titanic_survival_exploration.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 机器学习工程师纳米学位\n",
    "## 机器学习基础\n",
    "## 项目 0: 预测泰坦尼克号乘客生还率\n",
    "\n",
    "1912年，泰坦尼克号在第一次航行中就与冰山相撞沉没，导致了大部分乘客和船员身亡。在这个入门项目中，我们将探索部分泰坦尼克号旅客名单，来确定哪些特征可以最好地预测一个人是否会生还。为了完成这个项目，你将需要实现几个基于条件的预测并回答下面的问题。我们将根据代码的完成度和对问题的解答来对你提交的项目的进行评估。 \n",
    "\n",
    "> **提示**：这样的文字将会指导你如何使用 iPython Notebook 来完成项目。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "点击[这里](https://github.com/udacity/machine-learning/blob/master/projects/titanic_survival_exploration/titanic_survival_exploration.ipynb)查看本文件的英文版本。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 了解数据\n",
    "\n",
    "当我们开始处理泰坦尼克号乘客数据时，会先导入我们需要的功能模块以及将数据加载到 `pandas` DataFrame。运行下面区域中的代码加载数据，并使用 `.head()` 函数显示前几项乘客数据。 \n",
    "\n",
    "> **提示**：你可以通过单击代码区域，然后使用键盘快捷键 **Shift+Enter** 或 **Shift+ Return** 来运行代码。或者在选择代码后使用**播放**（run cell）按钮执行代码。像这样的 MarkDown 文本可以通过双击编辑，并使用这些相同的快捷键保存。[Markdown](http://daringfireball.net/projects/markdown/syntax) 允许你编写易读的纯文本并且可以转换为 HTML。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 检查你的Python版本\n",
    "from sys import version_info\n",
    "if version_info.major != 2 and version_info.minor != 7:\n",
    "    raise Exception('请使用Python 2.7来完成此项目')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\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  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "# 数据可视化代码\n",
    "from titanic_visualizations import survival_stats\n",
    "from IPython.display import display\n",
    "%matplotlib inline\n",
    "\n",
    "# 加载数据集\n",
    "in_file = 'titanic_data.csv'\n",
    "full_data = pd.read_csv(in_file)\n",
    "\n",
    "# 显示数据列表中的前几项乘客数据\n",
    "display(full_data.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "从泰坦尼克号的数据样本中，我们可以看到船上每位旅客的特征\n",
    "\n",
    "- **Survived**：是否存活（0代表否，1代表是）\n",
    "- **Pclass**：社会阶级（1代表上层阶级，2代表中层阶级，3代表底层阶级）\n",
    "- **Name**：船上乘客的名字\n",
    "- **Sex**：船上乘客的性别\n",
    "- **Age**:船上乘客的年龄（可能存在 `NaN`）\n",
    "- **SibSp**：乘客在船上的兄弟姐妹和配偶的数量\n",
    "- **Parch**：乘客在船上的父母以及小孩的数量\n",
    "- **Ticket**：乘客船票的编号\n",
    "- **Fare**：乘客为船票支付的费用\n",
    "- **Cabin**：乘客所在船舱的编号（可能存在 `NaN`）\n",
    "- **Embarked**：乘客上船的港口（C 代表从 Cherbourg 登船，Q 代表从 Queenstown 登船，S 代表从 Southampton 登船）\n",
    "\n",
    "因为我们感兴趣的是每个乘客或船员是否在事故中活了下来。可以将 **Survived** 这一特征从这个数据集移除，并且用一个单独的变量 `outcomes` 来存储。它也做为我们要预测的目标。\n",
    "\n",
    "运行该代码，从数据集中移除 **Survived** 这个特征，并将它存储在变量 `outcomes` 中。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\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>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>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>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>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>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>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  Pclass                                               Name  \\\n",
       "0            1       3                            Braund, Mr. Owen Harris   \n",
       "1            2       1  Cumings, Mrs. John Bradley (Florence Briggs Th...   \n",
       "2            3       3                             Heikkinen, Miss. Laina   \n",
       "3            4       1       Futrelle, Mrs. Jacques Heath (Lily May Peel)   \n",
       "4            5       3                           Allen, Mr. William Henry   \n",
       "\n",
       "      Sex   Age  SibSp  Parch            Ticket     Fare Cabin Embarked  \n",
