna9GihQ8qvu+++3xmVBBkAqreSPbs2f1rSseTh55YGr+TP3/+6KVUqVJp8loAAED8xT3g/Prrr/bHP/7RTpw4YRMmTLBMmTJZ6dKlT2o3rVu3zoNJ0aJFLXfu3Cke17Tx1PTr18/27dsXvWzevDnNXhMAADjLA063bt3sm2++sXfffdfy5s0bHZ8zffr0JLebNGmSNW7c2APQFVdckeS4wtHkyZP9eGpU+dF6O7EXAAAQTnGbJi5DhgyxMWPG2BtvvOFtqtWrV/v1t912mz333HPWu3dva9++vS1cuNDGjx9vixYt8uM9e/a0Vq1aWdmyZe3SSy+1kSNH+v21Xg4AAEDcAo7Ws1HbSNO+NcU71tSpU2327Nm+ps1LL73kM6RUoalVq5Yf16J+I0aMsIEDB9qPP/5o9evXt7lz53rrCgAAIFNECeMspFlUGtOj8TgJ3a4akD/eZxAeA/bF+wwApJEyfd+N9ymEwsZnm1tYPr/jPgYHAADgTCPgAACA0CHgAACA0CHgAACA0CHgAACA0InrOjgAMiBm9p05zOwD0gwVHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDpxDziNGze2TJky2a233prk+qVLl1rt2rUtR44cVqVKFZs1a1aS4+PGjbNy5cpZzpw5rVGjRrZ27dp0PnMAAJCo4hpwjh8/bsOHD7du3boluX7Hjh3WrFkzq1evni1ZssQ6duxobdq0seXLl/vxOXPmWJcuXaxPnz62ePFiK1mypDVp0sQOHjwYp1cCAAASSVwDTpYsWezCCy+0IkWKJLl+zJgxVqFCBRs2bJjVqFHDHn30Ua/wDB482I8PHTrUunfv7iGnZs2aNnbsWH8sVXUAAADi3qJKyaJFi6x169ZJruvQoYNXbiKRiFdtVNEJZM6c2dq1a+fHAQAAEjLgbNq0ySs4sfTz3r17befOnd6KSun4+vXrU33Mo0eP2v79+5NcAABAOCVkwDl8+LDlypUryXUFCxb0r0eOHPGvKR0/1Rgctbfy588fvZQqVSpNzh0AAMRfQgYczYwKgkxA1RvJnj27f03pePLQE6tfv362b9++6GXz5s1pcu4AACD+slgCKl269EntpnXr1nnlpWjRopY7d24/XqhQoSTHNW08NQpGQTgC8L8rc2RivE8hNDbG+wSAEEvICs5VV11l06dPT3LdpEmTomvmXHHFFUmOnzhxwiZPnuzHAQAAErKCc/fdd9tzzz1nvXv3tvbt29vChQtt/PjxPrtKevbsaa1atbKyZcvapZdeaiNHjrSff/7Z18sBAABIyAqO2lCzZ8+2+fPnW926dW3UqFFeoalVq5Yf16J+I0aMsEGDBvligGpXzZ0711tXAAAACVHBGTBgwEnX1alTx5YtW5bqfe666y6/AAAAZIgKDgAAwO9BwAEAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAEAAKGTEAv9IXVsbHjmsLEhAJw9qOAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQybABZ/v27da6dWvLnTu3lShRwoYMGRLvUwIAAAkii2VQbdu2taxZs9q8efNs69at1qVLF8uTJ49169Yt3qcGAADiLEMGnCVLltgXX3xhmzZtsiJFivh1Bw4csL59+9p9991n55yTYQtTAADgDMiQSWDRokV21VVXRcNNUNHZs2ePffvtt3E9NwAAEH8ZsoKjyk2FChWSXKexOMWLF7f169db9erVT7rP0aNH/RLYt2+ff92/f78lshNHD8X7FEIj0X/XGQV/k2cOf5NnDn+XZ8/f5P7//xwjkUj4As7hw4etcOHCJ11fsGBBO3jwYIr3GTx4sA0cOPCk60uVKpUm54jEk/+v8T4DICn+JpFo8megv0kNTcmfP3+4Ak7OnDntyJEjJ12/d+9ey5UrV4r36devn/Xo0SP684kTJ2z37t0elDJlypSm5xtmStIKiZs3b7Z8+fLF+3QAx98lEg1/k2eOKjcKN+eff/4pb5chA07p0qVt4cKFSa47dOiQ/fjjj1auXLkU75M9e3a/xCpQoECanufZRP+H5f+0SDT8XSLR8Dd5ZpyqcpOhBxlrgPFHH30UHUcj06dP9z+aSy65JK7nBgAA4i9DBpzLL7/cLr30Urv11lvts88+s3/961/2yCOPWP/+/ZkiDgAAMmbAkSlTpnjLqWHDhr72zcMPP2zdu3eP92mddfQ7ePLJJ09q/wHxxN8lEg1/k+kvU+S/zbMCAADIYDJsBQcAACA1BBwAABA6BBwAABA6BBwAABA6BBwAGdovv/xiW7ZsSXFfGm3dojWygHjSqvmLFy+2bdu2xftUzioEHAAZ0rFjx6xbt26+0a5WN9cy+KNGjUpym//85z920003xe0ccfaZOHGitW7d2rp27eqr648cOdJKlChhrVq18q+dO3e248ePx/s0zwoEHAAZkhb2nD9/vi1dutT3pXn++ed9nZE2bdqkuukukJZGjBhhDz30kF1wwQU2a9Ysq1+/vg0bNsxWrFhhO3bssFWrVnklZ9CgQfE+1bMC6+AAyJBUtRk9erQ1a9Ysep1aVS1atPANdN99913f4PDiiy+2X3/9Na7nirND+fLlPWirWqPthLQQ7WuvvWZ33XVX9DYzZ870ELRu3bq4nuvZIENuton4aN++/W8q0wJpSXvR5c2bN8l1agF8/PHHdvPNN1vdunVt6NChcTs/nH127drlrVL5wx/+4FsK1a5dO8ltzjvvPB+Tg7RHBQenTft8PfDAA1a4cOH/elu1CoC0dMcdd9j69ett6tSpVrRo0STHTpw4Yffff7/9/e9/9++p4CA9tG3b1g4dOuT/wCtQoMBJx/W3qGrOnj17bMaMGXE5x7MJAQenrUmTJt5jrlSpUrxPBfAKTocOHWzevHm+8a5aAckNGTLEHn/8cQZ1Il1s377dBxhXqFDBxo0bl+SYxoq1a9fOsmXL5uNzypYtG7fzPFsQcPCb6F/CmTNnjvdpAFErV6701sCVV16Z4vHJkyd7ywpIz6CjVlSsb775xlavXm0tW7a0HDlyxO3cziYEHAAAEDpMEwcAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAGQIWmlWM1MOR2ffvqpHT58OMl1U6ZM8Vktqfnggw/ss88+i65fkpyu00afABITs6gApLsjR474wpHaM0qruv7000++MeYPP/xgX3zxhW+5oH2msmTJEg0Tun2sG2+80XcQT75buK7TVg0BPeaFF17o6+RoHRI9lh5fWzgoJNWsWdPXydFU89ipvQ0aNPA1S95//30799xzfb8rLZGgx9ZyCUePHvX1TGK3igCQOAg4ANJdnTp1bNmyZb7VglYhVljImTOnh4UiRYpYnjx5fD+fcuXK+e07duxob775pt8mCDoKSQomuq0odGgVWe0oHuz9owpL48aNfZE13Vf/ubvkkkt8Of133nnHsmfP7o+h5//jH/9ob7zxht9v7dq1VqNGDd8cUee3ceNGu/vuu/1xFZK++uoru+6662znzp3+GAASDy0qAOlGIUQrEKs6o2Ch77///nvr1q2bNWrUyHdefuKJJ6xHjx5WsmRJX9JeFEz69Onj1Z633nrLqy5bt271kKOQoYtuG7vYn6pDCiP58uXz6ot2eB4/frzvT7V3715/7gkTJljlypXt2LFjSVZCfuGFF/z5FIS0gvfy5cs9IClM6bJw4UK79tprCTdAAmOzTQDpRpWRiy66yKswCgdqAanto6Ci8KPwoipM7PgWhZs2bdpY/vz5vZJy5513etBRiymg26hyo0pPtWrV/LpcuXJZ165drVatWl6heemll3wH8i5dunilKGvWrB6UOnfu7AFH5yJLlizxttSKFSu8haXWmW6vx9djynvvvef3U0jTc+uxACQYtagAID38+uuvkX379vlXadq0aWTevHmRwYMHR7p06eLfV69ePXL48GG/bu/evdH7TpkyJXLeeedFHnvsschrr73m1x04cCBy9dVXR+69995Un1OPo//UZc6cOZI3b95IlixZIrly5YrkzJnTr9dl1apV0ds/+eSTkQIFCkQqVqwYOeeccyILFy7062vWrBn58MMPI3v27Ilky5bNL7rvhg0b0vAdA/C/okUFIN2ocjN27FjLnTu377asVlVQFRF9r+pM/fr1fT8ffS+qkqxZs8Z3B1c7S9UUjZmpW7euNW3a1H9OidpQ2lVc7SpVW/bv329XX321V3OCSpHaVarsqI0ljz32mLe7unfvbg8++GC07aWB0Bp7o/NQxWnatGlWpkwZvwBIPLSoAKQrBYeHHnrIv9cA4E2bNkWPaYZSoUKF7OOPP/YNCYOAo+v79evnYUTBQkFJ7S6FDk0V14aal112mRUrVszHygQKFy7srS/NxtL1aofpZ83U0hgbhRqFJ4Wf5s2b28yZM71V9cknn9jo0aN9mnjQKlPgUkDSQGQNklY4030AJCYqOADSldaXue2223ysjGZSvfvuuzZy5EgPKQoa27ZtsxIlSng4mTRpkq9f88ADD3ioKF68uE2dOtUGDhzoA3/XrVtn9erVsxdffNHKly/vFSDNdgpozI6qNwopCieqzARfg/Ciiyo9mqUlqtA0bNjQg9f555/v3//73//2qeKtW7f2qeEas8Mu5UBiY5o4gHS1aNEin6LdsmVLDyU33HCDBwYNOlbrRwvwVa1aNcl9RowY4cfV3tqwYcNJa+LoP2MKG2ptVahQwad4y/Dhw71ipJZYcrqP2lQaIKyQo8dVu0ltK1WQypYt6yFLoUktrcWLF/sgaLW4FHYUek53oUEA6Y+AAyAutLpw+/bt7dZbb/Wp3Kq2qKqj1pFmP2mWkqZwx1JLq23btn48ltanUTgZMGBAkusVnNTe0iwqBZ1evXpFZ19penrFihU96Oh2Cjq6rao7mnauSpLaYJoW/swzz1jfvn39XBXONKvr9ddf9xldABITLSoA6UpVD03n1uBdBRWFh4AGEI8bN84mTpzoqw8XLFjQ21D/K42nUXBRSFGIeeqpp6LjabRNQ7DisW4XfK8WWqVKlaxTp042d+5cb0Up9Nx0000+NkftMq1y/PDDD9uYMWO8CgQg8TDIGEC6UrtJa80oPKjdIwofwX5Pqo5o3RmFHA0GVhsroNupwpO8UqMxNKrOnIoGDQcDkHVbLfJ3/fXXn3Q7XadqkhYG/O6773zw8quvvurVG42/GTx4sC9E+PLLL3tVSOOGkleUAMQfLSoA6UoVD1VMgn2mRIFB+0NpUPGpqIWlKk9KLSotBKixMqdDA401jifY5uFUNBhZ09EXLFjge1VpocLA+vXrvTWWfEwQgPgj4ADIMNQqUhVGU8gB4FQIOAAAIHSoqwIAgNAh4AAAgNAh4AAAgNAh4AAAgNAh4AAAgNAh4AAAgNAh4AAAAAub/w9L0u2bIDPjMAAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "fig.set(alpha=0.2)  # 设定图表颜色alpha参数\n",
    "\n",
    "Survived_0 = data_train.Pclass[data_train.Survived == 0].value_counts()\n",
    "Survived_1 = data_train.Pclass[data_train.Survived == 1].value_counts()\n",
    "df=pd.DataFrame({u'未获救':Survived_0,u'获救':Survived_1})\n",
    "df.plot(kind='bar', stacked=True)\n",
    "plt.title(u\"各乘客等级的获救情况\")\n",
    "plt.xlabel(u\"乘客等级\") \n",
    "plt.ylabel(u\"人数\") \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "613662C50043454394E17B230BE8EA05",
    "mdEditEnable": false
   },
   "source": [
    "明显等级为1的乘客，获救的概率高很多。恩，这个一定是影响最后获救结果的一个特征。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4A247574DA87475C872ABC0755B32F90",
    "mdEditEnable": false
   },
   "source": [
    "#### 看看各性别的获救情况"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false,
    "id": "CF47A03A505D43C6B34753EF10403A4E",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAHACAYAAACvXBIfAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAANbdJREFUeJzt3Qm8zfW+//HPNmS2zVEoQ6k4IRFXSYbmEQ2kbtNJkU5FhsPpUJ1EqtugTl3lSpOUipKbUpEMDTSQFMkswjaTrP/j/bn/33qste2Nyt5r7+9+PR+Pdbf1+6219m/9HP3e9/P9fL+/tFgsFjMAAICAFEr1AQAAABxqBBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOEDAhg8fbk888cQ+27dt22b9+vWziRMnZvveXbt2+euyM2HCBPvv//5vOxRrhW7cuNGee+45mz9//u9632+//WbLli2zSZMm2SOPPLLf127atMl/x+LFi+Pf7+qrr7bPP//8oH7XSy+9ZIMGDbJff/11n31Dhgyxp59+2jZs2JDle3fs2OGPrHz11Vd2/fXX25o1a+xgZWRk2M6dO7Pct337dtuyZctBfxYQqjRWMgbC1ahRIzv55JNt5MiRSdv1z75OnTp24okn2htvvJHle7/77js77rjjrG/fvnb//ffHg0j58uX9z127drWffvrJpk+fnuX7161bZ/fee68VK1bMChUqFP+9u3fvtm7duvlnR5YvX241a9a0yZMn21lnnRXfftNNN9nChQv9M2TPnj0eFBRWfv75Zw8Uif8J++CDD6x169bx368gULp0af/9q1ev9vPx8ccfW7169fw1VapUsalTp1r9+vVt7969/tkKTTo3mfXu3dvmzJlj06ZN8+9+zDHH2D//+c/45xx77LF+LtLS0vZ57/fff+/nWoGzR48etmrVKj/u4sWL26xZs+z888/3c6v3KlTqUatWLStSpIgHsjvuuMMeffRRO+qoo/zzrr32Wvuf//kff310bkXfQZ9711132eDBg7P8ewEKiiKpPgAAOadMmTJWvXp1/7P+P34FhBIlSljhwoU9uJQtW9b3RRd3KVWqVNLPCy+8MP55AwcOtLfeesurHrr46nOyowutLsrnnnuuVatWzbcp3IwZM8Y6dOhgRx99tI0fP96DTrQ/CjKRK664wj777DPfruM57LDD7KqrrvKL90UXXWTp6ekeuKLvkVhdGTt2rPXs2XOf4zr11FOTnp9xxhlJz08//XT78MMP93mfvmsUJho2bGhPPfWUn8MlS5Z4mHrmmWeyDDfRudT5b9q0qT/v3r27vfnmm/6ZOo9yxBFH+N9P9B0UyKpWrep/N++++65/F1XNomPR91AFLTMFPJ0noKAj4AABUlhRkClatGh8OEbDUZdddpk/14U6uhgrMKhqIapIaBhGoguvXhtVbnQxb9WqlVWqVCmpIqOLt6oOhx9+eDwERO+/884741UVVV4UcLRPF+Err7zSXnjhBb+4Z0XvUwWqZMmS/rk//PCDbz/nnHM8ZIh+99q1a70aUq5cufh7u3TpYhdffLFXcHQeVO1Rleibb76JV3CiCsoJJ5zgwULnTd8n8vDDD/vnK2B98cUXtnLlSh8K27p1qwc/DQe99tprHtbOO++8+Pt0LnTM0TmOzskvv/ziFaB///vfHv4qVKjgv799+/a2fv16/3vQ8JL+vlQVElWKFKQU6vS7Onbs6J+n0JRYBYvo3CZWdYCCin8FQIAuueQSv9CpEnHPPff4hVwXXf1//j/++KMPe2joR4FBD22Lhp407NSyZUu/GIuGjXSR1XDPokWL7Pnnn/cL9+jRo/1irYu/KikKKRo2imRXzYj26SKs9x6o2qDwoOPWd9Awj9539tlnezjQRV4PVYBeffXVpPdpv6pXCiF6TcWKFf1cKOTo3Ojx/vvv2/HHH+9hRFUgVUy0P3HIS8FQD50nnRP9WcNa77zzjocRDRUtXbrUj1HfSw+Fqqh/SSFEPUKiMKOAos/VcJPOeVQhU6DR8yOPPNKrO4khRQGnRo0a9ve//90rOgcKMPs790BBQQUHCNDQoUP958033+x9J507d/a+jygwKKyox+OWW26xu+++27frAq+woSrLJ598Eg8eZ555pl/ER40a5RUc9aHoYvu3v/3Ne2deeeUVr3zogq4QEYmGWjR089577/mfVZkQDcUciEJDNDyjAKUQomqTqkQPPvhg0mujysu3337r4ULHJ2+//bZ/d4WymTNnetXkP/7jP6xNmza+X4FHwUVBUFUhhST140Q0JKRjVRhSU7YCir5LtG3cuHEe6r788ks/PlVxFL6uu+46f67zrF6mKDxecMEF/vegcKNhLZ2vjz76yIfJdN71mTqPmzdv9jAVnU9V4xQo69at6+EmqrhlR+cNKOgIOECAouEbXWRVGTjttNP8oqchG1HlRcNO6o9JDBQyZcoU/6mgILqgKjholpDCQdRHoouuQlDU45OZhq/UjKvXafgl2qaQosByINdcc41f/BOHu3RhVxBTcInoe2m7Ao6Ch4LXf/3Xf/k+BY777rvPK1oaptKQUxRuRFUhBRoFDv35xRdftH/9619JzcEakovOSfQd1KfTvHlzD0C33nqrV5ZEoeQvf/lL/PxH5zSxKvTYY4/5OYno/Oqhvyt9jyicqNKTWJVK7BXSMSgIqQqXmc4BAQcg4AAFggKKhoM0pKLG45dfftkrOxqS0UVYF0z1mqjioKZeBZfEYQ5dkBUcFE5UmYl6eyLapqGgxOCiSov6bfSZmhmlmVOa/XOwFB4UbPS79RnqxdFwzkMPPZT0uujYFQ70HRMv7urNUThRb07UoDxv3jw/fgUKHbMC04oVK7xpWCFHoSHqbdHQnAKQhobUpyMjRozwXpy2bdv6DDSFLU3zVrVFAUezqSKqGuk7RE3UomNU5Uu/N7FPSnTsqnLpuLKi6o+GAvVdVWXT8Jo+T8Njeo++o4JeYh8RUGBpmjiAMLVt2zZ29NFHx8qXLx+76667NJ86VrhwYf9ZrFgxfxQqVCiWlpbmP4866qjYzp07Y2vWrImtXr3aXzdz5sz459WrV8+3FS9ePFakSBF/j/6sbRdddFHS7542bVps9uzZsSlTpvj+hx56KPbpp5/GPv744/hn6vePGzcu/rs++OCDfb7Dli1bYq1atfL9ZcuWjaWnp8cfpUuX9t+v47/99tuT3vf555/7/muuuSaWkZERq127dmz48OF+TvRZ0fEfdthhsTJlyvhPbb/++uuTPmf+/PmxVatWxfr27Rs7/fTT/dzMnTvX923fvj12+OGH+/eZOnWqn9utW7fG36s/L1myJP799P1l7Nixfsw6hujvQY+iRYv6695///19zoN+p76DPmvHjh2x3bt3+2ufe+652I8//hj/u/rtt9/87xAo6KjgAIFR/8aTTz7pwxuqxpxyyik+U6lJkya+Por6VzQ9W/0kou16aAhKVQBVBFR9yWrhOU1LVtVBw1cDBgzwyod6Q/Q7M08Zv/zyy73yElWCNENLQzH6HS1atIgPP+2Pmp8vvfRSr7qIKiTRcFUiVaNUmUp00kkneTOwhtQ0NKXqh5p99d2j4bXatWsnNQGrEpV5AT31LKmCo+qOqA8p6iHS5zRu3NhnWKkSpmGraHq96M/qdcp8LvW7VXnKvPifvp+qalk1EatXSudSzdN6v865ZP7eUfM2UNAxiwoIjIKGLoYNGjTwXhA1vap3Rhc9BY5hw4Z5w6xoVpQuzlrRVzRssz/q5dFQkUKBLqy6wOvPzZo185CUSIFCfSIaphKtOKxGWvX9RD0oGmrJqmFWwzQPPPCABxf1CkVNxQob6vnJ/FiwYEGWM4cUbrQCsfpq9HmaPaZhNAUEDddpxtjs2bP9uETDRYmBIerBUWNx1CCtlYfbtWvnw3zRdHR9RzVhq+fo99DQWeIjqwUGRcf4+uuve9CKmr917KJZVwD2RcABAqOqgfpIdMGN1quJfP31116J0OwqXfD/8Y9/eAjKvPjdoaReEfWNRGvUKCjoGKKFB6NwkUhVGlVv1LyrPpZoBV9VSVQ1yvyI+mMSqZ9FVab//M//9Fs0KBxFVQ9RJUfHoSnYOg+ZaZaTqkDqn1G1K6qKqJdI7+nVq5dXo7RooUKiqlqaPfV76LsnPqLbSGT+HprtpuNIDFA6rzpP6sMBsC+GqIAARdOkM9PsHw2xaNq3lv9XheXGG2/0ptScWP1WFaNnn302vsBgFHCiKoqabRV8tHBe5iqUmnkVgBKbcDVNOiuqFkUzxCJaCVjDdJoZpaqRFterXLmy71PVSMNWmkKuUKKhKp2bTp06xd+v1+o8acXkzMNiWgxRwUmixml9l8cff9xuv/32g16HJnFhwuzoODXUqPtdRX9HCj1qitZCiApW0Sy1iO7ppenpCl9AQUXAAQKmnpLEi636PlSN0Iq4ChqaVaQeGvWqaPqyhrMi0Wwk/dQFVEFFwUTDS+rziG5RoOqIekkUZrRAoB6ioSWt+6IQpQpERJUKTZvWZ0RTzBMrKxEdd+bZWlqzJrsenMw0dKOKVRQKNCtLs6JElR3NLNM6Nqpy6bl6hlTJ0crLUR+NpmrreyhAKWQkrkys4T0FJ62QrGn3M2bM8KqO1tzR99XnRZ8TncvMt/7LHOx0PjWtP/Hv7NNPP/VKk4JoRL1Uc+fO9b+37Ia0NLNLfU4aYgMKpFR3OQM49F566aXY5Zdf7rNynnjiidiyZctil1xySaxq1ao+26Zp06axWbNm+WsXLFgQO/nkk337nXfeGf8Mzf7RNs2C+vbbb/3PemjmUYkSJXxGkx4lS5aMz8x6+OGH/b2fffZZrHLlyj5T6Nlnn/Vtw4YNix1zzDH+ultvvdW3bdy4MXbZZZfFmjdv7ts/+eSTLL+PZlppv2Z5ZfXQ9xwwYECW77333ntjLVq08Pe/8MILsTvuuMP/rOOJaOaRzpe2N2jQwJ8nzl6KvnufPn18Vtdpp53mz1u2bBn74Ycf4q/V52vG2kknneSznCLRLKfp06fHZ1Fl/s+vZmo9//zzvn3GjBnx7b/++mts9OjR/udt27bFbrzxRn9Nv3794q/R36+23XzzzX6+69ev78ehWV5AQUXAAQKki7KCx7nnnhvbtGmTb9MF/Morr8xyCrKmFd90003xC7B89dVXftHUxVj7dRHN7oK5d+/e2IYNG+K/S3r16hV79dVXkz5Px9SmTZvYokWL4tuvuuoq/z2tW7fO9vNHjRrlr9HFPisNGzb0i3tWBg4c6PvHjBnj56VSpUqxQYMG7fO6PXv2xO677z4Ph5m/m6Ziz5s3L75NgU+hS/sy0zRufddE0bl88803/blCSOaAo32a8q7wqSCTlc6dO3tovOeee5K279q1K9akSROfdq791atXj73++utZfgZQUKTp/6S6igSg4NLsKg0jaXp0djT8peGb7Hpwfg8NkR1M70tOUg+Qhg+jqecHS03NGrJSvxCA/SPgAACA4DBNHAAABIeAAwAAgkPAAQAAwSmw6+BoXQqtbaFl2Q92US4AAJBaah3WrVa0QnpW922zgh5wFG6yW+0VAADkbVo9PFosNCsFNuBEN9TTCSpbtmyqDwcAABwErXyuAkXijXGzUmADTjQspXBDwAEAIH85UHsJTcYAACA4BBwAABAcAg4AAAhOge3BAQAgFfcgw/4VLVrUChcubH8WAQcAgBxet2XNmjV+o1ccHN0QVzfg/TPr1BFwAADIQVG4qVKlipUsWZLFZQ8QBrdv324///yzP69WrZr9UQQcAABycFgqCjcVK1ZM9eHkCyVKlPCfCjk6b390uIomYwAAckjUc6PKDQ5edL7+TM8SAQcAgBzGsFTuny8CDgAACA4BBwAABIeAAwAAsm2S3rNnj+3Pzp07D/qz9u7da7mFWVQAAOSyo/u9nau/b+n95/2h982aNctatWpllStXtmLFivk0bs0KK1++vAcWhRvd3XvXrl3xvpnHHnvMhg4dat9++63f8VuzofT++fPnW8OGDe2XX36xChUqWE6jggMAALKkgFKjRg1fy+enn36yadOmWaFChfzPK1assM8++8yndSc2BWsGVIMGDfy90rt3b7vkkkvssMMO8+e5EW6EgAMAALKkMPN7ZzxF69YoBMmPP/7oVaD169f782XLltnixYtt0aJFlpMYoiqIBqWn+giQmwZlpPoIAORjK1assKOPPtr/rH4cDUklPk983fnnn2+33XabPz/33HPtwQcftG+++cbmzp3rw1ty3HHH2e7du61Ro0ZeAcopBBwAAJCt6tWr29KlS/3P+qlgkvm5ZGRkePiJDBw40KZOnWqlS5f2qs13331nxx9/vN+KQQ7UvPxnEXAAAECW1Eh8IFFl5ssvv0xasblz5852xhln2BVXXBF/TUQrFOuzixTJuRhCDw4AAMjS1q1bbeXKlR5cypYt68NLO3bs8FlU0XPNoJJLL73Upk+fnvR+3RG8S5cu3pdz4okn+jY1G+vx+OOPW06iggMAALKk2VDbtm2Lz4A6/fTT7cwzz7QBAwbEh5m2bNniPTV6jYJPZu+9956VK1fOli9fbieccIIHpI0bNx5UA/OfQQUHAABkqU6dOr7+jcKIfqpCM3jwYCtevLgPLxUtWtSnfX/++ef77eFRH04020rVnEqVKuX4dHECDgAAyJIagzVMpQX9XnjhBatZs6b/WQ9Vb1SNUXCpVq2a5TUMUQEAkMv+6MrCua1ixYrxYaa//vWv9vTTTycNLSkAqYqjKk1W5s2bZ6VKlfKKz9q1a32bZlSpb0fDVI0bN/b35wQCDgAA2IduqfDyyy/b+PHjfQhKTcFqJBbdhuFf//qXB5hmzZolzYZSP44e0q5dO/8cVXn0GvXp1K5dOz47a926dT5clRMIOAAAYB9qGFaQadu2rY0bNy6pZ0br2SisaGG/bt26Jb1PQ1q6X5VogT/dh0oVnEQKOFrZOKum5EMlLZZ5cnoBocWI0tPTfWEiTXUrUFjJuGBhJWMgZdSrolsV1KpVa5+LPP7YeTvY6zdNxgAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACAHKe1b/bu3Wu5hYX+AAAIfT2yQYdmPazrr7/eLrjgArv44ouzfc1jjz1mQ4cO9UUCy5QpYz///LMv9jd//nxr2LChr2yc0zfaFCo4AADggJYuXWrPPvusB5b9KVmypDVo0MDDjfTu3dsuueQSX/lYciPcCAEHAAAckG64qXBy1VVX7fd1hQsX9p8//fST/9SKxK1atfJbM0Q321y8eLEtWrTIchJDVAAAYJ9bJehO4arGFCtWzO8grptuXnTRRT7EFPn1119t+/bttm3bNjviiCP83lS33Xab7zv33HPtwQcftG+++cbvSRXdGeq4447zm3E2atTIPvvsM8spBBwAAJBkxYoVHkBEQ0sKOAo9pUqVsldeecXDioKN7hOlkKMGYt1ZXPeJigwcONCmTp1qpUuX9qqNApNu0qn3yZ49eywnMUQFAACS1K5d23bt2uVBRj8HDx5s55xzjt8pXI/Jkydb0aJFbcOGDR5UFFrURKyKT6Rz5852xx132Jtvvhmv3kQUigg4AAAgVxUqVCjeFKzhJM2M6t69e3z/G2+8Yc2bN7cSJUr4c/289NJLbfr06UmfU7VqVevSpYv35Zx44om+TZ+rx+OPP56j34EhKgAAkC0NSa1atcqrLqJqzZgxY+zee+9Nep0qOuXLl8+yOblcuXK2fPlyO+GEE2zHjh22ceNGD1FBV3Cuu+46S0tLiz8qVark2ydNmmT169f38b1mzZrZnDlzkt6nOfZHHnmkjwd26NDB1q5dm6JvAABAuK688kobO3as/eMf/7DGjRvbNddc44FF1++DUb16de/D0TVeVM3RtT6np4unPOCou/qRRx7xBYH0mDVrli1YsMBDi06qnrdu3drOOussW7lypb9n5MiRdv/999tDDz3kDUxKgxdeeOE+Y3wAAODPUTDp1KmTNxGrQjNu3DivvkybNs3yspQOUSmQKMxofrymjUW6devmKyX+/e9/9+fq5P700089CA0bNsweeOABL41dfvnl8fKZqjnvvPOOT0sDACBPO0QrC+eG3377zaZMmWL33HOPLVmyxBuM3333XTvzzDO98VhTwY899tgs36tQpJEWjcZEIy2aUaXGZQ1TqSKkoa3gKjha/EdjeZlPjJqUOnbsmLSta9eufoK1gqIWB0rcr9USNTdf+wEAwJ/3wQcf2NVXX+0FhMsuu8xatmxpCxcu9BEVhZqZM2f6yEvTpk3ji/hFTcl6SLt27fwaf9RRR1n79u29uVgztLTtlFNOsYyMjDArOBqeUulLX1ZjcjqBQ4YM8XRXt27dpNfquZKj9ikNqjM78/4vvvgi29+ltKhHJHGuPgAASFanTh1vAbnvvvt8hlR064WIgo2uu1qsL+qfFU0j37Rpk/9ZC/zpPlSq4GSuCikUZdWUHETAUagZMWKED0GpdNWnTx9fSEgnNHEuvegkaKXErPYl7s+OgpPm8QMAgAOrWbOm99vsT9myZa1NmzZJ27T2jR5So0aNbK//hx9+uOWklAac8847L+m5ymAtWrTwpKegk0hpUMFGc+0z70vcn53+/fvHT3hUwcnuxAMAgPwt5bOoEqnRWCsbKtVpOCqRbsyloSwlyi1btiSN9yXuz47upaGkmfgAACA37N27N9WHUODOV8oqOC+//LKvYqiG4mhu/Pvvv+9jfGpEmjBhgk9Li2gOvpqVqlSpYvXq1fP90Rx8DVvpuRYeAgAgr4ju46SF8tSLoufRNQ9Zz65Wg/K6deuSVlPOVwGnbdu2dsstt/iU8BtvvNE7sTWENGDAAL8b6cknn2wNGzb017322mvera31b6R3797Wt29fX2hIw0yaMq4mY6aIAwDyEl2ka9WqZatXr/aQg4OjlhON2PyZ1Y5TFnCUZDWXXrdV1zo4FStW9D+r0VjpVqFGQUaBR/ev0Jx7rYYoN9xwgw9R9ejRw3tpNBd/4sSJOb7sMwAAv5eqELpYqwVDs4ewf2pALlKkyJ+udKXFCujyvwpG6enpPge/wPXjDEpP9REgN+WjBcUA4FBdvyl5AACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASnSKoPALnv6J0vpvoQkIuWpvoAACAFqOAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACE6eCjh9+/a1hQsX+p8nTZpk9evXt+LFi1uzZs1szpw5Sa8dOnSoHXnkkVaqVCnr0KGDrV27NkVHDQAA8po8E3CmTZtmw4cP9z8vWLDAQ8uVV15ps2bNstatW9tZZ51lK1eu9P0jR460+++/3x566CGbOnWq7dixwy688EKLxWIp/hYAACAvSIvlgVSwefNma9iwoS1dutS+/fZbe/jhh23Dhg02bty4+GvOOOMMa9q0qQ0bNszq1atnt956q/Xo0cP3bdmyxas5L7/8sp177rkH/TvT09MtIyPDypYtawXJ0f3eTvUhIBctvf+8VB8CABwyB3v9zhMVHIWVSpUqxZ9Pnz7dOnbsmPSarl272pQpU+znn3+2RYsWJe0vU6aMXXTRRb4fAAAg5QHn9ddf90rNCy+8EN+2bNkyq1u3btLr9HzJkiW+T303VatWzXI/AABAkVT+8jVr1tiNN95oDz74oB177LHx7eqpKVmyZNJry5cvb9u2bctyX+L+7OzatcsfiSUuAAAQppRWcK677jpr3ry53XTTTUnbS5QoYTt37kzatmnTJg82We1L3J+dIUOG+Jhd9KhRo8Yh/CYAACAvSVnAefLJJ+2dd97x6eBFihTxhzRo0MCqVKmyz3DT4sWLrXbt2lazZk1vKl6/fn2W+7PTv39/b0iKHsuXL8+hbwYAAApswDnzzDPt66+/ti+//NLmzZvnj6gnR/smTJiQ9PqxY8dau3btPPxoFlXifg1b6bn2Z6dYsWLebZ34AAAAYUpZD06dOnWy3H7MMcdYz5497eSTT/ap423btrXXXnvNZs6c6evfSO/evX1RwHLlyvlQ07333utNxgc7RRwAAIQtpU3G2dEKxgo1CjIDBgywE0880d59912rXr2677/hhht8iErr4KhZWBWfiRMnWqFCKZ8UBgAA8oA8sdBfKrDQHwoKFvoDEJJ8tdAfAADAoUTAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAITpFUHwAA4BAalJ7qI0BuGpSR6iPIs6jgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEJyUBpw1a9ZYp06dLD093WrVqmWPP/54fN+cOXOsadOmVrx4catfv75NmjQp6b2jR4+22rVrW4kSJaxNmzb2/fffp+AbAACAvCilAadDhw62detWmzJlij344IM2aNAge/75523dunV29tlnW4sWLWzmzJl27bXXWseOHW3u3Ln+Pr2+W7du1rdvX5sxY4ZVr17d2rdvb9u2bUvl1wEAAHlEkZz40L1791qhQoUOWL2pU6eOPfLII1ahQgVr1qyZTZs2zV599VVbvny51a1b1x599FF/bePGjW3+/Pk2ZMgQe+WVV2z48OHWs2dPDzkyatQoq1evnld1unfvnhNfCQAAhF7B6d27dzx8ZGXAgAHWq1ev/X5G1apVbcyYMR5udu/ebR999JG99tprVrlyZZs+fbpXdxJ17drVKzexWMyrNqroRAoXLmydO3f2/QAAAH8o4ChkDBw40Bo0aOBDSX369LGJEyfazp07vTLz9NNPW8WKFQ/680455RRr3bq1BxUNUy1btswrOIn0fNOmTbZ+/Xofispq/5IlS/7I1wEAAAU54MybN+//3lSokN155502duxY69Gjh5UuXdruu+8+q1mzprVr185atmxp/fv3P+jPfe655+yZZ56xww47zN544w3bsWOHlSxZMuk15cuX958KUZLV/v314Ozatcs2b96c9AAAAGH6XT04F154oaWlpfnMpXPOOceOP/54DwqqvBx++OG2ePFi77/R8JNed7D+8pe/+OP000/3GVOqxkRBJqLqjRQrVsx/an9iyNH+zKEnkfp3Bg8e/Hu+LgAAKAgVHA1DabaTqjTjx4/3xt7XX3/dg4Wae3/66Sef9fThhx/aiy++uN/P+vLLL2327NlJ29R0rGqQ+nAyDzcpPGk6ufaVKlUqy/2aNp4dVZQyMjLiDzUyAwCAMP2ugHPMMcfYeeed5yHntttu89lMCjhVqlSx9957zys7CiGavt2vXz/bs2dPtp+1cOFCbwxOfM2iRYvsl19+8SnfEyZMSHq9hsMUrFQZOvXUU5P2q2o0btw4358dVX7Kli2b9AAAAGH6XQHn/fff96Gkzz77zJ566ikPDQo3l19+uY0YMcKnbWvat2ZHqQ9n+/bt+x3uKlq0qK9xo0X93nrrLbvooov8+V//+lf75ptvvHlZfT+asaU+naivRzO0FLI0PVxr4+j3aj0dvRcAAOB3BZy2bdv60JOqNFphWMM8L730khUpUsSOOuooO+mkk7zXRT+1fX9VElV7tDrxhg0bvGKjdW00Nfzf//63D0NNnjzZpk6d6jOsFKZUoWnSpIm/V69XoLr77rt9MUANV6mCpKErAACAtJjmfB8kBRuFCL1FwSSauaThKA1VqbLTsGFD27hxo/3v//6vNyHnVWqO1vdRP05BG646ut/bqT4E5KKl95+X6kNAbhqUnuojQG4alGEFzeaDvH4fdAVHoUYrDWuY6uOPP/aqjVYf1nNN616wYIHdeOON/ktvv/12r7AAAACkwkEHHDX3qmqjHhzdNqFGjRreJKyG4rvuusvvB9W8eXM7//zzfShLM64AAADy/Do46nfRKsOayq21bhRwNCylBuBKlSrZdddd56/bsmWLlSlTxhfXi9atAQAAyJMBR43FRxxxhP9Za92IZjZpDOyGG26Iv06h5pNPPiHcAACA/HMvqkSq3OgWC4latWrFjS8BAED+DTiZvfzyy76uzerVqw/1RwMAAOR+wFF/zq233ur3lLrlllsO5UcDAADkfsD54Ycf7IwzzvCbbuqO4AAAAPk24Kxfv95XFNYCf7r5ptbK0Vo4AAAA+WIWVYMGDfzWCVrzRje41Ewp3SDztNNO85WMzzzzzJw7UgAAgEMdcLSmjW7LMHv2bJ8ivmrVKt+mxf1at27tIQcAACBfDVFpTZsff/zR7/K9ePFiv1O4bs9w6aWX2gMPPODDU+PHj8/ZowUAAMjJHhzduuG4447z+07p1g1XX321r2zco0ePP/qRAAAAud+Dkx0t9Hfvvfday5YtrWPHjj50NXLkyEPx0QAAAKldB+ecc86x1157zUaNGmVjxow5lB8NAACQupWMFXK6detmM2bMONQfDQAAkHtDVJmp6bho0aI58dEAAACpCTilSpXKiY8FAABIzRAVAABAqhFwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4KQ04mzZtsquvvtrKli1r1atXtzvvvNN27drl++bMmWNNmza14sWLW/369W3SpElJ7x09erTVrl3bSpQoYW3atLHvv/8+Rd8CAADkNSkNOFdddZX98MMP9tZbb9nTTz9tEydOtNtuu83WrVtnZ599trVo0cJmzpxp1157rXXs2NHmzp3r75syZYp169bN+vbtazNmzPBw1L59e9u2bVsqvw4AAMgjiqTqF8+fP9+DytKlS61q1aq+rVq1atasWTN/XrduXXv00Ud9e+PGjf31Q4YMsVdeecWGDx9uPXv29JAjo0aNsnr16nlVp3v37qn6SgAAoKBXcH755Re74oor4uFGjj32WNuzZ499+OGH1qFDh6TXd+3a1QNRLBbzqo0qOpHChQtb586dfT8AAEDKKjitWrXyRyINVamKs2bNGq/gJNJz9eysX7/eh6Ky2j9hwoRsf596e6L+Htm8efMh+y4AACBvyTOzqFauXGm33nqrDRw40Hbv3m0lS5ZM2l++fHn/uXPnTv+Z1f799eBoeCs9PT3+qFGjRo58DwAAkHp5IuBs3brVzj//fG8qvvnmm31mVBRkIqreSLFixfxnVvszh55E/fv3t4yMjPhj+fLlOfJdAABAAR6iivz222922WWX2d69e23MmDGWlpZmNWvWtCVLliS9bvHixV55qVy5spUqVcr3V6hQIWm/po1nR8EoCkcAACBsKa/g9OjRw7766it7++23rUyZMr5NvTmZ+2nGjh1r7dq18wB06qmnJu1XOBo3bpzvBwAASGkFZ+jQoTZy5Eh7/vnnfZhq4cKFvv3KK6+0YcOGWZ8+faxLly42bdo0e+6552z69Om+v1evXnbxxRdbrVq1rFGjRvbEE0/4+7VeDgAAQMoCjtazUV+Mpn1rinei119/3SZPnuxr2jzyyCM+Q0oVmiZNmvh+Leo3YsQIGzx4sK1evdpatmxp7733ng9dAQAApMWUMAogTRNXT48ajnWriILk6H5vp/oQkIuW3n9eqg8BuWlQeqqPALlpUIYVNJsP8vqd8h4cAACAQ42AAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSmS6gMAABw6R+98MdWHgFy0NNUHkIelvILTrl07S0tLsyuuuCJp+5w5c6xp06ZWvHhxq1+/vk2aNClp/+jRo6127dpWokQJa9OmjX3//fe5fOQAACCvSmnA2bNnjz3++OPWo0ePpO3r1q2zs88+21q0aGEzZ860a6+91jp27Ghz5871/VOmTLFu3bpZ3759bcaMGVa9enVr3769bdu2LUXfBAAA5CUpDThFihSx4447zipVqpS0feTIkVa3bl179NFHrXHjxta7d2+v8AwZMsT3Dx8+3Hr27Okh56STTrJRo0b5Z6mqAwAAkPIhqqxMnz7dOnTokLSta9euXrmJxWJetVFFJ1K4cGHr3Lmz7wcAAMiTAWfZsmVewUmk55s2bbL169f7UFRW+5csWZLtZ+7atcs2b96c9AAAAGHKkwFnx44dVrJkyaRt5cuX9587d+70n1nt318Pjoa30tPT448aNWrkyLEDAIDUy5MBRzOjoiATUfVGihUr5j+z2p859CTq37+/ZWRkxB/Lly/PkWMHAACplyfXwalZs+Y+w02LFy/2ykvlypWtVKlSvr9ChQpJ+zVtPDsKRlE4AgAAYcuTFZxWrVrZhAkTkraNHTs2vmbOqaeemrR/7969Nm7cON8PAACQJys4119/vQ0bNsz69OljXbp0sWnTptlzzz3ns6ukV69edvHFF1utWrWsUaNG9sQTT9jWrVt9vRwAAIA8WcHRMNTkyZNt6tSpdsopp9hTTz3lFZomTZr4fi3qN2LECLv77rt9MUANV7333ns+dAUAAJAW08IyBZCmiaunRw3HZcuWtYLk6H5vp/oQkIuW3n9eqg8BuYh/3wVLQfz3vfkgr995soIDAADwZxBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQHAIOAAAIDgEHAAAEh4ADAACCQ8ABAADBIeAAAIDgEHAAAEBwCDgAACA4BBwAABAcAg4AAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAEBwCDgAACA4BBwAABIeAAwAAgkPAAQAAwSHgAACA4BBwAABAcAg4AAAgOAQcAAAQnHwbcNauXWsdOnSwUqVK2ZFHHmlDhw5N9SEBAIA8oojlU506dbKiRYva+++/b6tWrbJu3bpZ6dKlrUePHqk+NAAAkGL5MuDMnDnTPv/8c1u2bJlVqlTJt23ZssX69etnN998sxUqlG8LUwAA4BDIl0lg+vTp1qpVq3i4iSo6GzdutK+//jqlxwYAAFIvX1ZwVLmpW7du0jb14lSrVs2WLFliDRs23Oc9u3bt8kckIyPDf27evNkKmr27tqf6EJCLCuL/xgsy/n0XLAXx3/fm//+dY7FYeAFnx44dVrFixX22ly9f3rZt25ble4YMGWKDBw/eZ3uNGjVy5BiBvCL9v1J9BABySkH+971lyxZLT08PK+CUKFHCdu7cuc/2TZs2WcmSJbN8T//+/e2OO+6IP9+7d69t2LDBg1JaWlqOHi/yRuJXmF2+fLmVLVs21YcD4BDi33fBEovFPNwcccQR+31dvgw4NWvWtGnTpiVt2759u61evdpq166d5XuKFSvmj0TlypXL0eNE3qP/+PEfQCBM/PsuONL3U7nJ103GajD+6KOP4n00MmHCBP8f9oknnpjSYwMAAKmXLwNO8+bNrVGjRnbFFVfY7Nmz7c0337Tbb7/dBg4cyBRxAACQPwOOvPrqqz7kdMYZZ/jaN7fddpv17Nkz1YeFPEr/W/nnP/+5zzAlgPyPf9/ISlrsQPOsAAAA8pl8W8EBAADIDgEHAAAEh4ADAACCQ8ABAADByZcL/QEHopuu6r5kunWHVrfWApCskQQABQcBB0F58skn7Z577vE7y+vmq7o/mW7hoVWutfLlP/7xD+vevXuqDxMAkMMIOAjGsGHDbPjw4f7o1KlT0n3JdINWLQip+5Hpth69e/dO6bECAHIW6+AgGLrZ3mOPPWYXX3xxtq9RyOnRo4etWLEiV48NAJC7qOAgGBqWOvLII/f7Gg1bacgKQP7TpUuXg37tiy++mKPHgryPCg6C0aFDB9u8ebONGzfOe28y27Jli11++eVWvHhxGz9+fEqOEcAfp3sN3nLLLVaxYsUDvla3bkDBRsBBMH7++Wfr2LGjzZ071+84X7duXe/D2blzp8+o0h3odZNW3cescuXKqT5cAL9T+/btbcSIEXbsscem+lCQDxBwEJyZM2faxx9/bMuWLfPm4hIlSljNmjU99JxyyimpPjwAf8Jvv/1mhQsXTvVhIB8g4AAAgOCwkjEAAAgOAQcAAASHgAMAAIJDwAEAAMEh4ADI1zRPYs+ePVnu27179wFfq1t86B5mAMJCwAGQL0ycONFKly5ty5cv9+fr16/3KcNawLFo0aL2xRdfZHn7Dr2nXLly/tDrxowZk/Sab775xtauXZtr3wNA7iDgAMgXtGhj9erVPbTIo48+ak2bNrUiRf7vjjMVKlTI8j2TJ0/223Po0a5dOytWrJjfXV4BKVodF0B4uBcVgHwhWtxt1apVvhL1jz/+aKeffrqtW7fOtyu07N271xd3rF+/frbhJdp2+OGH2y+//JKr3wFA7uH/dQGQ5x111FG2bds2//MNN9xgo0ePti+//NKeeuqpeJhp06aNL+HfunXr+PsUeDLTtrS0NH9o2ApAmAg4API0BZsVK1b4cJPcfvvtXsX54YcfvHcmGmr69ttvvYFYlZzEJuNrrrnGGjRo4I/Zs2fbrl27UvZdAOQehqgA5Gmq1Oh+Yqq4RDdcbNKkiZ199tkeen799df4axVwFGqivpyVK1fa4sWL/QaNAwcOjPfprFmzJkXfBkBuoYIDIE/TDVJ/+umnpG0KKnfddZcHmbJly/q2Y445xmdJ9enTx5+rwqNwpJuvPvPMM171mTdvXpazrQCEhwoOgDzfXFyxYsV9to8aNcorOAo1pUqVsu+++87Kly8fr+hoOrgeajpWVadTp04+k0ozsd56660UfBMAuYkKDoB8qVq1apaenh6fFaWfCjhVqlTx54MHD7YlS5bY3/72N7vgggts4cKFPkwVDWup2VgL/wEIEwEHQLDUk6MZV+rXETUYq59HAadZs2a2devWVB8igBzCEBWAfGfBggXef1O8ePH4kJRmTynQZGRk+HTxDRs2WPfu3X3NnK5du9r27dt93RtVcGrXrm2zZs3y9ynkRGvsAAgHAQdAvqA+mujeUldffbV9/vnnPrNKQeewww6z0047zW/dIJ9++qm99NJL3m/z0EMP+f4LL7zQp5I/8MAD/hoNWTVv3tz39erVK6XfDcChlxZjEBpAPjB+/Hhf5E+VGc2I0iJ90do4Ef3nTPt1/yndkuFAFi1aZHXq1KGCAwSIgAMAAIJDkzEAAAgOAQcAAASHgAMAAIJDwAEAAMEh4AAAgOAQcAAAQHAIOAAAIDgEHAAAYKH5fzCrF9lkvSvSAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "fig.set(alpha=0.2)  # 设定图表颜色alpha参数\n",
    "\n",
    "Survived_m = data_train.Survived[data_train.Sex == 'male'].value_counts()\n",
    "Survived_f = data_train.Survived[data_train.Sex == 'female'].value_counts()\n",
    "df=pd.DataFrame({u'男性':Survived_m, u'女性':Survived_f})\n",
    "df.plot(kind='bar', stacked=True)\n",
    "plt.title(u\"按性别看获救情况\")\n",
    "plt.xlabel(u\"性别\") \n",
    "plt.ylabel(u\"人数\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "C89799DF55AB47B283D543D89B63F3A3",
    "mdEditEnable": false
   },
   "source": [
    "歪果盆友果然很尊重lady，lady first践行得不错。性别无疑也要作为重要特征加入最后的模型之中。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "159A753C26E24D25845EFCBE173B3D37",
    "mdEditEnable": false
   },
   "source": [
    "再来个详细版的好了。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": false,
    "id": "409F149602FF49C19E127C9C074FE6B4",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#然后我们再来看看各种舱级别情况下各性别的获救情况\n",
    "fig=plt.figure()\n",
    "fig.set(alpha=0.65) # 设置图像透明度，无所谓\n",
    "plt.title(u\"根据舱等级和性别的获救情况\")\n",
    "\n",
    "ax1=fig.add_subplot(141)\n",
    "data_train.Survived[data_train.Sex == 'female'][data_train.Pclass != 3].value_counts().sort_index().plot(kind='bar', label=\"female highclass\", color='#FA2479')\n",
    "ax1.set_xticks([0,1])\n",
    "ax1.set_xticklabels([u\"未获救\", u\"获救\"], rotation=0)\n",
    "ax1.legend([u\"女性/高级舱\"], loc='best')\n",
    "\n",
    "ax2=fig.add_subplot(142, sharey=ax1)\n",
    "data_train.Survived[data_train.Sex == 'female'][data_train.Pclass == 3].value_counts().sort_index().plot(kind='bar', label='female, low class', color='pink')\n",
    "ax2.set_xticklabels([u\"未获救\", u\"获救\"], rotation=0)\n",
    "plt.legend([u\"女性/低级舱\"], loc='best')\n",
    "\n",
    "ax3=fig.add_subplot(143, sharey=ax1)\n",
    "data_train.Survived[data_train.Sex == 'male'][data_train.Pclass != 3].value_counts().sort_index().plot(kind='bar', label='male, high class',color='lightblue')\n",
    "ax3.set_xticklabels([u\"未获救\", u\"获救\"], rotation=0)\n",
    "plt.legend([u\"男性/高级舱\"], loc='best')\n",
    "\n",
    "ax4=fig.add_subplot(144, sharey=ax1)\n",
    "data_train.Survived[data_train.Sex == 'male'][data_train.Pclass == 3].value_counts().sort_index().plot(kind='bar', label='male low class', color='steelblue')\n",
    "ax4.set_xticklabels([u\"未获救\", u\"获救\"], rotation=0)\n",
    "plt.legend([u\"男性/低级舱\"], loc='best')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A8D55FBCD7EA43BF8C7CF42851098D95",
    "mdEditEnable": false
   },
   "source": [
    "#### 我们看看各登船港口的获救情况"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": false,
    "id": "18C26D1BD0BC41CE8C1CFF996D2225B2",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjgAAAHFCAYAAAD/kYOsAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAARYNJREFUeJzt3Qm8TfX+//GPoUyZxzJkjJIiURpEhhRXhQZRNxpUKDJfDaSbIRoURRpoIsRVuYpUhkh1pQylDHEVmech2f/H+/O7a//3Oc4p5Zyzt3Vez8dj33POWntYe+19W2+f75QlEolEDAAAIESyxvsAAAAA0hoBBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BBwAAhA4BB8Cf9uuvv9q+ffuMidABJCoCDpCOPvvsM+vcubN9//33afq8v/322x++7umnn27vvffen36OAwcO2FtvvXXU9uuvv95vsmrVKsuTJ4/NmjXL4kkBa+XKlX/qMUOGDLErrrjCQ1pG++STT2zy5MnRv99880174IEHjumxu3fvtvvuu88++uijo/YtXbrU/vGPf/jzp2bbtm2p7nv00Udt1KhRdqwOHz5s27dvT/E7pH1bt2495ucC0gsBB0hHTz75pF+QChcufEz3/+GHH/yC/eOPP9p///tfW7dunYejX375xV577TXr37+/bdmyxWrUqGGTJk1KNaB89dVXtn//fmvcuLEdOXLEDh065Pu6dOlijRo18gvT+++/7yFl9erVSR7/xhtv2A033GAtWrTwi6oujAoS2bJls9y5c/t98ubN6z9Lly7tF7SdO3cedRw65j179vhx6JhSuqkKpMem9Hi9T7337777zv7zn//Y7NmzbcKECX5O9T6uvvpqK168uFWuXNk+/PDDox6v4167dq39/PPPfiy66O7YscOyZMliH3zwgYdAXaT1Otqv861bctqm17nnnnuSbNc5OeOMM+zaa6+1gwcPpvhZLF++3NavX+/Pv3nzZnvhhRds3Lhx/pq66XPSe9M+3eenn36yb7/91s97cjru4cOH++vOnTvXzj33XH+MzJkzxwYOHBj9nFOioF21alV/bh2vPnc9XudE3wWFVn1eOi59/zZu3Bh97LPPPmt9+vSJ/r1hwwYrVKiQZc+e/ajbSSedZEWKFPHvHRBXWmwTQNqbP39+JEuWLGrDSfV26623JnnMZZdd5tuzZs0ayZYtm9/09+zZsyNjx46NFChQIHLOOedEbr755sjmzZsjP/3001Gve9ppp/nr6nbyySf740ePHu37OnToELnyyiv9908++cT3peT111+P5MuXL/LNN99E8uTJ48+lY9ItR44cftNjs2fP7j8bNGhw1HOULl36d9977O2BBx446vH/+c9/oueiYMGCkQoVKkTOP/9833bddddFHnzwwchTTz0VefHFFyOzZs066vEDBw485tcPbjr/yX311Ve+b+LEiUm2f/vtt779ySefjKSmcOHCf/oYdPvoo4+Oeq59+/ZF9+3duzdSrFixyKhRo3yfzsd5550X+T133HFH9LNfvHixP1fwHdE5PumkkyI5c+b037VP35XAbbfd5tsXLVrkf69fv97vM378+MiKFSuS3AYMGODfCyDessc3XgHhpH8J33bbbV71ePjhh61Vq1ZJ9t9///22YMECe+SRR5Jsf/nll73aky9fPv979OjR/q/2evXq+b+6e/bsadWrV7exY8faxIkT7ZZbbrEXX3zRbrrppuhzqCozaNAgu/jii73yoMeecsopvk+VkuC5s2ZNvYCr52vYsKEVK1bMpk+fbvnz57cHH3zQ348qBfpX/qWXXmpTp061EiVKpPgcqvao0vTQQw+l+jqqHrRr185y5MiRZLsqL6eeeqpXElQpOPnkk/14N23aZKVKlbJevXp5BUOVKFWQdG5UVVCV47TTTos+T5kyZWzZsmXR965qjKoYKenWrZtXiwJ6blWfVEXS+1Y1THbt2uXnUOdF2y+88EKvuogqUjpXFSpU8L8XLVrk1a6cOXN6dUNNTDqO119/3fc/9thj9uWXX3plSk1mqsCowlK0aFHfv2TJEpsyZYp/psHn9fbbb3vVR5+RjkPHqGN55plnkhy7nitXrlzRbXq8bmq2rFu3rh+zXkefk5rs9H3RZ6zKmipKOv+BwYMH+3F06NDB31NwLGoGrVKlSpLzqO/D7323gAwT74QFhM2vv/4aueaaayLVqlWLTJ482f8Vv2HDhuj+4cOH+7+Gp0yZkuLjb7jhhsgzzzwT2bFjR6RUqVKRadOmRfe99tprkbx580buu+8+r9So0nL48OEkj69cuXLk5ZdfjuTPnz/yyy+/RE4//fTIm2++6fsqVaoUeeihh/z3uXPnplrBSUmbNm0if//73/33n3/+2R+rf7GnRsdxxRVX/O5zBs+jf/XHUrXmr1Q+WrZsGX2OkSNHRi6++OLo3x07dvRK0Jw5c7z6Fdw2bdoU2b59u58XnfvA1q1bkzy3Kh2qqOlzlQsuuCDFYxg2bFiS96LP/siRI/77G2+84ZW9wNq1ayPLly+P7g9+BqZPnx6pUaNGpG7dupHLL7/cn19VrHr16vnPHj16+Ged0nF069bNn2PEiBH+Oagic9VVV0Vq1qzp1S391P1UtdH70vsLKnP67iT37LPP+j59B4PPbcGCBUfd74UXXvDnAeKNgAOkseeff96bdZYsWeJ/KxSUL18+8t133/mFRU0Buiil5q233vKLji5iF110kT9OF8Iff/zRfw8udKkFpCDg6AK4cOHCaMBRGFKTRHBc8+bNizaxKHQdOnQosmfPHm+mCZpIvv76a2+i+f777z20KUCsWbPGmyr02JkzZ/oF+tNPPz2quex4Ao5CgcKHQp6aYzp16uQX1qBp5IsvvvC/dY52797tz6Nzo0CXkueee+53g5Ga65LT+dA+fZ56Dd0U8hRsgnMX+7igCSloNgrUqlUr0rRpU2/uK1GihIeM3377LUnIVBPiP//5z0jbtm1TDc379+9P0nylbQpEZ599tgdefUYKIQory5Yt83MlXbp0iTRu3NgDjo5j6tSpkcGDB/t36sCBA34ffeZ9+/b133VsOp8bN270cxB7DPoOyX//+9/fDTj6jgPxRsAB0oHCSOyFoUmTJt4vQRefd99993cfqwtT8eLFI3Xq1PELhUKJfga/68Jyyy23+HO1atXKKzzBhUqvFQScGTNm+HMp4AwaNChStGhR70sR0AXszDPP9DCmKoEo/JQrV85f5+GHH/aKxrFWT5L3UTmegCMKD+rnogu3qia6cAb3Vx8S9b259NJLPQzpp8JHcrt27Yq0a9cukjt37sjTTz8dmTRpUnSfQpoqFUFFKzmdS71WbBhVWFXAUWVIFZBYCmRBhSOWgp+OX1U7PS72u/H555/7Y3Qsb7/9dqRQoUL+HmPdc889HnxiA44qO/qcevfu7e9Bry2q/JUsWTLJ4xVMb7/99mjAkV69ekU/t6B/lb6fQZ+q4KZQlpI/6oOj7yYQb/TBAdKB+n6oL4NG/qi/hYZTqz/KN998431zNEJJfWM0Akh9HdTHQn07NFpFfW7UR+KJJ57wvgwFChTwviWBYMSRnqNv377WvHlzf/4GDRp4nwy97l133eV9PoJ+Ieor8e677/rrBdQ/R6N89u7d631c5JxzzvE+K927d/eRMM8//7wPH9ZzXHbZZXb33Xdb165dvd/LWWed5SOL1C9D+/Ue0pJeT+9f7yOlId3ap34kOo/qi9S6deuj7qPh7v/+9799lJBGVbVp08YWLlzox64+N+ofpP5Qv0ejsII+NsG5Hz9+vPedUX8k9QfSsegcSPLzoFF06h+kz+aaa67xnxrBpb5WGjUl6kej78c777xja9as8c9c/XZE70v9nXTug/5dGhHVsWNHmz9/vlWqVMn7yDz++OM+Mkoju2Kpr5c+p+C1pHfv3v7+1TdL/Z9i+8wEcxzpmJKP/tNzqI9TMDxc/YD0D2X141E/KI2c0nNpf/A7EDfxTlhAWKg5RZUS/Wu7evXq3g9B/5LVyJUPPvggWtpXvwlVU2L/pazqw/vvv++VG42WUqXhjyomqtCoOUE/A6tWrfL+K2qqUBOEmmw0mklNVKpgJB+hFVSEVCVJzerVqyNnnXVWpH379tGmFTWNqJKgprTHH3/cm2eSO94KTlCVEvVB6dq1q1drdP8vv/zSz1PZsmW9SqPqxrp167w5K7XnUV+lxx57LHr+rr/+eu9nk5qggpP8pgqOqIqkc6l+NaKmOu0PPmtp3bp15NRTT/Vtr7zyivcB0nkLKieqyul39atSlSkYwRQ0LwU0Siy2gqPvij6DgwcP+mcdnGedJ1VnYumcbdu2LUkFR1S90+sFfW+Cm74XqoilRNUgNW/pe6DXVrOWml+D6lZwHCl9DkBGI14DaUSjZzTCRHOUaKRT+/btvXKgqoz+la2RQfrXsf41rXlLVKm59957fYTVnXfe6aN0NGpI1Qj9i1jP979m5CS3YI4U3Uf/QtYImED58uWjVZ8LLrjAtwX/ita/1vWvb1VoNDGcbq+88orvC0ZWBTTSRv9aV/WpVq1a/vtLL73ko4Z003PWr1/fKyGal0ajrTSaK61ozhqN6Pnb3/5mJUuW9PM1bdo0rzqIJsfT6J2gMqLRQuXKlTtq4jlVbWbMmOHnXM8zcuRIP+96319//bWP+FF1RM+rapUqPsknqdPItuDc//3vf49u12esKprmh1HFI6UKjkZJaSSUqjOaTE/30TFq9Jh+jhkzxu+nCpwqaaqMaV/sSDBRdUQj1kSP08gufZdUeTv//PNt8eLFPqePqjUa/RZLz1WwYMGjzrEeqwqgXldVoeCmilxKlRe9j3/9619e2dN+PV7vJ5gTKVYwXxIQTzRRAWlITRea6Ex0wX3uued+9/6aoO7yyy+P/h0Ml9ZFTBfUlCaeCy6kqdHFUM+jSfhiH6+Lko4tdlhvMJlb8guaJvvTxVfNRAo5GpKui28QiKRTp07+U0PSdf+Umoj+Kg0NV9PSmWee6ZMbqhlKAUQhSxTO9Leaj4ImNjWnBOde5+Ciiy7y4dRSsWJFb3bTTMxBc5wmPFRw0mem59J51esqNB0rBVM9Rs2COo7kAads2bLerNWkSRNr1qyZh19N/KfmxOB9KHhq0sbatWt7U6SCbUDNPAqQmqU4GAaupkwFNR23mh31OIU8BWMFTQXPY6XPTk2hsTS8XOcuOQU5vZ/YkKdAruAIJCIqOEAaCi6wopChOXBSqsKon4UEF9vkVKXRhUwhSRdn9Z1R3wr9rqqJBBfU5PQ4XWR1wU8+S/Gx+vjjjz14acbgoEKkuXd0AQ5uI0aM8IuhLtC60Kf0r3b141DlKrWbLvYp0fOrwqKKys0332wDBgzwcKOgFcwPpNmIVfXSc6gSk/zcq0+KApmqapohukePHt4XSX2edNOFWaFMVQlVhNTPRkEj6PtyLBSS1EdJISelCo6qZaqA6XPWEhHB5yP6HmjZBvVp0tw2CgvJKdjoO6Q5aBRgRNUYVXw047Xen86VZp5W2NF9VWE7VuqLo4pR7O2OO+446n6qnilwqr9S8Pw6fvVnUr8tIBFRwQESkC5YQcdgNXMEVNlRJUBNW7q4JqcJ+HSR0r/kNXmcmg+CZp1jpaYINUmpehPrxhtvPGoyOV3stE1BQU0XsVR9UIVEIe2PpLRopy7cauJTh1oFPoUuvS9RoFEH26uuusouueQSn/RP56NatWrekVjHFVTGgiqWOvuqEpScqhJ6LYXI2E7YAU1EqFsgaPpLLmg6jA046jSsztxq8lMA07EEk/gp3CjUqClTQVHHrQn7YpsL9Tnr89Z7VJUooIA2b948b/pU2FEzqEKUKoaqDqmpMq0qOPo+qVqnEKdjDSgY6nui40uJvhcKp2ouBeKBCg6QjtS8Elv1CG6pXfRVbdBN//JXM4uqEKqcqAKhmy5ganZR9UFrB6nKoVtAFQ6NOlIzlKoYmpE2dl0jXVBjjyOl5gwtAKmqzNlnn51ku55PISa46cKtn7pAqhKQnC7ImuVX/UJSuwWLgSZfQ0nBqGbNmt60o4qF7qtKQWwQUgD84osvvG+LKmJDhw71Y0+tgqH+OGqSS35LbdFRNcmJAuaKFSv8plFQsWss6XhU2VAwUTNX8oCj86JmSJ1PjZzSTMwKJfocNApKAUABS5Uafb6aTfjzzz9P0n9G4Ub7FJYkGFGn86KZkVVha9mypVd0VDHT56ZQqH45sXTcyYOkXje2/41uquDEjtrTa6t/z+233x6t0un4FXo02k/HnBJ9d9VHCYgXKjhAOlEnUDU/pDQMWf+Sr1OnzlEXdjX1qIKif+0rqOiComYiXbSDi5MuyOoYq58KEboAqrohasZRaNFFVsOY1QFV1OyiphFdMGMXplSYUp+K4CKvC5z63qgDbWy/HP2uZQzUuVXHpIu/7quLrpqamjZtetR71MVPzWoKOanRc6ija/I+H7pIa5/ChfrOxPatiQ0f6lysx+scqaKgZRyS00VfFAJSk9LCkEGTk6olQb8l9Y9R0AkoCOgcBH2tFF6Sd9hWWFXHZjUzaZkInUudG30mCmWix6gTsapQ6lOj/jWq4gV0DvT+FGr0Gtqv6QaCIBF8xxQMVWVRRUffA3VuDui7FrsoaPLvnr4Dem/6/sU2N+o7pEpicG71GqrY6ZwpfAf0HdV7VadshS01HSrEAXGT4eO2gExCQ4I1WV5KNPmbhht/+OGHSbZrKHDy6fr/SOxw6s8++8xnqg1oEcahQ4f6jLQa3qtJ/2IFQ5aXLl3qf2vItSaP05DsWJrEL/nQdk0Kp5l577333ujw8bSU0nnQ8PDkQ7H/SDDrcEoLWIrOyf3333/Udp0LTS64cuXK6DZNZBe71ILoM9RQfE2KqIkJk9u5c6cPFdcwak0ToOH2WjIjpaH5+hw0tDz58ht635ogUlMRBDTRYewSILFSOj8tWrSITugoZcqUOer7qckL9TmPGTMmxefVRJCaGqBKlSqRH374Icm+cePG+XB3DX/XcHMNKU9p+gAgo2TR/8QvXgFIREH/mtT2qQoR24yR6JU0jWRS9SW2I3K8qPKh6lTySk9GUzVHn3EwIeSx0kgwVYaCBVyBREXAAQAAoUMnYwAAEDoEHAAAEDqZdhSV2sE1h4PmCTlR+hIAAJDZRf63ZI1Ghf7egq6ZNuAo3GgqewAAcOLRVAQpTQ1hmT3gBAvE6QTFezQDAAA4NprTSwWKlBZ6jZVpA07QLKVwQ8ABAODE8kfdS+hkDAAAQoeAAwAAQoeAAwAAQifT9sEBAOCv0HIlWgIE6UNLqqS2VMyfQcABAOAY51/ZuHGj7dixI96HEnoFChSwEiVKHNc8dQQcAACOQRBuihUrZrlz52aS2HQKkfv27bNffvnF/z711FP/8nMRcAAAOIZmqSDcFC5cON6HE2q5cuXynwo5Ot9/tbmKTsYAAPyBoM+NKjdIf8F5Pp6+TgQcAACOEc1SJ855JuAAAIDQIeAAAJDJHD582I4cOXLM9w/uq07A6o/0R9RRWPdNaR2pjELAAQAgk6lfv7716NEj1fCzZ8+e6N+33nqrPfroo/77qlWr7PTTT4+GF/WRUZhJ7rbbbrO+ffvajz/+aFOnTo12Gi5atKi9+eablhEYRQUAwHEo2/u9DHuttYOa/qXH7dy50zvuahI9yZMnT5LVuIOQovt88cUXVqdOHf9bI5gUZrJmzWqPPfaYbzt06JCdfPLJHoSkQ4cO9vzzzyd5Pe3XCKitW7da69atbc2aNbZo0SJf3LpFixaWEQg4AACEXM2aNb36osCikKPKy+zZs23QoEH+u5qgHn74YevXr5/fd/v27T5cW0Hl7rvvtipVqliXLl1sw4YN1rhxY1u2bJk/5sCBA9Emq6+//to+/vhjO+WUU2z16tW+XSGqffv2tn//fq/kqLKTI0cOv7+2BUPC0wMBBwCAkJs5c6blzJnTbwot1157rdWuXdt69+7tFZndu3d7VUcUgA4ePGjvv/++3XDDDVapUiVr0qSJ7ytZsqT16tXLqzwKMvo7qARt3rzZFi5c6KFFzVEKOAozau5at26dTZ482UdHDR8+PFoFSql5K60QcBJdv/zxPoLw6Lcz3kcAABlu+/btPiOwwk1ATU7Zs2f3kCKFChXysKEwom179+61AQMG2Msvv2yfffaZlS5d2qs48sADD9gTTzxhb7zxhpUtW9Z/lwYNGngTV40aNeyuu+6y6tWrW9u2be2nn36yf//733bFFVfYW2+9ZU8//bQ/px6fngg4AACEvEPxkiVLvHoSLGSpCoqakxRO1ESlv9Xk1K1bNxs6dKiVKlXKLr30Uhs1apQ9++yz3iylmZwVfvQ86iysjsd6nLZr7SgZPHiwVahQIfraqvZoiYt58+ZZnz59fNv69es9GKU3RlEBABBi48ePt5UrV3r/Gd3OPvts7+yrxSzVAViBQ01IK1as8H42CiyqvijITJo0ySsuCjkFCxb0Zizdt1mzZt7Epf462hcYMmSI1apVy39XE5UqP2qaUn8fVYREYevcc89N9/dNBQcAgBCr8r+mJVETkZqhFE403FvNRP/4xz/891jvvPOON2up2WnTpk3WqlUrr8Co2qNmqJdeesmuu+46r/rELqeg6k0wLFzNYOXKlfOqkYKQfl+6dKktWLDAxo0bl+7vmwoOAACZwPLly+2OO+7wKosowGhUU9OmTb2fTiz1ybn66qvtqquu8mrLli1bvGnr888/96CioBSEmGBUlCpB6mfzySefRIOPKkB6Ho2Yuummm6xly5beZHY8q4QfKwIOAAAht3TpUu9H06ZNG29eCqgfjTobq0lq+vTpvk39atT/Rv1k1KFY4UdBRkaPHu3DvjU8XFWeWK+99ppVq1bN3n33Xb+/RklNmDDBV1/Xc6pJTE1lzZs3z5D3TBMVAABxmHwvo4wdO9buu+8+b2565plnkiy9oE7DH374oTc3qZKjCoxGPE2cONGbtrS0wvfff+9hR8PGFWrUJ+fKK6+0unXr2nPPPWc333xztENxEIQ07FzUnKVJAAcOHGjDhg3zwKSOzHpOza+jalB6oYIDAEBIHTx40JuT1Dykn2pmElVUgr4zqrDMmjXLR1Tdcsst3rn4zDPP9OrOBRdc4Dft1wzETz31lDdfnXfeeT5JoO6vio4o3GjGZHUsfvvtt61MmTL+U/17Xn31VZ+LRx2Np02bZq+//ro3f+k40gsVHAAAQipHjhw+iil2DpygwrJt27bo3wonXbt2TXIfTeCnJq1OnTp5dUbhRutSBbT9jDPO8CUZAvnz5/dRVdqukKQAo8qNOikHsxbXq1fP+wOpz46asdJLlkhKy31mAiq76YNQ2tRwuYTFRH9ph4n+APxF6nOi9ZTUpJI8LCBjz/exXr9pogIAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAEAAKFDwAEAIBM58r9ZjIMVv3+P1pCKvX/sUO1ElzABZ+/evfbiiy9aJp2WBwCANHfkyBEPI8G1tV+/fnb77bf774cPH/aJ9jTnTPC3VhqP9cgjj1jbtm1t8+bNvvJ48JyapThYtDNRJcxMxg8++KDNmDHD17TQCV+0aJF17NjRvvnmG19+/fHHH/dpnWPX1ujfv7/9/PPPVqdOHRs1apRVqlQpru8BAJAJZeSErH9ywtJffvklunK3lmAIbrFhRTMWq5KjEKS1qHQtDuh6rJmKNSOxgpFmJ9ZyD/v27fNlGhJZQgScxYsX+9TOs2fP9pOppNikSRNPjVq5VAuBaYn1Tz/91GrUqOHrWXTo0MEX/KpVq5ZPH92oUSNbtmyZ5cmTJ95vBwCAhFCsWDHbunWrL3qp66uqLhs3bvTrphQoUMB27Njh4UbrVun2448/2pQpUzz4fP31195MpYU2b7vtNg9HU6dO9XWptDq46DEnnXRSdKHNRBH3o1F6VFhp166dXXLJJb5tzJgxVrFiRRs+fLgHmu7du9uNN97oq5HK0KFDrXPnzv44Lfj18ssv+3LvquoAAID/o9Ch9ai0uKXCSfHixe3666//316zl156yT755BP76aeffG0qLYGgJq2FCxd6UUHrRW3YsMHmz59vW7Zs8VaTV155xas8Wolc1Rw9RtsTTdwDjpZa1wkcPHhwdNvcuXM9HcZSNUeVG6VMnWhVdAI6wa1bt/b9AADg/1NzkgoADRs2tPvuu8++/PLL6D5Vcj7//HNfEFOLZ0q1atXsgQce8NXCmzVrZg0aNLCRI0f6foWfokWLetVHoUkrjasPT8mSJS3RxLWJatOmTda3b19fsl3Lq6uZSdWbdevWeQUnlv7WCVWCVIfklPZrCXYAAPD/FSlSxBo3bmwPPfSQt4ioxUTNVlqoUlUdBRYVCdauXevbCxcu7NdiFR8UdkRNW+ou8u2333oFSH12tL9s2bKWqOIacAYMGOAnWH1p1HemW7du3uyk9j61F8YqWLCg/wx6e6e0X8EnNUHb4ok0xA0AgOPx66+/euVGIUatIxqsU7lyZe8wrH4zGjn1xRdfeFOWrr233nqrhxuFIVVoFHhE9500aZI3V/3tb3/z4sSSJUvs3HPPtUQVt4Cjkzpu3Dg/YUqWot+rVKnivbSDIBNQ9UbUlijaHxtytD956Iml/jsadQUAQGZx0kkneT/V008/3bt+qEuHWkw0Kkr7NMLq/vvvtx49ekQ7GkuhQoX8OqtrtegxusYq0Jxzzjm2fPly73h8xx13WKKKWx8cNTXt3r3bzj///Og2pUp1WtIHsXr16iT3X7VqlXdkUgpVtSel/eXLl0/19fr06WM7d+6M3lRaAwAg7E477TT7+9//7k1Vqr5o6LjCjbqDKMAE/VvVXJUzZ07bvn27N1mp6KD+O6LqT7169byic9NNN9ldd93l1+TatWtbosoaz6FrSoP/+c9/ottWrlzpzUwaTZW8P82ECRO8zKYPIPl+jcSaOHGi70+NKj9qDou9AQAQds2bN/fQoo7C+hkM537hhRfsmmuu8eKCpmoJaISU+tjMmTMneq3UY1W9UeDR8HBdr/W8iSxuTVQ6weqRrfLWM88842Gna9eu3vnpzjvv9B7dPXv29KSok6zmLCVIUV8dfSjlypWz6tWr+4nX7It6LAAAGepPTr6X0UaOHOkzD+u6q34zvXv39p8aQaX+N2o5Oeuss7x5StdhVW9uuOEGv79aWkS/a9Tz888/781Zeg5NsKvBQnpM1apVLdHEtZPxo48+6idNgUapUCdUJzxXrlyeIO+55x7vgKwRUqrQ1KxZ0x+n0VYjRoyIzmR88cUX26xZs5jkDwCAZCpWrOghR31xNNmf5rtR9UXFAlVvRMPIr7vuOps+fbpPuqsmK1VstJqA5p6bN2+etWnTxq/ValHR41WAaNWqlbeorFixwkdfJZIskUy6+JNGUan9UP1xErq5KiOnAA+7BP9XFoDEpQ63a9as8ZYD9VM50bz++us+K7GqLQo7Ki5oXapYmg9HFZvLL7/c/1bfHHU2VnFBUUHLOzRt2tRnPw6o6rN06dJoASIjzvexXr8JOASczIOAAyCTBpwTTVoEnLjPZAwAAJDWCDgAAByjTNrocUKeZwIOAAB/QPPGSDAvDNJXcJ6D837CjaICAOBEoBmA1blWk+SJpjbRvGxI+8qNwo3Os863zvtfRcABAOAYaII7CUIO0o/CTXC+/yoCDgAAx0AVG63dpJn4tYgl0oeapY6nchMg4AAA8Cfo4psWF2CkLzoZAwCA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0CHgAACA0Il7wGnfvr1lyZIleitSpIhvnz59ulWtWtVy5sxptWvXtkWLFiV53ODBg61kyZKWJ08ea9GihW3atClO7wAAACSauAecpUuX2tNPP20rVqzw28KFC2358uUeWtq0aeN/16tXz6644grbsGGDP2bMmDE2aNAge+KJJ2z27Nm2f/9+a968uUUikXi/HQAAkACyROKYCvTSefPmtXnz5ln16tWj2zt06GDbtm2ziRMnRrfVr1/fatWqZUOGDLHKlSvbvffeax07dvR9u3fv9mrO+PHj7aqrrjqm1961a5flz5/fdu7cafny5bOE1S9/vI8gPPrtjPcRAACO07Fev+NawVmzZo3t27fPzjjjjCTb586day1btkyyrW3btjZz5kz75ZdfbOXKlUn2KyRdffXVvh8AACBrvJun1O+mfPnyXoHp2rWrHThwwNatW2cVK1ZMcl/9vXr1at+nfjclSpRIcX9qDh486Kkv9gYAAMIpezxfPFu2bDZixAhvnlIn4Z49e3rAUZ+a3LlzJ7lvwYIFbe/evSnui92fmoEDB1r//v3T5X0AAIDEEteA07Rp0yR/q4pTp04dHzmloBNrx44dHmxy5cp11L7Y/anp06eP3X///dG/VcEpXbp0mrwPAACQWOI+iipWlSpV7PDhw1a8ePGjmptWrVrlTVllypTxTsVbtmxJcX9qcuTI4Z2RYm8AACCc4hZwNOLpkksuSTK0+8MPP/QOw40aNbJp06Yluf+ECROsYcOGVqxYMR9FFbtfzVb6W/sBAADi1kTVoEED69Spkw8Jv/POO30OHDUh9e3b15o1a2bnn3++nXvuuX6/yZMn24IFC3z+G+nevbv16tXLChQo4M1Mjz76qHcyPtYh4gAAINziFnCKFi1qM2bMsC5duljdunWtcOHC/rs6GmtklUKNgowCzznnnGMffPCBlSpVyh97++23exOV5sFRX5rGjRvbO++8Y1mzJlSLGwAAyIwT/cUTE/1lQkz0BwAnvBNioj8AAID0QMABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChQ8ABAAChkz3eB4DfV/bAG/E+hNBYG+8DAABkGCo4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdAg4AAAgdBIq4PTq1cu+/fZb/3369OlWtWpVy5kzp9WuXdsWLVqU5L6DBw+2kiVLWp48eaxFixa2adOmOB01AABINAkTcObMmWNDhw7135cvX+6hpU2bNrZw4UKrV6+eXXHFFbZhwwbfP2bMGBs0aJA98cQTNnv2bNu/f781b97cIpFInN8FAABIBFkiCZAKdu3aZeeee66tXbvWVqxYYU8++aRt27bNJk6cGL1P/fr1rVatWjZkyBCrXLmy3XvvvdaxY0fft3v3bq/mjB8/3q666qpjfs38+fPbzp07LV++fJaoyvZ+L96HEBprBzWN9yEAAI7TsV6/E6KCo7BSpEiR6N9z5861li1bJrlP27ZtbebMmfbLL7/YypUrk+zPmzevXX311b4fAAAg7gFnypQpXql5/fXXo9vWrVtnFStWTHI//b169Wrfp343JUqUSHE/AABAXFcT37hxo9155502bNgwO+OMM6Lb1acmd+7cSe5bsGBB27t3b4r7Yven5uDBg36LLXEBAIBwimsFp3379nbhhRfaXXfdlWR7rly57MCBA0m27dixw4NNSvti96dm4MCB3mYX3EqXLp2G7wQAACSSuAWc5557zv7973/7cPDs2bP7Tc4++2wrVqzYUc1Nq1atsvLly1uZMmW8U/GWLVtS3J+aPn36eIek4LZ+/fp0emcAACDTBpzGjRvbN998Y0uWLLGvvvrKb0GfHO2bNm1akvtPmDDBGjZs6OFHo6hi96vZSn9rf2py5Mjhva1jbwAAIJzi1genQoUKKW6vVKmSde7c2c4//3wfOt6gQQObPHmyLViwwOe/ke7du/ukgAUKFPCmpkcffdQ7GR/rEHEAABBuce1knBrNYKxQoyDTt29fO+ecc+yDDz6wUqVK+f7bb7/dm6g0D446C6vi884771jWrHEfFAYAABJAQkz0Fw9M9Jf5MNEfAJz4TqiJ/gAAANISAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIROugScI0eOpMfTAgAApF/A6d69uw0fPjzV/X379rVu3br9lacGAACIT8CJRCL2wAMP2Nlnn21NmjSxnj172jvvvGMHDhywjRs32ujRo61w4cLHf3QAAADpHXC++uqr/3tQ1qzWo0cPmzBhgnXs2NFOOeUUe+yxx6xMmTLWsGFDu/jii61Pnz5/5XgAAACOW/Y/c+fmzZtblixZLFeuXHbllVfamWeeabt27bJs2bJZ8eLFbdWqVd7/pkSJEn4/AACAhK/gqBlq2LBhXqV5++23rXLlyjZlyhTLnTu33XPPPfbjjz/aggUL7OOPP7Y33ngj/Y4aAAAgrQJOpUqVrGnTph5yunTpYkOHDvWAU6xYMZs1a5ZXdvLnz2+9evWy3r172+HDh//M0wMAAGR8wPnwww+tWrVq9sUXX9ioUaMsR44cHm5uuOEGGzFihHXo0MGefvppe/XVV70fzr59+9LmKAEAANIr4DRo0MCbnlSlyZkzp61fv97efPNNy549u51++ul23nnn2cCBA/2ntufLl+/PPD0AAEDGB5xTTz3VrrnmGmvUqJGtWLHCHnzwQatatar985//9H44quCos/GkSZN8PwAAQEIHHM19M2fOHG+mmjdvnldtateu7X/v37/fli9fbnfeeadXd7p27epNVgAAAAkdcDTse9u2bd4HZ9myZVa6dGlr3bq1dyh+6KGHrFSpUnbhhRdas2bNvClLI64AAAASfh6cOnXq2Lp163xiP811o4CjDsfjxo2zIkWKWPv27f1+u3fvtrx589rBgwe9IzIAAEBGyhJR29Nx2LJli3cmPvnkk6PbDh065Ms2JHInY01QqOa0nTt3JvRxlu39XrwPITTWDmoa70MAAGTQ9fu4VxNX5SY23EjdunVt5syZx/vUAAAAf8lxB5zkxo8fb4sWLbKff/45rZ8aAAAg4wOO+ufce++9dtlll1mnTp3S8qkBAAAyPuD88MMPVr9+fZ8HZ+rUqWn1tAAAABkfcNTJ+JFHHrFzzz3XF9/UXDnq/AMAAHBCDBM/++yz7YILLvA5b44cOWKffvqprVy50i699FJfdLNx48bpd6QAAABpHXA0p83evXvts88+s48//th++ukn36bJ/erVq+chBwAA4IRqotKEfWvWrLGlS5faqlWrfKVwLc9w3XXX2eOPP+7NU2+//Xb6Hi0AAEB69sHR0g1VqlTxdae0dMMtt9ziMxt37Njxrz4lAABAxvfBSY0m+nv00Uft4osvtpYtW3rT1ZgxY9LiqQEAAOI7D86VV15pkydPtpdfftleffXVtHxqAACA+M1krJDToUMHmz9/flo/NQAAQMY1USWnTscnnXRSejw1AABAxldwJE+ePEctwJmSjRs3WqtWrXxiwHLlytmzzz4b3af1rGrVqmU5c+a0qlWr2vTp05M8duzYsVa+fHnLlSuXXX755fb999+nx1sBAAAnoHQJOMeqRYsWtmfPHl95fNiwYdavXz977bXXbPPmzdakSROrU6eOLViwwNq1a+edlxcvXuyP0/3VDNarVy9vCtPEg40aNfJ5egAAANKliepYqHpToUIFe/rpp61QoUJWu3ZtX+Zh0qRJtn79eqtYsaINHz7c71ujRg0fij5w4EB76623bOjQoda5c2cPOaJOzZqHR1Wde+65J15vCQAAZPYKTokSJXyklcLNoUOH7JNPPvERWEWLFrW5c+d6dSdW27ZtvXITiUS8aqOKTiBbtmw+B4/2AwAAxLWJKqD1rbTcg4KKmqnWrVvnFZxY+nvHjh2+uKeaolLav3r16gw+cgAAkIgSIuCMGzfOXnzxRe+YPHXqVNu/f7/lzp07yX0KFizoPw8cOOA/U9r/e31wNPngrl27ktwAAEA4xa0PTqxq1ar57bLLLvMRU6rGBEEmoOpNsCaWaH9syNH+5KEnlvrv9O/fP93eAwAASBxxq+AsWbLEVyaPpU7Hp5xyivfDSd7cpAU+NZxc+zQMPaX9Gjaemj59+tjOnTujN3VkBgAA4RS3gPPtt996x+DDhw9Ht61cudK2bt3qQ76nTZuW5P4TJkywhg0b+iKfl1xySZL9R44csYkTJ/r+1Kjyky9fviQ3AAAQTnELOM2bN/fZjjXHjSb1e/fdd+3qq6/2v++44w5bunSp9ezZ07766isfLq5+OqrCSLdu3XzeHA0P19w4Gi6u+XT0WAAAgLgFHM1ArNmJt23b5hUbzWujoeHPP/+8N0PNmDHDZs+e7SOsRo0a5RWamjVr+mN1/xEjRtgjjzzikwGquWrWrFnedAUAAJAloollMiGNolKfHvXHSeTmqrK934v3IYTG2kFN430IAIAMun4nxDBxAACAtETAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoUPAAQAAoRPXgLNjxw675ZZbLF++fFaqVCnr0aOHHTx40PctWrTIatWqZTlz5rSqVava9OnTkzx27NixVr58ecuVK5ddfvnl9v3338fpXQAAgEQT14Bz88032w8//GDvvvuujR492t555x3r0qWLbd682Zo0aWJ16tSxBQsWWLt27axly5a2ePFif9zMmTOtQ4cO1qtXL5s/f76Ho0aNGtnevXvj+XYAAECCyB6vF162bJkHlbVr11qJEiV826mnnmq1a9f2vytWrGjDhw/37TVq1PD7Dxw40N566y0bOnSode7c2UOOvPzyy1a5cmWv6txzzz3xeksAACCzV3C2bt1qN954YzTcyBlnnGGHDx+2jz/+2Fq0aJHk/m3btvVAFIlEvGqjik4gW7Zs1rp1a98PAAAQtwpO3bp1/RZLTVWq4mzcuNErOLH0t/rsbNmyxZuiUto/bdq0VF9PfXuC/j2ya9euNHsvAAAgsSTMKKoNGzbYvffeaw888IAdOnTIcufOnWR/wYIF/eeBAwf8Z0r7f68Pjpq38ufPH72VLl06Xd4HAACIv4QIOHv27LFmzZp5p+K7777bR0YFQSag6o3kyJHDf6a0P3noidWnTx/buXNn9LZ+/fp0eS8AACATN1EFfvvtN7v++uvtyJEj9uqrr1qWLFmsTJkytnr16iT3W7VqlVdeihYtanny5PH9hQoVSrJfw8ZTo2AUhCMAABBuca/gdOzY0b7++mt77733LG/evL5NfXOS96eZMGGCNWzY0APQJZdckmS/wtHEiRN9PwAAQFwrOIMHD7YxY8bYa6+95s1U3377rW9v06aNDRkyxHr27Gk33XSTzZkzx8aNG2dz5871/d26dbNrrrnGypUrZ9WrV7eRI0f64zVfDgAAQNwCjuazUb8YDfvWEO9YU6ZMsRkzZvicNk8//bSPkFKFpmbNmr5fk/qNGDHC+vfvbz///LNdfPHFNmvWLG+6AgAAyBJRwsiENExcfXrU4VhLRSSqsr3fi/chhMbaQU3jfQgAgAy6fse9Dw4AAEBaI+AAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQIeAAAIDQietq4gBOQP3yx/sIwqPfzngfARBaVHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoEHAAAEDoxD3gNGzY0LJkyWI33nhjku2LFi2yWrVqWc6cOa1q1ao2ffr0JPvHjh1r5cuXt1y5ctnll19u33//fQYfOQAASFRxDTiHDx+2Z5991jp27Jhk++bNm61JkyZWp04dW7BggbVr185atmxpixcv9v0zZ860Dh06WK9evWz+/PlWqlQpa9Soke3duzdO7wQAACSSuAac7NmzW5UqVaxIkSJJto8ZM8YqVqxow4cPtxo1alj37t29wjNw4EDfP3ToUOvcubOHnPPOO89efvllfy5VdQAAAOLeRJWSuXPnWosWLZJsa9u2rVduIpGIV21U0Qlky5bNWrdu7fsBAAASMuCsW7fOKzix9PeOHTtsy5Yt3hSV0v7Vq1dn8JECAIBElN0S0P79+y137txJthUsWNB/HjhwwH+mtP/3+uAcPHjQb4Fdu3al8VEDAIBEkZAVHI2MCoJMQNUbyZEjh/9MaX/y0BNL/Xfy588fvZUuXTpdjh0AAMRfQgacMmXKHNXctGrVKg8mRYsWtTx58qS4X8PGU9OnTx/buXNn9LZ+/fp0O34AABBfCRlw6tata9OmTUuybcKECdE5cy655JIk+48cOWITJ070/alR5SdfvnxJbgAAIJwSsg/ObbfdZkOGDLGePXvaTTfdZHPmzLFx48b56Crp1q2bXXPNNVauXDmrXr26jRw50vbs2ePz5QAAACRkBUfNUDNmzLDZs2fbBRdcYKNGjfIKTc2aNX2/JvUbMWKEPfLIIz4ZoJqrZs2a5U1XAAAAWSKaWCYT0igq9elRf5xEbq4q2/u9eB9CaKwd1DTehxAO/fLH+wjCo9/OeB8BENrrd0JWcAAAAI4HAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIQOAQcAAIROQs5kDADAn8L8TGmjX3jmZiLgAPhTyh54I96HEBpr430AQIjRRAUAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEKHgAMAAEIne7wPAACA41X2wBvxPoRQWGvhQQUHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEDgEHAACEzgkbcDZt2mQtWrSwPHnyWMmSJW3w4MHxPiQAAJAgstsJqlWrVnbSSSfZhx9+aD/99JN16NDBTjnlFOvYsWO8Dw0AAMTZCRlwFixYYF9++aWtW7fOihQp4tt2795tvXv3trvvvtuyZj1hC1MAACANnJBJYO7cuVa3bt1ouAkqOtu3b7dvvvkmrscGAADi74QMOKrcVKxYMck29cU59dRTbfXq1XE7LgAAkBhOyCaq/fv3W+HChY/aXrBgQdu7d2+Kjzl48KDfAjt37vSfu3btskR25OC+eB9CaCT6Z32i4DuZdvhOph2+l5nnO7nrf8cYiUTCF3By5cplBw4cOGr7jh07LHfu3Ck+ZuDAgda/f/+jtpcuXTpdjhGJJ/9T8T4CICm+k0g0+U+g76T63ubPnz9cAadMmTI2Z86cJNv27dtnP//8s5UvXz7Fx/Tp08fuv//+6N9Hjhyxbdu2eSUoS5Ys6X7MYaUkrZC4fv16y5cvX7wPB3B8L5Fo+E6mHVVuFG5OO+20373fCRlw1MF4wIAB3swUpLdp06b5l+acc85J8TE5cuTwW6wCBQpkyPFmBjr3/J8WiYbvJRIN38m08XuVmxO6k/GFF15o1atXtxtvvNE+++wz+9e//mVdu3a1Bx54gCHiAADgxAw4MmnSJK/I1K9f3+e+6dKli3Xu3DnehwUAABLACdlEJcWLF7epU6fG+zAyPYXMhx9++KjmPyCe+F4i0fCdzHhZIn80zgoAAOAEc8I2UQEAAKSGgAMAAEKHgAMAAEKHgAMAAEKHgIPjovW9Ulv/C0hvW7ZssdmzZ6e4b968ebZhw4YMPyYgsGfPHvvggw9s7NixNn369OgaiAH+25m+CDj4Q1rj66qrrrIffvghuu3XX3+1Tp06+YycmlGyVatWJ8QibQgPrUdXr149GzFiRIr7X3zxRd+vZVyAjPTbb79Z37597dRTT7UOHTrY6NGjrWPHjlaiRAnr0aOHfyd79+5tTz75ZLwPNdQIOPhDd911l23dujW6tIXW8br22mvtrbfesldffdUWLVpkGzdutPvuuy/eh4pMRBcHBe3XX389xf2jRo3yhXmHDRuW4ceGzK1Fixb29ttv28yZM23NmjU2f/58/6lq47vvvutBR/tbtmwZ70MNtRN2oj9knPfee8/ef/99K1KkiP/drVs3mzt3rn366adWtWpV3zZo0CBr3rx5nI8UmYkCtv41nDNnzhT3n3zyyda9e3d7/PHH7cEHH8zw40Pm9Morr9iXX35pixcvtqJFix5V2VGlW9VH/YPwzDPPjNtxZgZUcPCHSpYsacuXL/ffn3nmGb+NHz8+Gm5EfR1YvBQZSU2mNWvW/N37nHvuuf4vZyCjvPDCC9anT5+jws3IkSOtSZMm1qtXL183UUEI6YsKDv7Q008/bTfccINXbjTx9UsvvWRXXnlldP/kyZO9GWvIkCFxPU5kLqrQqJP771Ffh2zZsmXYMQHLli2ziy+++Kjtqup8+OGHdsEFF3jH46FDh8bl+DITAg7+0BVXXGHr1q2zlStXWqVKlY5apv6rr77ycHPHHXfE7RiR+ag6M2PGDLvwwgtTvc+sWbPsnHPOydDjQuamfwSmFLxV2cma9f8aTdR3LHt2Lr/pjbWoAJyQJkyYYHfeeactWLDAzjrrrKP2q2nq/PPP907Gt956a1yOEZnPZZdd5qP3+vfvn+p9NKJKTazq24j0Qx8cACckNZs2bdrU6tat66P5gn81qyPnlClT7NJLL/V9hBtkpLvvvtueeOIJb5JKbX4mTWHAqNP0RwUHwAlL//l65JFHfMi4Ao5G+m3bts2yZMlinTt3tgEDBtAUgAyn5noNxNDovRtvvNEHamgqjTfeeMMrO5obh+kL0h8BB8AJT52JP/vsM9u0aZMVLlzY6tSpY6ecckq8DwuZ2PPPP2+PPfZYktm0FXT69etn7du3j+uxZRYEHAAA0on62mzevNmD9xlnnBHvw8lUCDgAACB06GQMAABCh4ADAABCh4ADAABCh/GTADKMZnA96aSTkmzT6vQfffSRr93z008/+TDvm266Kcl9Dh8+bDt37vSOmn+VuhtqjhwNG9dxaNiuVnVOfjwAwoEKDoAM065dOw8YBQsW9LCSI0cOe+6553yNHk2QpvlDNJ29Qk5sMFm0aJHPVqzRKIGBAwdaq1at/Het/aPJ02J16dLFVxLXEHKFKN30ej/++KPt3r3bypQp49uC1wgoBB3L2AuFLgCJi4ADIMNogcE9e/bY9u3bbdWqVZYnTx6fCK1169b26aef+sKtWqW+SpUqNn36dH+MVl9u3LixV3C0r3z58tHnCyo6OXPmtOLFiyd5rfvvv9+mTp3qz6v75cuXzwON1qYKnkPbFai+++676OM0A622aTFPPW/ym7Zr/9tvv51BZw3AX0ETFYAMoyahhg0bWp8+fWzatGkebBQydLvuuuvs0UcftSVLlvgEaU2aNPHHKFB0797dw021atWsWbNmvl2VGM1cLJq5ODlVaObOnethRHORKODkzZvXg5WCin7fsWOH7dq1y38P5M6d23/Onj3bSpUqleL7KFeunB8XgMRFwAGQobR+lJqW1EykkPPNN9/4dq0Z1aJFCw86asoKVl4OaGr7MWPG+O+HDh3ylcRHjBgR3a8KjypDBQoUsPXr1/vz6rV039T67qi5rFChQn6f5BRuypYtm8bvHkBGIeAAyFAXXXSR94vRcgpt2rTxaooqMFpLSvvUH0c/u3XrZldffXX0cVq/Z+/evf679qkfz5VXXul/q3+OAtLpp5/u1ZpZs2bZXXfd5Qtyanp8LX6oyoxeo0KFCtGKj8KQwk3Lli19wU4AIaKZjAEgvR06dCgybNiwSKFChSLjx4+PHDly5Kj7/Pbbb5HvvvsuctFFF0Xy5s0bGTp0aCRHjhyRPHnyRIoXLx4pUqRIpEKFCpGNGzdGvv766ySPPXDgQGT//v3Rv7/55pvIQw89FNm9e3dkz549kffff98fq9/12meccUbko48+iuzduzeyY8eOJI/TfxrXrFmT6nvR/ilTpqTZuQGQ9qjgAMgQ//3vf+1f//qXffzxx/bUU095RSb5gphqZrrtttu878zChQu9342am9QZWdUYVWDUb0fNW+osrGasoBqj0U/z5s3zEVVy9tln+03mz59v119/vTeJ6bmCUViqIE2aNMmrSQDChYADIEOoY+4nn3ziv2u4tlZV1lDuWL179/Yh2gouaqaSoEPw7bffbuPGjfNtCin58+f3uWxi+8wE4UXUgXj06NHelNWzZ08PQjfffLM/TqOpFJwUhurXr2/Nmze3CRMmpNhZGcCJiWHiADKc+tIo4KjPjPrEaNi2flen4eQdfjWsPFeuXN4Z+JdffvFtQRCJHb6dnCpFb775po+mev/9961y5co2duxY+/rrr23p0qV27bXX2i233GIrVqywf/zjH4QbIGSo4ADIcC+99JINGDDAOnXqFB3qrflq1NF3yJAhSe6r4eCXXHKJNykdywR8AYUbdUK+4oorots0QkvVo6BJq1GjRh6AsmXLFr1P8BqqOAE4cVHBAZBhtmzZ4vPOKESoiUijmwKvvPKKNzNp6LYqPJrNWE1QWsZBfWQ0Gmr58uXRSov66+j34LZ169ZoOFm9erVNmTLFLr/88iSvr9dQXyDd1q1b55UhjcCKrRrt37/ffy5YsMCPMaWbHDhwIEPOGYC/hgoOgAyjodyaoViT5GkNKFVORo0a5fsUTtQ3ZtCgQR4y2rZt6xP8qQqj5isFC82F06NHD18mQX1pFJhiJxFUeArCjyYFvPDCC6P7zzvvPJ9nJ1ieQaFI8+PoNWIn7dOaWA8//LDPpqxh5ClRpen8889Pt/ME4Phl0VCqNHgeAPhDGgWlMHEs/V1UVdF91ZQU24Qky5Yt8w7C6kicGgWZ5JMFBtuDqg+A8CLgAACA0KEPDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAACB0CDgAAsLD5f5bR9ckIslmWAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "fig.set(alpha=0.2)  # 设定图表颜色alpha参数\n",
    "\n",
    "Survived_0 = data_train.Embarked[data_train.Survived == 0].value_counts()\n",
    "Survived_1 = data_train.Embarked[data_train.Survived == 1].value_counts()\n",
    "df=pd.DataFrame({u'未获救':Survived_0,u'获救':Survived_1})\n",
    "df.plot(kind='bar', stacked=True)\n",
    "plt.title(u\"各登船港口乘客的获救情况\")\n",
    "plt.xlabel(u\"登船港口\") \n",
    "plt.ylabel(u\"人数\") \n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "908B60F9DD894D929DA6D5728CD1929A",
    "mdEditEnable": false
   },
   "source": [
    "#### 看看 堂兄弟/妹，孩子/父母有几人，对是否获救的影响"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": false,
    "id": "D5C98FE13EBB435F85FAF4F21938EA2C",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                PassengerId\n",
      "SibSp Survived             \n",
      "0     0                 398\n",
      "      1                 210\n",
      "1     0                  97\n",
      "      1                 112\n",
      "2     0                  15\n",
      "      1                  13\n",
      "3     0                  12\n",
      "      1                   4\n",
      "4     0                  15\n",
      "      1                   3\n",
      "5     0                   5\n",
      "8     0                   7\n",
      "                PassengerId\n",
      "Parch Survived             \n",
      "0     0                 445\n",
      "      1                 233\n",
      "1     0                  53\n",
      "      1                  65\n",
      "2     0                  40\n",
      "      1                  40\n",
      "3     0                   2\n",
      "      1                   3\n",
      "4     0                   4\n",
      "5     0                   4\n",
      "      1                   1\n",
      "6     0                   1\n"
     ]
    }
   ],
   "source": [
    "gg = data_train.groupby(['SibSp','Survived'])\n",
    "df = pd.DataFrame(gg.count()['PassengerId'])\n",
    "print(df)\n",
    "\n",
    "gp = data_train.groupby(['Parch','Survived'])\n",
    "df = pd.DataFrame(gp.count()['PassengerId'])\n",
    "print(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "630CA44341D84F3D85B265A347016F96",
    "mdEditEnable": false
   },
   "source": [
    "好吧，没看出特别特别明显的规律(为自己的智商感到捉急…)，先作为备选特征，放一放。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "18CB6E8DC7194E9490C41F27B755EBB7",
    "mdEditEnable": false
   },
   "source": [
    "#### tickets cabin的分析\n",
    "ticket是船票编号，应该是unique的，和最后的结果没有太大的关系，先不纳入考虑的特征范畴\n",
    "\n",
    "cabin只有204个乘客有值，我们先看看它的一个分布"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": false,
    "id": "52EFF0E17D9141F28A47633B2969D050",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Cabin\n",
       "B96 B98        4\n",
       "G6             4\n",
       "C23 C25 C27    4\n",
       "C22 C26        3\n",
       "F33            3\n",
       "              ..\n",
       "E34            1\n",
       "C7             1\n",
       "C54            1\n",
       "E36            1\n",
       "C148           1\n",
       "Name: count, Length: 147, dtype: int64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.Cabin.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AE065A86E10C46A78140AE25C0C9E628",
    "mdEditEnable": false
   },
   "source": [
    "这三三两两的…如此不集中…我们猜一下，也许，前面的ABCDE是指的甲板位置、然后编号是房间号？…好吧，我瞎说的，别当真…\n",
    "\n",
    "关键是Cabin这鬼属性，应该算作类目型的，本来缺失值就多，还如此不集中，注定是个棘手货…第一感觉，这玩意儿如果直接按照类目特征处理的话，太散了，估计每个因子化后的特征都拿不到什么权重。加上有那么多缺失值，要不我们先把Cabin缺失与否作为条件(虽然这部分信息缺失可能并非未登记，maybe只是丢失了而已，所以这样做未必妥当)，先在有无Cabin信息这个粗粒度上看看Survived的情况好了。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false,
    "id": "C46525090059436083CF94D6599456E0",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "fig.set(alpha=0.2)  # 设定图表颜色alpha参数\n",
    "\n",
    "Survived_cabin = data_train.Survived[pd.notnull(data_train.Cabin)].value_counts()\n",
    "Survived_nocabin = data_train.Survived[pd.isnull(data_train.Cabin)].value_counts()\n",
    "df=pd.DataFrame({u'有':Survived_cabin, u'无':Survived_nocabin}).transpose()\n",
    "df.plot(kind='bar', stacked=True)\n",
    "plt.title(u\"按Cabin有无看获救情况\")\n",
    "plt.xlabel(u\"Cabin有无\") \n",
    "plt.ylabel(u\"人数\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "C0F627D831324F4A8E02278E49A895B6",
    "mdEditEnable": false
   },
   "source": [
    "有Cabin记录的似乎获救概率稍高一些，先这么着放一放吧。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2BED89CF8A7D45E6832F5D54E84CAF75",
    "mdEditEnable": false
   },
   "source": [
    "## 简单数据预处理\n",
    "大体数据的情况看了一遍，对感兴趣的属性也有个大概的了解了。 \n",
    "下一步干啥？咱们该处理处理这些数据，为机器学习建模做点准备了。\n",
    "\n",
    "对了，我这里说的数据预处理，其实就包括了很多`Kaggler`津津乐道的`feature engineering`过程，灰常灰常有必要！\n",
    "\n",
    "- **『特征工程(feature engineering)太重要了！』**\n",
    "- **『特征工程(feature engineering)太重要了！』**\n",
    "- **『特征工程(feature engineering)太重要了！』**\n",
    "\n",
    "恩，重要的事情说三遍。\n",
    "\n",
    "先从最突出的数据属性开始吧，对，`Cabin`和`Age`，有丢失数据实在是对下一步工作影响太大。\n",
    "\n",
    "先说`Cabin`，暂时我们就按照刚才说的，按`Cabin`有无数据，将这个属性处理成`Yes`和`No`两种类型吧。\n",
    "\n",
    "再说`Age`：\n",
    "\n",
    "通常遇到缺值的情况，我们会有几种常见的处理方式\n",
    "\n",
    "* 如果缺值的样本占总数比例极高，我们可能就直接舍弃了，作为特征加入的话，可能反倒带入`noise`，影响最后的结果了;\n",
    "* 如果缺值的样本适中，而该属性非连续值特征属性(比如说类目属性)，那就把`NaN`作为一个新类别，加到类别特征中;\n",
    "* 如果缺值的样本适中，而该属性为连续值特征属性，有时候我们会考虑给定一个`step`(比如这里的`age`，我们可以考虑每隔2/3岁为一个步长)，然后把它离散化，之后把`NaN`作为一个`type`加到属性类目中;\n",
    "* 有些情况下，缺失的值个数并不是特别多，那我们也可以试着根据已有的值，拟合一下数据，补充上。\n",
    "\n",
    "本例中，后两种处理方式应该都是可行的，我们先试试拟合补全吧(虽然说没有特别多的背景可供我们拟合，这不一定是一个多么好的选择)\n",
    "\n",
    "我们这里用`scikit-learn`中的`RandomForest`来拟合一下缺失的年龄数据(注：`RandomForest`是一个用在原始数据中做不同采样，建立多颗`DecisionTree`，再进行`average`等等来降低过拟合现象，提高结果的机器学习算法，我们之后会介绍到)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": false,
    "id": "77DB5AD318FB481C8F098E7DC145348D",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "### 使用 RandomForestClassifier 填补缺失的年龄属性\n",
    "def set_missing_ages(df):\n",
    "\n",
    "    # 把已有的数值型特征取出来丢进Random Forest Regressor中\n",
    "    age_df = df[['Age','Fare', 'Parch', 'SibSp', 'Pclass']]\n",
    "\n",
    "    # 乘客分成已知年龄和未知年龄两部分\n",
    "    known_age = age_df[age_df.Age.notnull()].values\n",
    "    unknown_age = age_df[age_df.Age.isnull()].values\n",
    "\n",
    "    # y即目标年龄\n",
    "    y = known_age[:, 0]\n",
    "\n",
    "    # X即特征属性值\n",
    "    X = known_age[:, 1:]\n",
    "\n",
    "    # fit到RandomForestRegressor之中\n",
    "    rfr = RandomForestRegressor(random_state=0, n_estimators=2000, n_jobs=-1)\n",
    "    rfr.fit(X, y)\n",
    "\n",
    "    # 用得到的模型进行未知年龄结果预测\n",
    "    predictedAges = rfr.predict(unknown_age[:, 1::])\n",
    "\n",
    "    # 用得到的预测结果填补原缺失数据\n",
    "    df.loc[ (df.Age.isnull()), 'Age' ] = predictedAges \n",
    "\n",
    "    return df, rfr\n",
    "\n",
    "def set_Cabin_type(df):\n",
    "    df.loc[ (df.Cabin.notnull()), 'Cabin' ] = \"Yes\"\n",
    "    df.loc[ (df.Cabin.isnull()), 'Cabin' ] = \"No\"\n",
    "    return df\n",
    "\n",
    "data_train, rfr = set_missing_ages(data_train)\n",
    "data_train = set_Cabin_type(data_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": false,
    "id": "5036CC74FAE84F658FB4EF8FB8998E2A",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</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>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>No</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>Yes</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>No</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>Yes</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>No</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Moran, Mr. James</td>\n",
       "      <td>male</td>\n",
       "      <td>23.838953</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>330877</td>\n",
       "      <td>8.4583</td>\n",
       "      <td>No</td>\n",
       "      <td>Q</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>7</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>McCarthy, Mr. Timothy J</td>\n",
       "      <td>male</td>\n",
       "      <td>54.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>17463</td>\n",
       "      <td>51.8625</td>\n",
       "      <td>Yes</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>8</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Palsson, Master. Gosta Leonard</td>\n",
       "      <td>male</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>349909</td>\n",
       "      <td>21.0750</td>\n",
       "      <td>No</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>9</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Johnson, Mrs. Oscar W (Elisabeth Vilhelmina Berg)</td>\n",
       "      <td>female</td>\n",
       "      <td>27.000000</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>347742</td>\n",
       "      <td>11.1333</td>\n",
       "      <td>No</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>10</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>Nasser, Mrs. Nicholas (Adele Achem)</td>\n",
       "      <td>female</td>\n",
       "      <td>14.000000</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>237736</td>\n",
       "      <td>30.0708</td>\n",
       "      <td>No</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "5            6         0       3   \n",
       "6            7         0       1   \n",
       "7            8         0       3   \n",
       "8            9         1       3   \n",
       "9           10         1       2   \n",
       "\n",
       "                                                Name     Sex        Age  \\\n",
       "0                            Braund, Mr. Owen Harris    male  22.000000   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.000000   \n",
       "2                             Heikkinen, Miss. Laina  female  26.000000   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.000000   \n",
       "4                           Allen, Mr. William Henry    male  35.000000   \n",
       "5                                   Moran, Mr. James    male  23.838953   \n",
       "6                            McCarthy, Mr. Timothy J    male  54.000000   \n",
       "7                     Palsson, Master. Gosta Leonard    male   2.000000   \n",
       "8  Johnson, Mrs. Oscar W (Elisabeth Vilhelmina Berg)  female  27.000000   \n",
       "9                Nasser, Mrs. Nicholas (Adele Achem)  female  14.000000   \n",
       "\n",
       "   SibSp  Parch            Ticket     Fare Cabin Embarked  \n",
       "0      1      0         A/5 21171   7.2500    No        S  \n",
       "1      1      0          PC 17599  71.2833   Yes        C  \n",
       "2      0      0  STON/O2. 3101282   7.9250    No        S  \n",
       "3      1      0            113803  53.1000   Yes        S  \n",
       "4      0      0            373450   8.0500    No        S  \n",
       "5      0      0            330877   8.4583    No        Q  \n",
       "6      0      0             17463  51.8625   Yes        S  \n",
       "7      3      1            349909  21.0750    No        S  \n",
       "8      0      2            347742  11.1333    No        S  \n",
       "9      1      0            237736  30.0708    No        C  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data_train.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "954507CD225B48F19CC9BC688024355A",
    "mdEditEnable": false
   },
   "source": [
    "恩。目的达到，`OK`了。\n",
    "\n",
    "因为逻辑回归建模时，需要输入的特征都是数值型特征，我们通常会先对类目型的特征因子化。 \n",
    "什么叫做因子化呢？举个例子：\n",
    "\n",
    "以`Cabin`为例，原本一个属性维度，因为其取值可以是\\[‘yes’,’no’\\]，而将其平展开为`Cabin_yes`,`Cabin_no`两个属性\n",
    "\n",
    "原本`Cabin`取值为`yes`的，在此处的”`Cabin_yes`”下取值为1，在”`Cabin_no`”下取值为0\n",
    "原本`Cabin`取值为`no`的，在此处的”`Cabin_yes`”下取值为0，在”`Cabin_no`”下取值为1\n",
    "我们使用`pandas`的”`get_dummies`”来完成这个工作，并拼接在原来的”`data_train`”之上，如下所示:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": false,
    "id": "10DE8C414A214AD9842CA4C772644829",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin_No</th>\n",
       "      <th>Cabin_Yes</th>\n",
       "      <th>Embarked_C</th>\n",
       "      <th>Embarked_Q</th>\n",
       "      <th>Embarked_S</th>\n",
       "      <th>Sex_female</th>\n",
       "      <th>Sex_male</th>\n",
       "      <th>Pclass_1</th>\n",
       "      <th>Pclass_2</th>\n",
       "      <th>Pclass_3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived   Age  SibSp  Parch     Fare  Cabin_No  Cabin_Yes  \\\n",
       "0            1         0  22.0      1      0   7.2500      True      False   \n",
       "1            2         1  38.0      1      0  71.2833     False       True   \n",
       "2            3         1  26.0      0      0   7.9250      True      False   \n",
       "3            4         1  35.0      1      0  53.1000     False       True   \n",
       "4            5         0  35.0      0      0   8.0500      True      False   \n",
       "\n",
       "   Embarked_C  Embarked_Q  Embarked_S  Sex_female  Sex_male  Pclass_1  \\\n",
       "0       False       False        True       False      True     False   \n",
       "1        True       False       False        True     False      True   \n",
       "2       False       False        True        True     False     False   \n",
       "3       False       False        True        True     False      True   \n",
       "4       False       False        True       False      True     False   \n",
       "\n",
       "   Pclass_2  Pclass_3  \n",
       "0     False      True  \n",
       "1     False     False  \n",
       "2     False      True  \n",
       "3     False     False  \n",
       "4     False      True  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dummies_Cabin = pd.get_dummies(data_train['Cabin'], prefix= 'Cabin')\n",
    "\n",
    "dummies_Embarked = pd.get_dummies(data_train['Embarked'], prefix= 'Embarked')\n",
    "\n",
    "dummies_Sex = pd.get_dummies(data_train['Sex'], prefix= 'Sex')\n",
    "\n",
    "dummies_Pclass = pd.get_dummies(data_train['Pclass'], prefix= 'Pclass')\n",
    "\n",
    "df = pd.concat([data_train, dummies_Cabin, dummies_Embarked, dummies_Sex, dummies_Pclass], axis=1)\n",
    "df.drop(['Pclass', 'Name', 'Sex', 'Ticket', 'Cabin', 'Embarked'], axis=1, inplace=True)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DFF9132E01D14C268CD38CB8B1C56419",
    "mdEditEnable": false
   },
   "source": [
    "`bingo`，我们很成功地把这些类目属性全都转成`0`，`1`的数值属性了。\n",
    "\n",
    "这样，看起来，是不是我们需要的属性值都有了，且它们都是数值型属性呢。\n",
    "\n",
    "有一种临近结果的宠宠欲动感吧，莫急莫急，我们还得做一些处理，仔细看看`Age`和`Fare`两个属性，乘客的数值幅度变化，也忒大了吧！！如果大家了解逻辑回归与梯度下降的话，会知道，各属性值之间scale差距太大，将对收敛速度造成几万点伤害值！甚至不收敛！ (╬▔皿▔)…所以我们先用`scikit-learn`里面的`preprocessing`模块对这俩货做一个`scaling`，所谓`scaling`，其实就是将一些变化幅度较大的特征化到\\[-1,1\\]之内。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": false,
    "id": "FBA4537B8FDB444DA85DF162F8A28D95",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin_No</th>\n",
       "      <th>Cabin_Yes</th>\n",
       "      <th>Embarked_C</th>\n",
       "      <th>Embarked_Q</th>\n",
       "      <th>Embarked_S</th>\n",
       "      <th>Sex_female</th>\n",
       "      <th>Sex_male</th>\n",
       "      <th>Pclass_1</th>\n",
       "      <th>Pclass_2</th>\n",
       "      <th>Pclass_3</th>\n",
       "      <th>Age_scaled</th>\n",
       "      <th>Fare_scaled</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>-0.561377</td>\n",
       "      <td>-0.502445</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.613173</td>\n",
       "      <td>0.786845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>-0.267740</td>\n",
       "      <td>-0.488854</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.392945</td>\n",
       "      <td>0.420730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>0.392945</td>\n",
       "      <td>-0.486337</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived   Age  SibSp  Parch     Fare  Cabin_No  Cabin_Yes  \\\n",
       "0            1         0  22.0      1      0   7.2500      True      False   \n",