       "0    male  22.0      1      0         A/5 21171   7.2500   NaN        S  \n",
       "1  female  38.0      1      0          PC 17599  71.2833   C85        C  \n",
       "2  female  26.0      0      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3  female  35.0      1      0            113803  53.1000  C123        S  \n",
       "4    male  35.0      0      0            373450   8.0500   NaN        S  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 从数据集中移除 'Survived' 这个特征，并将它存储在一个新的变量中。\n",
    "outcomes = full_data['Survived']\n",
    "data = full_data.drop('Survived', axis = 1)\n",
    "\n",
    "# 显示已移除 'Survived' 特征的数据集\n",
    "display(data.head())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "这个例子展示了如何将泰坦尼克号的 **Survived** 数据从 DataFrame 移除。注意到 `data`（乘客数据）和 `outcomes` （是否存活）现在已经匹配好。这意味着对于任何乘客的 `data.loc[i]` 都有对应的存活的结果 `outcome[i]`。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 计算准确率\n",
    "为了验证我们预测的结果，我们需要一个标准来给我们的预测打分。因为我们最感兴趣的是我们预测的**准确率**，既正确预测乘客存活的比例。运行下面的代码来创建我们的 `accuracy_score` 函数以对前五名乘客的预测来做测试。\n",
    "\n",
    "**思考题**：在前五个乘客中，如果我们预测他们全部都存活，你觉得我们预测的准确率是多少？"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions have an accuracy of 60.00%.\n"
     ]
    }
   ],
   "source": [
    "def accuracy_score(truth, pred):\n",
    "    \"\"\" 返回 pred 相对于 truth 的准确率 \"\"\"\n",
    "    \n",
    "    # 确保预测的数量与结果的数量一致\n",
    "    if len(truth) == len(pred): \n",
    "        \n",
    "        # 计算预测准确率（百分比）\n",
    "        return \"Predictions have an accuracy of {:.2f}%.\".format((truth == pred).mean()*100)\n",
    "    \n",
    "    else:\n",
    "        return \"Number of predictions does not match number of outcomes!\"\n",
    "    \n",
    "# 测试 'accuracy_score' 函数\n",
    "predictions = pd.Series(np.ones(5, dtype = int)) #五个预测全部为1，既存活\n",
    "print accuracy_score(outcomes[:5], predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> **提示**：如果你保存 iPython Notebook，代码运行的输出也将被保存。但是，一旦你重新打开项目，你的工作区将会被重置。请确保每次都从上次离开的地方运行代码来重新生成变量和函数。\n",
    "\n",
    "### 最简单的预测\n",
    "\n",
    "如果我们要预测泰坦尼克号上的乘客是否存活，但是我们又对他们一无所知，那么最好的预测就是船上的人无一幸免。这是因为，我们可以假定当船沉没的时候大多数乘客都遇难了。下面的 `predictions_0` 函数就预测船上的乘客全部遇难。  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predictions_0(data):\n",
    "    \"\"\" 不考虑任何特征，预测所有人都无法生还 \"\"\"\n",
    "\n",
    "    predictions = []\n",
    "    for _, passenger in data.iterrows():\n",
    "        \n",
    "        # 预测 'passenger' 的生还率\n",
    "        predictions.append(0)\n",
    "    \n",
    "    # 返回预测结果\n",
    "    return pd.Series(predictions)\n",
    "\n",
    "# 进行预测\n",
    "predictions = predictions_0(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**问题1**：对比真实的泰坦尼克号的数据，如果我们做一个所有乘客都没有存活的预测，这个预测的准确率能达到多少？\n",
    "\n",
    "**回答**： 61.62%\n",
    "\n",
    "**提示**：运行下面的代码来查看预测的准确率。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions have an accuracy of 61.62%.\n"
     ]
    }
   ],
   "source": [
    "print accuracy_score(outcomes, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 考虑一个特征进行预测\n",
    "\n",
    "我们可以使用 `survival_stats` 函数来看看 **Sex** 这一特征对乘客的存活率有多大影响。这个函数定义在名为 `titanic_visualizations.py` 的 Python 脚本文件中，我们的项目提供了这个文件。传递给函数的前两个参数分别是泰坦尼克号的乘客数据和乘客的 生还结果。第三个参数表明我们会依据哪个特征来绘制图形。\n",
    "\n",
    "运行下面的代码绘制出依据乘客性别计算存活率的柱形图。 "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfgAAAGDCAYAAADHzQJ9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa49b9e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "survival_stats(data, outcomes, 'Sex')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "观察泰坦尼克号上乘客存活的数据统计，我们可以发现大部分男性乘客在船沉没的时候都遇难了。相反的，大部分女性乘客都在事故中**生还**。让我们以此改进先前的预测：如果乘客是男性，那么我们就预测他们遇难；如果乘客是女性，那么我们预测他们在事故中活了下来。\n",