       "1            2         1  38.0      1      0  71.2833     False       True   \n",
       "2            3         1  26.0      0      0   7.9250      True      False   \n",
       "3            4         1  35.0      1      0  53.1000     False       True   \n",
       "4            5         0  35.0      0      0   8.0500      True      False   \n",
       "\n",
       "   Embarked_C  Embarked_Q  Embarked_S  Sex_female  Sex_male  Pclass_1  \\\n",
       "0       False       False        True       False      True     False   \n",
       "1        True       False       False        True     False      True   \n",
       "2       False       False        True        True     False     False   \n",
       "3       False       False        True        True     False      True   \n",
       "4       False       False        True       False      True     False   \n",
       "\n",
       "   Pclass_2  Pclass_3  Age_scaled  Fare_scaled  \n",
       "0     False      True   -0.561377    -0.502445  \n",
       "1     False     False    0.613173     0.786845  \n",
       "2     False      True   -0.267740    -0.488854  \n",
       "3     False     False    0.392945     0.420730  \n",
       "4     False      True    0.392945    -0.486337  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 导入sklearn.preprocessing库中的StandardScaler类，用于数据标准化\n",
    "import sklearn.preprocessing as preprocessing\n",
    "\n",
    "# 创建StandardScaler对象，用于将数据标准化（均值为0，标准差为1）\n",
    "scaler = preprocessing.StandardScaler()\n",
    "\n",
    "# 对'Age'列进行标准化处理\n",
    "# 使用fit方法计算'Age'列的均值和标准差，并返回标准化的参数\n",
    "age_scale_param = scaler.fit(df['Age'].values.reshape(-1,1))\n",
    "\n",
    "# 使用fit_transform方法将'Age'列进行标准化，并将结果存入新的'Age_scaled'列\n",
    "# 注意：这里调用了标准化参数age_scale_param\n",
    "df['Age_scaled'] = scaler.fit_transform(df['Age'].values.reshape(-1,1), age_scale_param)\n",
    "\n",
    "# 对'Fare'列进行标准化处理\n",
    "# 使用fit方法计算'Fare'列的均值和标准差，并返回标准化的参数\n",
    "fare_scale_param = scaler.fit(df['Fare'].values.reshape(-1,1))\n",
    "\n",
    "# 使用fit_transform方法将'Fare'列进行标准化，并将结果存入新的'Fare_scaled'列\n",
    "# 同样地，调用了fare_scale_param参数\n",
    "df['Fare_scaled'] = scaler.fit_transform(df['Fare'].values.reshape(-1,1), fare_scale_param)\n",
    "\n",
    "# 显示DataFrame的前5行数据，查看新增的'Age_scaled'和'Fare_scaled'列\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F6D48EE010B042EC870A17B8FFFD61B5",
    "mdEditEnable": false
   },
   "source": [
    "恩，好看多了，万事俱备，只欠建模。马上就要看到成效了，哈哈。我们把需要的属性值抽出来，转成`scikit-learn`里面`LogisticRegression`可以处理的格式。\n",
    "## 逻辑回归建模\n",
    "我们把需要的`feature`字段取出来，转成`numpy`格式，使用`scikit-learn`中的`LogisticRegression`建模。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": false,
    "id": "DCE1A3BC05B7443E8810C1EF8B5C14C8",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-2 {\n",
       "  /* Definition of color scheme common for light and dark mode */\n",
       "  --sklearn-color-text: #000;\n",
       "  --sklearn-color-text-muted: #666;\n",
       "  --sklearn-color-line: gray;\n",
       "  /* Definition of color scheme for unfitted estimators */\n",
       "  --sklearn-color-unfitted-level-0: #fff5e6;\n",
       "  --sklearn-color-unfitted-level-1: #f6e4d2;\n",
       "  --sklearn-color-unfitted-level-2: #ffe0b3;\n",
       "  --sklearn-color-unfitted-level-3: chocolate;\n",
       "  /* Definition of color scheme for fitted estimators */\n",
       "  --sklearn-color-fitted-level-0: #f0f8ff;\n",
       "  --sklearn-color-fitted-level-1: #d4ebff;\n",
       "  --sklearn-color-fitted-level-2: #b3dbfd;\n",
       "  --sklearn-color-fitted-level-3: cornflowerblue;\n",
       "\n",
       "  /* Specific color for light theme */\n",
       "  --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, white)));\n",
       "  --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, black)));\n",
       "  --sklearn-color-icon: #696969;\n",
       "\n",
       "  @media (prefers-color-scheme: dark) {\n",
       "    /* Redefinition of color scheme for dark theme */\n",
       "    --sklearn-color-text-on-default-background: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-background: var(--sg-background-color, var(--theme-background, var(--jp-layout-color0, #111)));\n",
       "    --sklearn-color-border-box: var(--sg-text-color, var(--theme-code-foreground, var(--jp-content-font-color1, white)));\n",
       "    --sklearn-color-icon: #878787;\n",
       "  }\n",
       "}\n",
       "\n",
       "#sk-container-id-2 {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 pre {\n",
       "  padding: 0;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 input.sk-hidden--visually {\n",
       "  border: 0;\n",
       "  clip: rect(1px 1px 1px 1px);\n",
       "  clip: rect(1px, 1px, 1px, 1px);\n",
       "  height: 1px;\n",
       "  margin: -1px;\n",
       "  overflow: hidden;\n",
       "  padding: 0;\n",
       "  position: absolute;\n",
       "  width: 1px;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-dashed-wrapped {\n",
       "  border: 1px dashed var(--sklearn-color-line);\n",
       "  margin: 0 0.4em 0.5em 0.4em;\n",
       "  box-sizing: border-box;\n",
       "  padding-bottom: 0.4em;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-container {\n",
       "  /* jupyter's `normalize.less` sets `[hidden] { display: none; }`\n",
       "     but bootstrap.min.css set `[hidden] { display: none !important; }`\n",
       "     so we also need the `!important` here to be able to override the\n",
       "     default hidden behavior on the sphinx rendered scikit-learn.org.\n",
       "     See: https://github.com/scikit-learn/scikit-learn/issues/21755 */\n",
       "  display: inline-block !important;\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-text-repr-fallback {\n",
       "  display: none;\n",
       "}\n",
       "\n",
       "div.sk-parallel-item,\n",
       "div.sk-serial,\n",
       "div.sk-item {\n",
       "  /* draw centered vertical line to link estimators */\n",
       "  background-image: linear-gradient(var(--sklearn-color-text-on-default-background), var(--sklearn-color-text-on-default-background));\n",
       "  background-size: 2px 100%;\n",
       "  background-repeat: no-repeat;\n",
       "  background-position: center center;\n",
       "}\n",
       "\n",
       "/* Parallel-specific style estimator block */\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item::after {\n",
       "  content: \"\";\n",
       "  width: 100%;\n",
       "  border-bottom: 2px solid var(--sklearn-color-text-on-default-background);\n",
       "  flex-grow: 1;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel {\n",
       "  display: flex;\n",
       "  align-items: stretch;\n",
       "  justify-content: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  position: relative;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item:first-child::after {\n",
       "  align-self: flex-end;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item:last-child::after {\n",
       "  align-self: flex-start;\n",
       "  width: 50%;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-parallel-item:only-child::after {\n",
       "  width: 0;\n",
       "}\n",
       "\n",
       "/* Serial-specific style estimator block */\n",
       "\n",
       "#sk-container-id-2 div.sk-serial {\n",
       "  display: flex;\n",
       "  flex-direction: column;\n",
       "  align-items: center;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  padding-right: 1em;\n",
       "  padding-left: 1em;\n",
       "}\n",
       "\n",
       "\n",
       "/* Toggleable style: style used for estimator/Pipeline/ColumnTransformer box that is\n",
       "clickable and can be expanded/collapsed.\n",
       "- Pipeline and ColumnTransformer use this feature and define the default style\n",
       "- Estimators will overwrite some part of the style using the `sk-estimator` class\n",
       "*/\n",
       "\n",
       "/* Pipeline and ColumnTransformer style (default) */\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable {\n",
       "  /* Default theme specific background. It is overwritten whether we have a\n",
       "  specific estimator or a Pipeline/ColumnTransformer */\n",
       "  background-color: var(--sklearn-color-background);\n",
       "}\n",
       "\n",
       "/* Toggleable label */\n",
       "#sk-container-id-2 label.sk-toggleable__label {\n",
       "  cursor: pointer;\n",
       "  display: flex;\n",
       "  width: 100%;\n",
       "  margin-bottom: 0;\n",
       "  padding: 0.5em;\n",
       "  box-sizing: border-box;\n",
       "  text-align: center;\n",
       "  align-items: start;\n",
       "  justify-content: space-between;\n",
       "  gap: 0.5em;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 label.sk-toggleable__label .caption {\n",
       "  font-size: 0.6rem;\n",
       "  font-weight: lighter;\n",
       "  color: var(--sklearn-color-text-muted);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 label.sk-toggleable__label-arrow:before {\n",
       "  /* Arrow on the left of the label */\n",
       "  content: \"▸\";\n",
       "  float: left;\n",
       "  margin-right: 0.25em;\n",
       "  color: var(--sklearn-color-icon);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 label.sk-toggleable__label-arrow:hover:before {\n",
       "  color: var(--sklearn-color-text);\n",
       "}\n",
       "\n",
       "/* Toggleable content - dropdown */\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content {\n",
       "  max-height: 0;\n",
       "  max-width: 0;\n",
       "  overflow: hidden;\n",
       "  text-align: left;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content pre {\n",
       "  margin: 0.2em;\n",
       "  border-radius: 0.25em;\n",
       "  color: var(--sklearn-color-text);\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-toggleable__content.fitted pre {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 input.sk-toggleable__control:checked~div.sk-toggleable__content {\n",
       "  /* Expand drop-down */\n",
       "  max-height: 200px;\n",
       "  max-width: 100%;\n",
       "  overflow: auto;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {\n",
       "  content: \"▾\";\n",
       "}\n",
       "\n",
       "/* Pipeline/ColumnTransformer-specific style */\n",
       "\n",
       "#sk-container-id-2 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-label.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator-specific style */\n",
       "\n",
       "/* Colorize estimator box */\n",
       "#sk-container-id-2 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-estimator.fitted input.sk-toggleable__control:checked~label.sk-toggleable__label {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-label label.sk-toggleable__label,\n",
       "#sk-container-id-2 div.sk-label label {\n",
       "  /* The background is the default theme color */\n",
       "  color: var(--sklearn-color-text-on-default-background);\n",
       "}\n",
       "\n",
       "/* On hover, darken the color of the background */\n",
       "#sk-container-id-2 div.sk-label:hover label.sk-toggleable__label {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "/* Label box, darken color on hover, fitted */\n",
       "#sk-container-id-2 div.sk-label.fitted:hover label.sk-toggleable__label.fitted {\n",
       "  color: var(--sklearn-color-text);\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Estimator label */\n",
       "\n",
       "#sk-container-id-2 div.sk-label label {\n",
       "  font-family: monospace;\n",
       "  font-weight: bold;\n",
       "  display: inline-block;\n",
       "  line-height: 1.2em;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-label-container {\n",
       "  text-align: center;\n",
       "}\n",
       "\n",
       "/* Estimator-specific */\n",
       "#sk-container-id-2 div.sk-estimator {\n",
       "  font-family: monospace;\n",
       "  border: 1px dotted var(--sklearn-color-border-box);\n",
       "  border-radius: 0.25em;\n",
       "  box-sizing: border-box;\n",
       "  margin-bottom: 0.5em;\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-0);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-estimator.fitted {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-0);\n",
       "}\n",
       "\n",
       "/* on hover */\n",
       "#sk-container-id-2 div.sk-estimator:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-2);\n",
       "}\n",
       "\n",
       "#sk-container-id-2 div.sk-estimator.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-2);\n",
       "}\n",
       "\n",
       "/* Specification for estimator info (e.g. \"i\" and \"?\") */\n",
       "\n",
       "/* Common style for \"i\" and \"?\" */\n",
       "\n",
       ".sk-estimator-doc-link,\n",
       "a:link.sk-estimator-doc-link,\n",
       "a:visited.sk-estimator-doc-link {\n",
       "  float: right;\n",
       "  font-size: smaller;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1em;\n",
       "  height: 1em;\n",
       "  width: 1em;\n",
       "  text-decoration: none !important;\n",
       "  margin-left: 0.5em;\n",
       "  text-align: center;\n",
       "  /* unfitted */\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted,\n",
       "a:link.sk-estimator-doc-link.fitted,\n",
       "a:visited.sk-estimator-doc-link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "div.sk-estimator:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link:hover,\n",
       ".sk-estimator-doc-link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "div.sk-estimator.fitted:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover,\n",
       "div.sk-label-container:hover .sk-estimator-doc-link.fitted:hover,\n",
       ".sk-estimator-doc-link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "/* Span, style for the box shown on hovering the info icon */\n",
       ".sk-estimator-doc-link span {\n",
       "  display: none;\n",
       "  z-index: 9999;\n",
       "  position: relative;\n",
       "  font-weight: normal;\n",
       "  right: .2ex;\n",
       "  padding: .5ex;\n",
       "  margin: .5ex;\n",
       "  width: min-content;\n",
       "  min-width: 20ex;\n",
       "  max-width: 50ex;\n",
       "  color: var(--sklearn-color-text);\n",
       "  box-shadow: 2pt 2pt 4pt #999;\n",
       "  /* unfitted */\n",
       "  background: var(--sklearn-color-unfitted-level-0);\n",
       "  border: .5pt solid var(--sklearn-color-unfitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link.fitted span {\n",
       "  /* fitted */\n",
       "  background: var(--sklearn-color-fitted-level-0);\n",
       "  border: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "\n",
       ".sk-estimator-doc-link:hover span {\n",
       "  display: block;\n",
       "}\n",
       "\n",
       "/* \"?\"-specific style due to the `<a>` HTML tag */\n",
       "\n",
       "#sk-container-id-2 a.estimator_doc_link {\n",
       "  float: right;\n",
       "  font-size: 1rem;\n",
       "  line-height: 1em;\n",
       "  font-family: monospace;\n",
       "  background-color: var(--sklearn-color-background);\n",
       "  border-radius: 1rem;\n",
       "  height: 1rem;\n",
       "  width: 1rem;\n",
       "  text-decoration: none;\n",
       "  /* unfitted */\n",
       "  color: var(--sklearn-color-unfitted-level-1);\n",
       "  border: var(--sklearn-color-unfitted-level-1) 1pt solid;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 a.estimator_doc_link.fitted {\n",
       "  /* fitted */\n",
       "  border: var(--sklearn-color-fitted-level-1) 1pt solid;\n",
       "  color: var(--sklearn-color-fitted-level-1);\n",
       "}\n",
       "\n",
       "/* On hover */\n",
       "#sk-container-id-2 a.estimator_doc_link:hover {\n",
       "  /* unfitted */\n",
       "  background-color: var(--sklearn-color-unfitted-level-3);\n",
       "  color: var(--sklearn-color-background);\n",
       "  text-decoration: none;\n",
       "}\n",
       "\n",
       "#sk-container-id-2 a.estimator_doc_link.fitted:hover {\n",
       "  /* fitted */\n",
       "  background-color: var(--sklearn-color-fitted-level-3);\n",
       "}\n",
       "</style><div id=\"sk-container-id-2\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LogisticRegression(penalty=&#x27;l1&#x27;, solver=&#x27;liblinear&#x27;, tol=1e-06)</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator fitted sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-2\" type=\"checkbox\" checked><label for=\"sk-estimator-id-2\" class=\"sk-toggleable__label fitted sk-toggleable__label-arrow\"><div><div>LogisticRegression</div></div><div><a class=\"sk-estimator-doc-link fitted\" rel=\"noreferrer\" target=\"_blank\" href=\"https://scikit-learn.org/1.6/modules/generated/sklearn.linear_model.LogisticRegression.html\">?<span>Documentation for LogisticRegression</span></a><span class=\"sk-estimator-doc-link fitted\">i<span>Fitted</span></span></div></label><div class=\"sk-toggleable__content fitted\"><pre>LogisticRegression(penalty=&#x27;l1&#x27;, solver=&#x27;liblinear&#x27;, tol=1e-06)</pre></div> </div></div></div></div>"
      ],
      "text/plain": [
       "LogisticRegression(penalty='l1', solver='liblinear', tol=1e-06)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import linear_model\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "# 用正则取出我们要的属性值\n",
    "train_df = df.filter(regex='Survived|Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass_.*')\n",
    "\n",
    "# 使用SimpleImputer填充缺失值，选择使用均值填充\n",
    "imputer = SimpleImputer(strategy='mean')\n",
    "train_df_imputed = imputer.fit_transform(train_df)\n",
    "\n",
    "# y即第0列：Survival结果\n",
    "y = train_df_imputed[:, 0]\n",
    "\n",
    "# X即第1列及以后：特征属性值\n",
    "X = train_df_imputed[:, 1:]\n",
    "\n",
    "# 标准化特征数据\n",
    "scaler = StandardScaler()\n",
    "X_scaled = scaler.fit_transform(X)\n",
    "\n",
    "# 创建LogisticRegression模型\n",
    "clf = linear_model.LogisticRegression(solver='liblinear', C=1.0, penalty='l1', tol=1e-6)\n",
    "\n",
    "# 训练模型\n",
    "clf.fit(X_scaled, y)\n",
    "\n",
    "# 输出训练好的模型\n",
    "clf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "D5A6805C05604A8D8DDB2382248F29D0",
    "mdEditEnable": false
   },
   "source": [
    "`Good`，很顺利，我们得到了一个 `model`。\n",
    "\n",
    "先淡定！淡定！你以为把`test.csv`直接丢进`model`里就能拿到结果啊…骚年，图样图森破啊！我们的”`test_data`”也要做和”`train_data`”一样的预处理啊！！"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false,
    "id": "E984C35272844F5182912EE3898B3A81",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin_No</th>\n",
       "      <th>Cabin_Yes</th>\n",
       "      <th>Embarked_C</th>\n",
       "      <th>Embarked_Q</th>\n",
       "      <th>Embarked_S</th>\n",
       "      <th>Sex_female</th>\n",
       "      <th>Sex_male</th>\n",
       "      <th>Pclass_1</th>\n",
       "      <th>Pclass_2</th>\n",
       "      <th>Pclass_3</th>\n",
       "      <th>Age_scaled</th>\n",
       "      <th>Fare_scaled</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892</td>\n",
       "      <td>34.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.8292</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>0.307526</td>\n",
       "      <td>-0.496637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>47.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.0000</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>1.256242</td>\n",
       "      <td>-0.511497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>62.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9.6875</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>2.394702</td>\n",
       "      <td>-0.463335</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>27.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.6625</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>-0.261704</td>\n",
       "      <td>-0.481704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>12.2875</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>-0.641190</td>\n",
       "      <td>-0.416740</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId   Age  SibSp  Parch     Fare  Cabin_No  Cabin_Yes  Embarked_C  \\\n",
       "0          892  34.5      0      0   7.8292      True      False       False   \n",
       "1          893  47.0      1      0   7.0000      True      False       False   \n",
       "2          894  62.0      0      0   9.6875      True      False       False   \n",
       "3          895  27.0      0      0   8.6625      True      False       False   \n",
       "4          896  22.0      1      1  12.2875      True      False       False   \n",
       "\n",
       "   Embarked_Q  Embarked_S  Sex_female  Sex_male  Pclass_1  Pclass_2  Pclass_3  \\\n",
       "0        True       False       False      True     False     False      True   \n",
       "1       False        True        True     False     False     False      True   \n",
       "2        True       False       False      True     False      True     False   \n",
       "3       False        True       False      True     False     False      True   \n",
       "4       False        True        True     False     False     False      True   \n",
       "\n",
       "   Age_scaled  Fare_scaled  \n",
       "0    0.307526    -0.496637  \n",
       "1    1.256242    -0.511497  \n",
       "2    2.394702    -0.463335  \n",
       "3   -0.261704    -0.481704  \n",
       "4   -0.641190    -0.416740  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 读取 test.csv 数据\n",
    "data_test = pd.read_csv(\"test.csv\")\n",
    "\n",
    "# 对 Fare 列中的缺失值进行处理，将缺失值填充为 0\n",
    "data_test.loc[ (data_test.Fare.isnull()), 'Fare' ] = 0\n",
    "\n",
    "# 接着我们对 test_data 进行与 train_data 中一致的特征变换\n",
    "\n",
    "# 1. 用 RandomForestRegressor 模型填充缺失的年龄数据\n",
    "# 提取包含 Age, Fare, Parch, SibSp, Pclass 列的数据\n",
    "tmp_df = data_test[['Age','Fare', 'Parch', 'SibSp', 'Pclass']]\n",
    "\n",
    "# 获取 Age 列为缺失的行\n",
    "null_age = tmp_df[data_test.Age.isnull()].values\n",
    "\n",
    "# 提取非缺失的特征用于预测年龄（去除 Age 列）\n",
    "X = null_age[:, 1:]\n",
    "\n",
    "# 用训练好的 RandomForestRegressor 模型预测缺失的年龄\n",
    "predictedAges = rfr.predict(X)\n",
    "\n",
    "# 将预测的年龄值填充回原数据中缺失的位置\n",
    "data_test.loc[ (data_test.Age.isnull()), 'Age' ] = predictedAges\n",
    "\n",
    "# 2. 对 Cabin 列进行处理，可能会有预定义的 set_Cabin_type 函数\n",
    "data_test = set_Cabin_type(data_test)\n",
    "\n",
    "# 3. 对分类特征进行独热编码处理\n",
    "# 对 Cabin 列进行独热编码\n",
    "dummies_Cabin = pd.get_dummies(data_test['Cabin'], prefix='Cabin')\n",
    "\n",
    "# 对 Embarked 列进行独热编码\n",
    "dummies_Embarked = pd.get_dummies(data_test['Embarked'], prefix='Embarked')\n",
    "\n",
    "# 对 Sex 列进行独热编码\n",
    "dummies_Sex = pd.get_dummies(data_test['Sex'], prefix='Sex')\n",
    "\n",
    "# 对 Pclass 列进行独热编码\n",
    "dummies_Pclass = pd.get_dummies(data_test['Pclass'], prefix='Pclass')\n",
    "\n",
    "# 4. 将独热编码后的列和原数据合并\n",
    "df_test = pd.concat([data_test, dummies_Cabin, dummies_Embarked, dummies_Sex, dummies_Pclass], axis=1)\n",
    "\n",
    "# 5. 删除原来的不再需要的列\n",
    "df_test.drop(['Pclass', 'Name', 'Sex', 'Ticket', 'Cabin', 'Embarked'], axis=1, inplace=True)\n",
    "\n",
    "# 6. 对 Age 和 Fare 进行特征缩放（标准化）\n",
    "df_test['Age_scaled'] = scaler.fit_transform(df_test['Age'].values.reshape(-1,1), age_scale_param)\n",
    "df_test['Fare_scaled'] = scaler.fit_transform(df_test['Fare'].values.reshape(-1,1), fare_scale_param)\n",
    "\n",
    "# 查看处理后的数据前几行\n",
    "df_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "25901C4D9F6B47678E87FC4D898D1CDA",
    "mdEditEnable": false
   },
   "source": [
    "不错不错，数据很OK，差最后一步了。 \n",
    "下面就做预测取结果吧！！"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "collapsed": false,
    "id": "E140D03943F24EE6BC56983AFC1D8182",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "# 确保在训练时使用 DataFrame 进行训练\n",
    "# 使用正则表达式从原始数据框 `df` 中选择需要的特征列\n",
    "train_df = df.filter(regex='Survived|Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass_.*')\n",
    "\n",
    "# 提取特征矩阵 X_train（去除目标列 'Survived'）\n",
    "X_train = train_df.drop(columns='Survived')  # 特征\n",
    "\n",
    "# 提取目标变量 y_train（Survived）\n",
    "y_train = train_df['Survived']  # 目标\n",
    "\n",
    "# 训练 LogisticRegression 模型\n",
    "# 选择 'liblinear' 求解器，L1 正则化，容忍度设置为 1e-6\n",
    "clf = linear_model.LogisticRegression(solver='liblinear', penalty='l1', tol=1e-6)\n",
    "\n",
    "# 使用训练数据 X_train 和目标变量 y_train 进行模型训练\n",
    "clf.fit(X_train, y_train)\n",
    "\n",
    "# 使用相同的 DataFrame 进行预测\n",
    "# 从测试数据框 `df_test` 中选择与训练数据中相同的特征列\n",
    "test = df_test.filter(regex='Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass_.*')\n",
    "\n",
    "# 使用训练好的模型 `clf` 对测试数据进行预测\n",
    "predictions = clf.predict(test)\n",
    "\n",
    "# 保存预测结果\n",
    "# 创建一个 DataFrame，包含 PassengerId 和预测结果 Survived\n",
    "result = pd.DataFrame({'PassengerId': data_test['PassengerId'].values, 'Survived': predictions.astype(np.int32)})\n",
    "\n",
    "# 将预测结果保存为 CSV 文件，不包含索引\n",
    "result.to_csv(\"logistic_regression_predictions.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false,
    "id": "5AD6B7EEB61B420C894E2964A3AB7993",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived\n",
       "0          892         0\n",
       "1          893         0\n",
       "2          894         0\n",
       "3          895         0\n",
       "4          896         1"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"logistic_regression_predictions.csv\").head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6D6B7CAACAEF4264824F93794CC731FE",
    "mdEditEnable": false
   },
   "source": [
    "\n",
    "啧啧，挺好，格式正确，去`make a submission`啦啦啦！\n",
    "\n",
    "在`Kaggle`的`Make a submission`页面，提交上结果。如下： \n",
    "\n",
    "![提交结果](https://www.z4a.net/images/2018/11/28/result_2.jpg)\n",
    "\n",
    "`0.76555`，恩，结果还不错。毕竟，这只是我们简单分析处理过后出的一个`baseline`模型嘛。\n",
    "\n",
    "## 逻辑回归系统优化\n",
    "### 模型系数关联分析\n",
    "亲，你以为结果提交上了，就完事了？ \n",
    "我不会告诉你，这只是万里长征第一步啊(泪牛满面)！！！这才刚撸完`baseline model`啊！！！还得优化啊！！！\n",
    "\n",
    "看过`Andrew Ng`老师的`machine Learning`课程的同学们，知道，我们应该分析分析模型现在的状态了，是过/欠拟合？，以确定我们需要更多的特征还是更多数据，或者其他操作。我们有一条很著名的learning curves对吧。\n",
    "\n",
    "不过在现在的场景下，先不着急做这个事情，我们这个baseline系统还有些粗糙，先再挖掘挖掘。\n",
    "\n",
    "首先，`Name`和`Ticket`两个属性被我们完整舍弃了(好吧，其实是因为这俩属性，几乎每一条记录都是一个完全不同的值，我们并没有找到很直接的处理方式)。\n",
    "\n",
    "然后，我们想想，年龄的拟合本身也未必是一件非常靠谱的事情，我们依据其余属性，其实并不能很好地拟合预测出未知的年龄。再一个，以我们的日常经验，小盆友和老人可能得到的照顾会多一些，这样看的话，年龄作为一个连续值，给一个固定的系数，应该和年龄是一个正相关或者负相关，似乎体现不出两头受照顾的实际情况，所以，说不定我们把年龄离散化，按区段分作类别属性会更合适一些。\n",
    "\n",
    "上面只是我瞎想的，who knows是不是这么回事呢，老老实实先把得到的model系数和feature关联起来看看。\n",
    "\n",
    "**LR模型系数：**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false,
    "id": "34F8D229D4764822AF8E21C87060A4B0",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>columns</th>\n",
       "      <th>coef</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SibSp</td>\n",
       "      <td>[-0.3442335383470622]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Parch</td>\n",
       "      <td>[-0.10491788507492303]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Cabin_No</td>\n",
       "      <td>[-0.1499906003268399]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Cabin_Yes</td>\n",
       "      <td>[0.752097754166376]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Embarked_C</td>\n",
       "      <td>[0.0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Embarked_Q</td>\n",
       "      <td>[0.0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Embarked_S</td>\n",
       "      <td>[-0.4172610602198502]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Sex_female</td>\n",
       "      <td>[2.1065598457659522]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Sex_male</td>\n",
       "      <td>[-0.5274274881321755]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Pclass_1</td>\n",
       "      <td>[0.34116836134281964]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Pclass_2</td>\n",
       "      <td>[0.0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Pclass_3</td>\n",
       "      <td>[-1.1941323311741345]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>Age_scaled</td>\n",
       "      <td>[-0.5237629090103685]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Fare_scaled</td>\n",
       "      <td>[0.08443572751412937]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        columns                    coef\n",
       "0         SibSp   [-0.3442335383470622]\n",
       "1         Parch  [-0.10491788507492303]\n",
       "2      Cabin_No   [-0.1499906003268399]\n",
       "3     Cabin_Yes     [0.752097754166376]\n",
       "4    Embarked_C                   [0.0]\n",
       "5    Embarked_Q                   [0.0]\n",
       "6    Embarked_S   [-0.4172610602198502]\n",
       "7    Sex_female    [2.1065598457659522]\n",
       "8      Sex_male   [-0.5274274881321755]\n",
       "9      Pclass_1   [0.34116836134281964]\n",
       "10     Pclass_2                   [0.0]\n",
       "11     Pclass_3   [-1.1941323311741345]\n",
       "12   Age_scaled   [-0.5237629090103685]\n",
       "13  Fare_scaled   [0.08443572751412937]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame({\"columns\":list(train_df.columns)[1:], \"coef\":list(clf.coef_.T)})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false,
    "id": "CAEA1C315A1D43E0B045C0FBF9C2C4F3",
    "jupyter": {
     "outputs_hidden": false
    },
    "mdEditEnable": false,
    "scrolled": false
   },
   "source": [
    "首先，大家回去前两篇文章里瞄一眼公式就知道，这些系数为正的特征，和最后结果是一个正相关，反之为负相关。\n",
    "\n",
    "我们先看看那些权重绝对值非常大的feature，在我们的模型上：\n",
    "\n",
    "- Sex属性，如果是female会极大提高最后获救的概率，而male会很大程度拉低这个概率。\n",
    "- Pclass属性，1等舱乘客最后获救的概率会上升，而乘客等级为3会极大地拉低这个概率。\n",
    "- 有Cabin值会很大程度拉升最后获救概率(这里似乎能看到了一点端倪，事实上从最上面的有无Cabin记录的Survived分布图上看出，即使有Cabin记录的乘客也有一部分遇难了，估计这个属性上我们挖掘还不够)\n",
    "- Age是一个负相关，意味着在我们的模型里，年龄越小，越有获救的优先权(还得回原数据看看这个是否合理）\n",
    "- 有一个登船港口S会很大程度拉低获救的概率，另外俩港口压根就没啥作用(这个实际上非常奇怪，因为我们从之前的统计图上并没有看到S港口的获救率非常低，所以也许可以考虑把登船港口这个feature去掉试试)。\n",
    "- 船票Fare有小幅度的正相关(并不意味着这个feature作用不大，有可能是我们细化的程度还不够，举个例子，说不定我们得对它离散化，再分至各个乘客等级上？)\n",
    "\n",
    "噢啦，观察完了，我们现在有一些想法了，但是怎么样才知道，哪些优化的方法是promising的呢？\n",
    "\n",
    "因为test.csv里面并没有Survived这个字段(好吧，这是废话，这明明就是我们要预测的结果)，我们无法在这份数据上评定我们算法在该场景下的效果…\n",
    "\n",
    "而『每做一次调整就make a submission，然后根据结果来判定这次调整的好坏』其实是行不通的…\n",
    "\n",
    "### 交叉验证\n",
    "\n",
    "重点又来了：\n",
    "\n",
    "- **『要做交叉验证(cross validation)!』 **\n",
    "- **『要做交叉验证(cross validation)!』 **\n",
    "- **『要做交叉验证(cross validation)!』 **\n",
    "\n",
    "恩，重要的事情说三遍。我们通常情况下，这么做cross validation：把train.csv分成两部分，一部分用于训练我们需要的模型，另外一部分数据上看我们预测算法的效果。\n",
    "\n",
    "我们用scikit-learn的cross_validation来帮我们完成小数据集上的这个工作。\n",
    "\n",
    "先简单看看cross validation情况下的打分\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "collapsed": false,
    "id": "0399477A313B4B0484FF4E558E6F06A8",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "\nAll the 5 fits failed.\nIt is very likely that your model is misconfigured.\nYou can try to debug the error by setting error_score='raise'.\n\nBelow are more details about the failures:\n--------------------------------------------------------------------------------\n5 fits failed with the following error:\nTraceback (most recent call last):\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/model_selection/_validation.py\", line 866, in _fit_and_score\n    estimator.fit(X_train, y_train, **fit_params)\n    ~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/base.py\", line 1389, in wrapper\n    return fit_method(estimator, *args, **kwargs)\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/linear_model/_logistic.py\", line 1231, in fit\n    check_classification_targets(y)\n    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/utils/multiclass.py\", line 222, in check_classification_targets\n    raise ValueError(\n    ...<3 lines>...\n    )\nValueError: Unknown label type: unknown. Maybe you are trying to fit a classifier, which expects discrete classes on a regression target with continuous values.\n",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[43], line 12\u001b[0m\n\u001b[1;32m     10\u001b[0m y \u001b[38;5;241m=\u001b[39m all_data\u001b[38;5;241m.\u001b[39mvalues[:,\u001b[38;5;241m0\u001b[39m]\n\u001b[1;32m     11\u001b[0m \u001b[38;5;66;03m# print(cross_validation.cross_val_score(clf, X, y, cv=5))\u001b[39;00m\n\u001b[0;32m---> 12\u001b[0m \u001b[38;5;28mprint\u001b[39m(\u001b[43mcross_val_score\u001b[49m\u001b[43m(\u001b[49m\u001b[43mclf\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mX\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43my\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcv\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m5\u001b[39;49m\u001b[43m)\u001b[49m)\n",
      "File \u001b[0;32m/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/utils/_param_validation.py:216\u001b[0m, in \u001b[0;36mvalidate_params.<locals>.decorator.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    210\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m    211\u001b[0m     \u001b[38;5;28;01mwith\u001b[39;00m config_context(\n\u001b[1;32m    212\u001b[0m         skip_parameter_validation\u001b[38;5;241m=\u001b[39m(\n\u001b[1;32m    213\u001b[0m             prefer_skip_nested_validation \u001b[38;5;129;01mor\u001b[39;00m global_skip_validation\n\u001b[1;32m    214\u001b[0m         )\n\u001b[1;32m    215\u001b[0m     ):\n\u001b[0;32m--> 216\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    217\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m InvalidParameterError \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[1;32m    218\u001b[0m     \u001b[38;5;66;03m# When the function is just a wrapper around an estimator, we allow\u001b[39;00m\n\u001b[1;32m    219\u001b[0m     \u001b[38;5;66;03m# the function to delegate validation to the estimator, but we replace\u001b[39;00m\n\u001b[1;32m    220\u001b[0m     \u001b[38;5;66;03m# the name of the estimator by the name of the function in the error\u001b[39;00m\n\u001b[1;32m    221\u001b[0m     \u001b[38;5;66;03m# message to avoid confusion.\u001b[39;00m\n\u001b[1;32m    222\u001b[0m     msg \u001b[38;5;241m=\u001b[39m re\u001b[38;5;241m.\u001b[39msub(\n\u001b[1;32m    223\u001b[0m         \u001b[38;5;124mr\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mparameter of \u001b[39m\u001b[38;5;124m\\\u001b[39m\u001b[38;5;124mw+ must be\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    224\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mparameter of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mfunc\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__qualname__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m must be\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    225\u001b[0m         \u001b[38;5;28mstr\u001b[39m(e),\n\u001b[1;32m    226\u001b[0m     )\n",
      "File \u001b[0;32m/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/model_selection/_validation.py:684\u001b[0m, in \u001b[0;36mcross_val_score\u001b[0;34m(estimator, X, y, groups, scoring, cv, n_jobs, verbose, params, pre_dispatch, error_score)\u001b[0m\n\u001b[1;32m    681\u001b[0m \u001b[38;5;66;03m# To ensure multimetric format is not supported\u001b[39;00m\n\u001b[1;32m    682\u001b[0m scorer \u001b[38;5;241m=\u001b[39m check_scoring(estimator, scoring\u001b[38;5;241m=\u001b[39mscoring)\n\u001b[0;32m--> 684\u001b[0m cv_results \u001b[38;5;241m=\u001b[39m \u001b[43mcross_validate\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m    685\u001b[0m \u001b[43m    \u001b[49m\u001b[43mestimator\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mestimator\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    686\u001b[0m \u001b[43m    \u001b[49m\u001b[43mX\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mX\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    687\u001b[0m \u001b[43m    \u001b[49m\u001b[43my\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43my\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    688\u001b[0m \u001b[43m    \u001b[49m\u001b[43mgroups\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mgroups\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    689\u001b[0m \u001b[43m    \u001b[49m\u001b[43mscoring\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m{\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mscore\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m:\u001b[49m\u001b[43m \u001b[49m\u001b[43mscorer\u001b[49m\u001b[43m}\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    690\u001b[0m \u001b[43m    \u001b[49m\u001b[43mcv\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcv\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    691\u001b[0m \u001b[43m    \u001b[49m\u001b[43mn_jobs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mn_jobs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    692\u001b[0m \u001b[43m    \u001b[49m\u001b[43mverbose\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mverbose\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    693\u001b[0m \u001b[43m    \u001b[49m\u001b[43mparams\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mparams\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    694\u001b[0m \u001b[43m    \u001b[49m\u001b[43mpre_dispatch\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mpre_dispatch\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    695\u001b[0m \u001b[43m    \u001b[49m\u001b[43merror_score\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43merror_score\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m    696\u001b[0m \u001b[43m\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    697\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m cv_results[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtest_score\u001b[39m\u001b[38;5;124m\"\u001b[39m]\n",
      "File \u001b[0;32m/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/utils/_param_validation.py:216\u001b[0m, in \u001b[0;36mvalidate_params.<locals>.decorator.<locals>.wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    210\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m    211\u001b[0m     \u001b[38;5;28;01mwith\u001b[39;00m config_context(\n\u001b[1;32m    212\u001b[0m         skip_parameter_validation\u001b[38;5;241m=\u001b[39m(\n\u001b[1;32m    213\u001b[0m             prefer_skip_nested_validation \u001b[38;5;129;01mor\u001b[39;00m global_skip_validation\n\u001b[1;32m    214\u001b[0m         )\n\u001b[1;32m    215\u001b[0m     ):\n\u001b[0;32m--> 216\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    217\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m InvalidParameterError \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[1;32m    218\u001b[0m     \u001b[38;5;66;03m# When the function is just a wrapper around an estimator, we allow\u001b[39;00m\n\u001b[1;32m    219\u001b[0m     \u001b[38;5;66;03m# the function to delegate validation to the estimator, but we replace\u001b[39;00m\n\u001b[1;32m    220\u001b[0m     \u001b[38;5;66;03m# the name of the estimator by the name of the function in the error\u001b[39;00m\n\u001b[1;32m    221\u001b[0m     \u001b[38;5;66;03m# message to avoid confusion.\u001b[39;00m\n\u001b[1;32m    222\u001b[0m     msg \u001b[38;5;241m=\u001b[39m re\u001b[38;5;241m.\u001b[39msub(\n\u001b[1;32m    223\u001b[0m         \u001b[38;5;124mr\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mparameter of \u001b[39m\u001b[38;5;124m\\\u001b[39m\u001b[38;5;124mw+ must be\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    224\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mparameter of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mfunc\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__qualname__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m must be\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m    225\u001b[0m         \u001b[38;5;28mstr\u001b[39m(e),\n\u001b[1;32m    226\u001b[0m     )\n",
      "File \u001b[0;32m/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/model_selection/_validation.py:431\u001b[0m, in \u001b[0;36mcross_validate\u001b[0;34m(estimator, X, y, groups, scoring, cv, n_jobs, verbose, params, pre_dispatch, return_train_score, return_estimator, return_indices, error_score)\u001b[0m\n\u001b[1;32m    410\u001b[0m parallel \u001b[38;5;241m=\u001b[39m Parallel(n_jobs\u001b[38;5;241m=\u001b[39mn_jobs, verbose\u001b[38;5;241m=\u001b[39mverbose, pre_dispatch\u001b[38;5;241m=\u001b[39mpre_dispatch)\n\u001b[1;32m    411\u001b[0m results \u001b[38;5;241m=\u001b[39m parallel(\n\u001b[1;32m    412\u001b[0m     delayed(_fit_and_score)(\n\u001b[1;32m    413\u001b[0m         clone(estimator),\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    428\u001b[0m     \u001b[38;5;28;01mfor\u001b[39;00m train, test \u001b[38;5;129;01min\u001b[39;00m indices\n\u001b[1;32m    429\u001b[0m )\n\u001b[0;32m--> 431\u001b[0m \u001b[43m_warn_or_raise_about_fit_failures\u001b[49m\u001b[43m(\u001b[49m\u001b[43mresults\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43merror_score\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    433\u001b[0m \u001b[38;5;66;03m# For callable scoring, the return type is only know after calling. If the\u001b[39;00m\n\u001b[1;32m    434\u001b[0m \u001b[38;5;66;03m# return type is a dictionary, the error scores can now be inserted with\u001b[39;00m\n\u001b[1;32m    435\u001b[0m \u001b[38;5;66;03m# the correct key.\u001b[39;00m\n\u001b[1;32m    436\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mcallable\u001b[39m(scoring):\n",
      "File \u001b[0;32m/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/model_selection/_validation.py:517\u001b[0m, in \u001b[0;36m_warn_or_raise_about_fit_failures\u001b[0;34m(results, error_score)\u001b[0m\n\u001b[1;32m    510\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m num_failed_fits \u001b[38;5;241m==\u001b[39m num_fits:\n\u001b[1;32m    511\u001b[0m     all_fits_failed_message \u001b[38;5;241m=\u001b[39m (\n\u001b[1;32m    512\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124mAll the \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mnum_fits\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m fits failed.\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    513\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mIt is very likely that your model is misconfigured.\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    514\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mYou can try to debug the error by setting error_score=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mraise\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m.\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    515\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mBelow are more details about the failures:\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;132;01m{\u001b[39;00mfit_errors_summary\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    516\u001b[0m     )\n\u001b[0;32m--> 517\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(all_fits_failed_message)\n\u001b[1;32m    519\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[1;32m    520\u001b[0m     some_fits_failed_message \u001b[38;5;241m=\u001b[39m (\n\u001b[1;32m    521\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;132;01m{\u001b[39;00mnum_failed_fits\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m fits failed out of a total of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mnum_fits\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m.\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    522\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mThe score on these train-test partitions for these parameters\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m   (...)\u001b[0m\n\u001b[1;32m    526\u001b[0m         \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mBelow are more details about the failures:\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;132;01m{\u001b[39;00mfit_errors_summary\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    527\u001b[0m     )\n",
      "\u001b[0;31mValueError\u001b[0m: \nAll the 5 fits failed.\nIt is very likely that your model is misconfigured.\nYou can try to debug the error by setting error_score='raise'.\n\nBelow are more details about the failures:\n--------------------------------------------------------------------------------\n5 fits failed with the following error:\nTraceback (most recent call last):\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/model_selection/_validation.py\", line 866, in _fit_and_score\n    estimator.fit(X_train, y_train, **fit_params)\n    ~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/base.py\", line 1389, in wrapper\n    return fit_method(estimator, *args, **kwargs)\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/linear_model/_logistic.py\", line 1231, in fit\n    check_classification_targets(y)\n    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^\n  File \"/opt/homebrew/Homebrew/Cellar/jupyterlab/4.3.5/libexec/lib/python3.13/site-packages/sklearn/utils/multiclass.py\", line 222, in check_classification_targets\n    raise ValueError(\n    ...<3 lines>...\n    )\nValueError: Unknown label type: unknown. Maybe you are trying to fit a classifier, which expects discrete classes on a regression target with continuous values.\n"
     ]
    }
   ],
   "source": [
    "# from sklearn import cross_validation\n",
    "# 参考https://blog.csdn.net/cheneyshark/article/details/78640887 ， 0.18版本中，cross_validation被废弃\n",
    "# 改为下面的从model_selection直接import cross_val_score 和 train_test_split\n",
    "from sklearn.model_selection import cross_val_score, train_test_split\n",
    "\n",
    " #简单看看打分情况\n",
    "clf = linear_model.LogisticRegression(solver='liblinear',C=1.0, penalty='l1', tol=1e-6)\n",
    "all_data = df.filter(regex='Survived|Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass_.*')\n",
    "X = all_data.values[:,1:]\n",
    "y = all_data.values[:,0]\n",
    "# print(cross_validation.cross_val_score(clf, X, y, cv=5))\n",
    "print(cross_val_score(clf, X, y, cv=5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "325C5FFAE140492B829C3750CC0F2C3B",
    "mdEditEnable": false
   },
   "source": [
    "结果是下面酱紫的： \n",
    "[0.81564246 0.81005587 0.78651685 0.78651685 0.81355932]\n",
    "\n",
    "似乎比Kaggle上的结果略高哈，毕竟用的是不是同一份数据集评估的。\n",
    "\n",
    "等等，既然我们要做交叉验证，那我们干脆先把交叉验证里面的bad case拿出来看看，看看人眼审核，是否能发现什么蛛丝马迹，是我们忽略了哪些信息，使得这些乘客被判定错了。再把bad case上得到的想法和前头系数分析的合在一起，然后逐个试试。\n",
    "\n",
    "下面我们做数据分割，并且在原始数据集上瞄一眼bad case："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false,
    "id": "FDB2637F9C574ED88B41F198D3C677C8",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</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>23</th>\n",
       "      <td>24</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Sloper, Mr. William Thompson</td>\n",
       "      <td>male</td>\n",
       "      <td>28.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>113788</td>\n",
       "      <td>35.5000</td>\n",
       "      <td>A6</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>26</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Asplund, Mrs. Carl Oscar (Selma Augusta Emilia...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.00</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>347077</td>\n",
       "      <td>31.3875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>50</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Arnold-Franchi, Mrs. Josef (Josefine Franchi)</td>\n",
       "      <td>female</td>\n",
       "      <td>18.00</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>349237</td>\n",
       "      <td>17.8000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>56</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Woolner, Mr. Hugh</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>19947</td>\n",
       "      <td>35.5000</td>\n",
       "      <td>C52</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>66</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Moubarek, Master. Gerios</td>\n",
       "      <td>male</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2661</td>\n",
       "      <td>15.2458</td>\n",
       "      <td>NaN</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>79</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>Caldwell, Master. Alden Gates</td>\n",
       "      <td>male</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>248738</td>\n",
       "      <td>29.0000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>82</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Sheerlinck, Mr. Jan Baptist</td>\n",
       "      <td>male</td>\n",
       "      <td>29.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>345779</td>\n",
       "      <td>9.5000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>119</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Baxter, Mr. Quigg Edmond</td>\n",
       "      <td>male</td>\n",
       "      <td>24.00</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>PC 17558</td>\n",
       "      <td>247.5208</td>\n",
       "      <td>B58 B60</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>139</th>\n",
       "      <td>140</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>Giglio, Mr. Victor</td>\n",
       "      <td>male</td>\n",
       "      <td>24.00</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17593</td>\n",
       "      <td>79.2000</td>\n",
       "      <td>B86</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>165</th>\n",
       "      <td>166</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Goldsmith, Master. Frank John William \"Frankie\"</td>\n",
       "      <td>male</td>\n",
       "      <td>9.00</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>363291</td>\n",
       "      <td>20.5250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     PassengerId  Survived  Pclass  \\\n",
       "23            24         1       1   \n",
       "25            26         1       3   \n",
       "49            50         0       3   \n",
       "55            56         1       1   \n",
       "65            66         1       3   \n",
       "78            79         1       2   \n",
       "81            82         1       3   \n",
       "118          119         0       1   \n",
       "139          140         0       1   \n",
       "165          166         1       3   \n",
       "\n",
       "                                                  Name     Sex    Age  SibSp  \\\n",
       "23                        Sloper, Mr. William Thompson    male  28.00      0   \n",
       "25   Asplund, Mrs. Carl Oscar (Selma Augusta Emilia...  female  38.00      1   \n",
       "49       Arnold-Franchi, Mrs. Josef (Josefine Franchi)  female  18.00      1   \n",
       "55                                   Woolner, Mr. Hugh    male    NaN      0   \n",
       "65                            Moubarek, Master. Gerios    male    NaN      1   \n",
       "78                       Caldwell, Master. Alden Gates    male   0.83      0   \n",
       "81                         Sheerlinck, Mr. Jan Baptist    male  29.00      0   \n",
       "118                           Baxter, Mr. Quigg Edmond    male  24.00      0   \n",
       "139                                 Giglio, Mr. Victor    male  24.00      0   \n",
       "165    Goldsmith, Master. Frank John William \"Frankie\"    male   9.00      0   \n",
       "\n",
       "     Parch    Ticket      Fare    Cabin Embarked  \n",
       "23       0    113788   35.5000       A6        S  \n",
       "25       5    347077   31.3875      NaN        S  \n",
       "49       0    349237   17.8000      NaN        S  \n",
       "55       0     19947   35.5000      C52        S  \n",
       "65       1      2661   15.2458      NaN        C  \n",
       "78       2    248738   29.0000      NaN        S  \n",
       "81       0    345779    9.5000      NaN        S  \n",
       "118      1  PC 17558  247.5208  B58 B60        C  \n",
       "139      0  PC 17593   79.2000      B86        C  \n",
       "165      2    363291   20.5250      NaN        S  "
      ]
     },
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 分割数据，按照 训练数据:cv数据 = 7:3的比例\n",
    "# split_train, split_cv = cross_validation.train_test_split(df, test_size=0.3, random_state=0)\n",
    "split_train, split_cv = train_test_split(df, test_size=0.3, random_state=42)\n",
    "\n",
    "train_df = split_train.filter(regex='Survived|Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass_.*')\n",
    "# 生成模型\n",
    "clf = linear_model.LogisticRegression(solver='liblinear',C=1.0, penalty='l1', tol=1e-6)\n",
    "clf.fit(train_df.values[:,1:], train_df.values[:,0])\n",
    "\n",
    "# 对cross validation数据进行预测\n",
    "\n",
    "cv_df = split_cv.filter(regex='Survived|Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass_.*')\n",
    "predictions = clf.predict(cv_df.values[:,1:])\n",
    "\n",
    "origin_data_train = pd.read_csv(\"/home/kesci/input/titanic8120/train.csv\")\n",
    "bad_cases = origin_data_train.loc[origin_data_train['PassengerId'].isin(split_cv[predictions != cv_df.values[:,0]]['PassengerId'].values)]\n",
    "bad_cases.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FCBECB2990084A9587F17BB87BE509ED",
    "mdEditEnable": false
   },
   "source": [
    "大家可以自己跑一遍试试，拿到bad cases之后，仔细看看。也会有一些猜测和想法。其中会有一部分可能会印证在系数分析部分的猜测，那这些优化的想法优先级可以放高一些。\n",
    "\n",
    "现在有了”train_df” 和 “vc_df” 两个数据部分，前者用于训练model，后者用于评定和选择模型。可以开始可劲折腾了。\n",
    "\n",
    "我们随便列一些可能可以做的优化操作：\n",
    "\n",
    "- Age属性不使用现在的拟合方式，而是根据名称中的『Mr』『Mrs』『Miss』等的平均值进行填充。\n",
    "- Age不做成一个连续值属性，而是使用一个步长进行离散化，变成离散的类目feature。\n",
    "- Cabin再细化一些，对于有记录的Cabin属性，我们将其分为前面的字母部分(我猜是位置和船层之类的信息) 和 后面的数字部分(应该是房间号，有意思的事情是，如果你仔细看看原始数据，你会发现，这个值大的情况下，似乎获救的可能性高一些)。\n",
    "- Pclass和Sex俩太重要了，我们试着用它们去组出一个组合属性来试试，这也是另外一种程度的细化。\n",
    "- 单加一个Child字段，Age<=12的，设为1，其余为0(你去看看数据，确实小盆友优先程度很高啊)\n",
    "- 如果名字里面有『Mrs』，而Parch>1的，我们猜测她可能是一个母亲，应该获救的概率也会提高，因此可以多加一个Mother字段，此种情况下设为1，其余情况下设为0\n",
    "- 登船港口可以考虑先去掉试试(Q和C本来就没权重，S有点诡异)\n",
    "- 把堂兄弟/兄妹 和 Parch 还有自己 个数加在一起组一个Family_size字段(考虑到大家族可能对最后的结果有影响)\n",
    "- Name是一个我们一直没有触碰的属性，我们可以做一些简单的处理，比如说男性中带某些字眼的(‘Capt’, ‘Don’, ‘Major’, ‘Sir’)可以统一到一个Title，女性也一样。\n",
    "\n",
    "大家接着往下挖掘，可能还可以想到更多可以细挖的部分。我这里先列这些了，然后我们可以使用手头上的”train_df”和”cv_df”开始试验这些feature engineering的tricks是否有效了。\n",
    "\n",
    "试验的过程比较漫长，也需要有耐心，而且我们经常会面临很尴尬的状况，就是我们灵光一闪，想到一个feature，然后坚信它一定有效，结果试验下来，效果还不如试验之前的结果。恩，需要坚持和耐心，以及不断的挖掘。\n",
    "\n",
    "我最好的结果是在『Survived~C(Pclass)+C(Title)+C(Sex)+C(Age_bucket)+C(Cabin_num_bucket)Mother+Fare+Family_Size』下取得的，结果如下(抱歉，博主君commit的时候手抖把页面关了，于是没截着图，下面这张图是在我得到最高分之后，用这次的结果重新make commission的，截了个图，得分是0.79426，不是目前我的最高分哈，因此排名木有变…)：\n",
    "\n",
    "![做完feature engineering调整之后的结果](https://www.z4a.net/images/2018/11/28/result_3.jpg)\n",
    "\n",
    "### learning curves\n",
    "\n",
    "有一个很可能发生的问题是，我们不断地做feature engineering，产生的特征越来越多，用这些特征去训练模型，会对我们的训练集拟合得越来越好，同时也可能在逐步丧失泛化能力，从而在待预测的数据上，表现不佳，也就是发生过拟合问题。\n",
    "\n",
    "从另一个角度上说，如果模型在待预测的数据上表现不佳，除掉上面说的过拟合问题，也有可能是欠拟合问题，也就是说在训练集上，其实拟合的也不是那么好。\n",
    "\n",
    "额，这个欠拟合和过拟合怎么解释呢。这么说吧：\n",
    "\n",
    "- 过拟合就像是你班那个学数学比较刻板的同学，老师讲过的题目，一字不漏全记下来了，于是老师再出一样的题目，分分钟精确出结果。but数学考试，因为总是碰到新题目，所以成绩不咋地。\n",
    "- 欠拟合就像是，咳咳，和博主level差不多的差生。连老师讲的练习题也记不住，于是连老师出一样题目复习的周测都做不好，考试更是可想而知了。\n",
    "而在机器学习的问题上，对于过拟合和欠拟合两种情形。我们优化的方式是不同的。\n",
    "\n",
    "对过拟合而言，通常以下策略对结果优化是有用的：\n",
    "\n",
    "- 做一下feature selection，挑出较好的feature的subset来做training\n",
    "- 提供更多的数据，从而弥补原始数据的bias问题，学习到的model也会更准确\n",
    "而对于欠拟合而言，我们通常需要更多的feature，更复杂的模型来提高准确度。\n",
    "\n",
    "著名的learning curve可以帮我们判定我们的模型现在所处的状态。我们以样本数为横坐标，训练和交叉验证集上的错误率作为纵坐标，两种状态分别如下两张图所示：过拟合(overfitting/high variace)，欠拟合(underfitting/high bias)\n",
    "\n",
    "![过拟合](https://www.z4a.net/images/2018/11/28/high_variance.jpg)\n",
    "![欠拟合](https://www.z4a.net/images/2018/11/28/10067a39f8c5849405a.jpg)\n",
    "\n",
    "我们也可以把错误率替换成准确率(得分)，得到另一种形式的learning curve(sklearn 里面是这么做的)。\n",
    "\n",
    "回到我们的问题，我们用scikit-learn里面的learning_curve来帮我们分辨我们模型的状态。举个例子，这里我们一起画一下我们最先得到的baseline model的learning curve。\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false,
    "id": "C76CD67974F947BC8B7F5D32E2134DD3",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.6/site-packages/sklearn/model_selection/_split.py:2053: FutureWarning: You should specify a value for 'cv' instead of relying on the default value. The default value will change from 3 to 5 in version 0.22.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/C76CD67974F947BC8B7F5D32E2134DD3/porrqjs2sy.png\">"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(0.8065696844854024, 0.018258876711338634)"
      ]
     },
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\r\n",
    "import matplotlib.pyplot as plt\r\n",
    "# from sklearn.learning_curve import learning_curve  修改以fix learning_curve DeprecationWarning\r\n",
    "from sklearn.model_selection import learning_curve\r\n",
    "\r\n",
    "# 用sklearn的learning_curve得到training_score和cv_score，使用matplotlib画出learning curve\r\n",
    "def plot_learning_curve(estimator, title, X, y, ylim=None, cv=None, n_jobs=1, \r\n",
    "                        train_sizes=np.linspace(.05, 1., 20), verbose=0, plot=True):\r\n",
    "    \"\"\"\r\n",
    "    画出data在某模型上的learning curve.\r\n",
    "    参数解释\r\n",
    "    ----------\r\n",
    "    estimator : 你用的分类器。\r\n",
    "    title : 表格的标题。\r\n",
    "    X : 输入的feature，numpy类型\r\n",
    "    y : 输入的target vector\r\n",
    "    ylim : tuple格式的(ymin, ymax), 设定图像中纵坐标的最低点和最高点\r\n",
    "    cv : 做cross-validation的时候，数据分成的份数，其中一份作为cv集，其余n-1份作为training(默认为3份)\r\n",
    "    n_jobs : 并行的的任务数(默认1)\r\n",
    "    \"\"\"\r\n",
    "    train_sizes, train_scores, test_scores = learning_curve(\r\n",
    "        estimator, X, y, cv=cv, n_jobs=n_jobs, train_sizes=train_sizes, verbose=verbose)\r\n",
    "    \r\n",
    "    train_scores_mean = np.mean(train_scores, axis=1)\r\n",
    "    train_scores_std = np.std(train_scores, axis=1)\r\n",
    "    test_scores_mean = np.mean(test_scores, axis=1)\r\n",
    "    test_scores_std = np.std(test_scores, axis=1)\r\n",
    "    \r\n",
    "    if plot:\r\n",
    "        plt.figure()\r\n",
    "        plt.title(title)\r\n",
    "        if ylim is not None:\r\n",
    "            plt.ylim(*ylim)\r\n",
    "        plt.xlabel(u\"训练样本数\")\r\n",
    "        plt.ylabel(u\"得分\")\r\n",
    "        plt.gca().invert_yaxis()\r\n",
    "        plt.grid()\r\n",
    "    \r\n",
    "        plt.fill_between(train_sizes, train_scores_mean - train_scores_std, train_scores_mean + train_scores_std, \r\n",
    "                         alpha=0.1, color=\"b\")\r\n",
    "        plt.fill_between(train_sizes, test_scores_mean - test_scores_std, test_scores_mean + test_scores_std, \r\n",
    "                         alpha=0.1, color=\"r\")\r\n",
    "        plt.plot(train_sizes, train_scores_mean, 'o-', color=\"b\", label=u\"训练集上得分\")\r\n",
    "        plt.plot(train_sizes, test_scores_mean, 'o-', color=\"r\", label=u\"交叉验证集上得分\")\r\n",
    "    \r\n",
    "        plt.legend(loc=\"best\")\r\n",
    "        \r\n",
    "        plt.draw()\r\n",
    "        plt.gca().invert_yaxis()\r\n",
    "        plt.show()\r\n",
    "    \r\n",
    "    midpoint = ((train_scores_mean[-1] + train_scores_std[-1]) + (test_scores_mean[-1] - test_scores_std[-1])) / 2\r\n",
    "    diff = (train_scores_mean[-1] + train_scores_std[-1]) - (test_scores_mean[-1] - test_scores_std[-1])\r\n",
    "    return midpoint, diff\r\n",
    "\r\n",
    "plot_learning_curve(clf, u\"学习曲线\", X, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2B1321FDECB743C78976C4E94B469645",
    "mdEditEnable": false
   },
   "source": [
    "在实际数据上看，我们得到的learning curve没有理论推导的那么光滑哈，但是可以大致看出来，训练集和交叉验证集上的得分曲线走势还是符合预期的。\n",
    "\n",
    "目前的曲线看来，我们的model并不处于overfitting的状态(overfitting的表现一般是训练集上得分高，而交叉验证集上要低很多，中间的gap比较大)。因此我们可以再做些feature engineering的工作，添加一些新产出的特征或者组合特征到模型中。\n",
    "\n",
    "## 模型融合(model ensemble)\n",
    "\n",
    "好了，终于到这一步了，我们要祭出机器学习/数据挖掘上通常最后会用到的大杀器了。恩，模型融合。\n",
    "\n",
    "『强迫症患者』打算继续喊喊口号… \n",
    "\n",
    "- **『模型融合(model ensemble)很重要！』 **\n",
    "- **『模型融合(model ensemble)很重要！』 **\n",
    "- **『模型融合(model ensemble)很重要！』 **\n",
    "\n",
    "重要的事情说三遍，恩，噢啦。\n",
    "\n",
    "先解释解释，一会儿再回到我们的问题上哈。 \n",
    "啥叫模型融合呢，我们还是举几个例子直观理解一下好了。\n",
    "\n",
    "大家都看过知识问答的综艺节目中，求助现场观众时候，让观众投票，最高的答案作为自己的答案的形式吧，每个人都有一个判定结果，最后我们相信答案在大多数人手里。\n",
    "\n",
    "再通俗一点举个例子。你和你班某数学大神关系好，每次作业都『模仿』他的，于是绝大多数情况下，他做对了，你也对了。突然某一天大神脑子犯糊涂，手一抖，写错了一个数，于是…恩，你也只能跟着错了。 \n",
    "我们再来看看另外一个场景，你和你班5个数学大神关系都很好，每次都把他们作业拿过来，对比一下，再『自己做』，那你想想，如果哪天某大神犯糊涂了，写错了，but另外四个写对了啊，那你肯定相信另外4人的是正确答案吧？\n",
    "\n",
    "最简单的模型融合大概就是这么个意思，比如分类问题，当我们手头上有一堆在同一份数据集上训练得到的分类器(比如logistic regression，SVM，KNN，random forest，神经网络)，那我们让他们都分别去做判定，然后对结果做投票统计，取票数最多的结果为最后结果。\n",
    "\n",
    "bingo，问题就这么完美的解决了。\n",
    "\n",
    "模型融合可以比较好地缓解，训练过程中产生的过拟合问题，从而对于结果的准确度提升有一定的帮助。\n",
    "\n",
    "话说回来，回到我们现在的问题。你看，我们现在只讲了logistic regression，如果我们还想用这个融合思想去提高我们的结果，我们该怎么做呢？\n",
    "\n",
    "既然这个时候模型没得选，那咱们就在数据上动动手脚咯。大家想想，如果模型出现过拟合现在，一定是在我们的训练上出现拟合过度造成的对吧。\n",
    "\n",
    "那我们干脆就不要用全部的训练集，每次取训练集的一个subset，做训练，这样，我们虽然用的是同一个机器学习算法，但是得到的模型却是不一样的；同时，因为我们没有任何一份子数据集是全的，因此即使出现过拟合，也是在子训练集上出现过拟合，而不是全体数据上，这样做一个融合，可能对最后的结果有一定的帮助。对，这就是常用的Bagging。\n",
    "\n",
    "我们用scikit-learn里面的Bagging来完成上面的思路，过程非常简单。代码如下："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false,
    "id": "0B5FC2631C72456186DCEC6929924625",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import BaggingRegressor\n",
    "\n",
    "train_df = df.filter(regex='Survived|Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass.*|Mother|Child|Family|Title')\n",
    "train_np = train_df.values\n",
    "\n",
    "# y即Survival结果\n",
    "y = train_np[:, 0]\n",
    "\n",
    "# X即特征属性值\n",
    "X = train_np[:, 1:]\n",
    "\n",
    "# fit到BaggingRegressor之中\n",
    "clf = linear_model.LogisticRegression(C=1.0, penalty='l1', tol=1e-6)\n",
    "bagging_clf = BaggingRegressor(clf, n_estimators=20, max_samples=0.8, max_features=1.0, bootstrap=True, bootstrap_features=False, n_jobs=-1)\n",
    "bagging_clf.fit(X, y)\n",
    "\n",
    "test = df_test.filter(regex='Age_.*|SibSp|Parch|Fare_.*|Cabin_.*|Embarked_.*|Sex_.*|Pclass.*|Mother|Child|Family|Title')\n",
    "predictions = bagging_clf.predict(test)\n",
    "result = pd.DataFrame({'PassengerId':data_test['PassengerId'].values, 'Survived':predictions.astype(np.int32)})\n",
    "result.to_csv(\"/home/kesci/work/savedata/titanic/logistic_regression_bagging_predictions.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false,
    "id": "7EABF7D54061460CA6CA37873DC2424F",
    "jupyter": {
     "outputs_hidden": false
    }
   },
   "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>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>893</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>894</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>895</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>896</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>897</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>898</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>899</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>900</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>901</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived\n",
       "0          892         0\n",
       "1          893         0\n",
       "2          894         0\n",
       "3          895         0\n",
       "4          896         0\n",
       "5          897         0\n",
       "6          898         0\n",
       "7          899         0\n",
       "8          900         1\n",
       "9          901         0"
      ]
     },
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.read_csv(\"/home/kesci/work/savedata/titanic/logistic_regression_bagging_predictions.csv\").head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4FD550B25C30489B8204F4C788E79F18",
    "mdEditEnable": false
   },
   "source": [
    "然后你再Make a submission，恩，发现对结果还是有帮助的。\n",
    "\n",
    "![ensemble之后的结果](https://www.z4a.net/images/2018/11/28/10063240f91a36cc150.jpg)\n",
    "\n",
    "## 总结\n",
    "文章稍微有点长，非常感谢各位耐心看到这里。 \n",
    "总结的部分，我就简短写几段，出现的话，很多在文中有对应的场景，大家有兴趣再回头看看。\n",
    "\n",
    "对于任何的机器学习问题，不要一上来就追求尽善尽美，先用自己会的算法撸一个baseline的model出来，再进行后续的分析步骤，一步步提高。\n",
    "\n",
    "在问题的结果过程中：\n",
    "\n",
    "- **『对数据的认识太重要了！』**\n",
    "- **『数据中的特殊点/离群点的分析和处理太重要了！』**\n",
    "- **『特征工程(feature engineering)太重要了！』**\n",
    "- **『模型融合(model ensemble)太重要了！』**\n",
    "\n",
    "本文中用机器学习解决问题的过程大概如下图所示： \n",
    "\n",
    "![用机器学习解决问题的过程](https://www.z4a.net/images/2018/11/28/100f3a39031e73bd0a8.jpg)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.13.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}