    "\n",
    "将下面的代码补充完整，让函数可以进行正确预测。  \n",
    "\n",
    "**提示**：您可以用访问 dictionary（字典）的方法来访问船上乘客的每个特征对应的值。例如， `passenger['Sex']` 返回乘客的性别。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predictions_1(data):\n",
    "    \"\"\" 只考虑一个特征，如果是女性则生还 \"\"\"\n",
    "    \n",
    "    predictions = []\n",
    "    for _, passenger in data.iterrows():\n",
    "        \n",
    "        # TODO 1\n",
    "        # 移除下方的 'pass' 声明\n",
    "        # 输入你自己的预测条件\n",
    "        if passenger['Sex']=='female':\n",
    "            predictions.append(1)\n",
    "        else:\n",
    "            predictions.append(0)\n",
    "        \n",
    "    \n",
    "    # 返回预测结果\n",
    "    return pd.Series(predictions)\n",
    "\n",
    "# 进行预测\n",
    "predictions = predictions_1(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions have an accuracy of 79.35%.\n"
     ]
    }
   ],
   "source": [
    "print accuracy_score(outcomes, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**问题2**：当我们预测船上女性乘客全部存活，而剩下的人全部遇难，那么我们预测的准确率会达到多少？\n",
    "\n",
    "**回答**: 78.68%\n",
    "\n",
    "**提示**：你需要在下面添加一个代码区域，实现代码并运行来计算准确率。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 考虑两个特征进行预测\n",
    "\n",
    "仅仅使用乘客性别（Sex）这一特征，我们预测的准确性就有了明显的提高。现在再看一下使用额外的特征能否更进一步提升我们的预测准确度。例如，综合考虑所有在泰坦尼克号上的男性乘客：我们是否找到这些乘客中的一个子集，他们的存活概率较高。让我们再次使用 `survival_stats` 函数来看看每位男性乘客的年龄（Age）。这一次，我们将使用第四个参数来限定柱形图中只有男性乘客。\n",
    "\n",
    "运行下面这段代码，把男性基于年龄的生存结果绘制出来。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfsAAAGDCAYAAAAs+rl+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nbJSzDrwUMlwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb45fb38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "survival_stats(data, outcomes, 'Age', [\"Sex == 'male'\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "仔细观察泰坦尼克号存活的数据统计，在船沉没的时候，大部分小于10岁的男孩都活着，而大多数10岁以上的男性都随着船的沉没而**遇难**。让我们继续在先前预测的基础上构建：如果乘客是女性，那么我们就预测她们全部存活；如果乘客是男性并且小于10岁，我们也会预测他们全部存活；所有其它我们就预测他们都没有幸存。  \n",
    "\n",
    "将下面缺失的代码补充完整，让我们的函数可以实现预测。  \n",
    "**提示**: 您可以用之前 `predictions_1` 的代码作为开始来修改代码，实现新的预测函数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predictions_2(data):\n",
    "    \"\"\" 考虑两个特征: \n",
    "            - 如果是女性则生还\n",
    "            - 如果是男性并且小于10岁则生还 \"\"\"\n",
    "    \n",
    "    predictions = []\n",
    "    for _, passenger in data.iterrows():\n",
    "        \n",
    "        # TODO 2\n",
    "        # 移除下方的 'pass' 声明\n",
    "        # 输入你自己的预测条件\n",
    "        if passenger['Sex']=='female':\n",
    "            predictions.append(1)\n",
    "        else:\n",
    "            \n",
    "            if passenger['Age']< 10:\n",
    "                predictions.append(1)\n",
    "            else:\n",
    "                predictions.append(0)\n",
    "        \n",
    "    \n",
    "    \n",
    "    # 返回预测结果\n",
    "    return pd.Series(predictions)\n",
    "\n",
    "# 进行预测\n",
    "predictions = predictions_2(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions have an accuracy of 79.35%.\n"
     ]
    }
   ],
   "source": [
    "print accuracy_score(outcomes, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**问题3**：当预测所有女性以及小于10岁的男性都存活的时候，预测的准确率会达到多少？\n",
    "\n",
    "**回答**: 79.35%\n",
    "\n",
    "**提示**：你需要在下面添加一个代码区域，实现代码并运行来计算准确率。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "### 你自己的预测模型\n",
    "\n",
    "添加年龄（Age）特征与性别（Sex）的结合比单独使用性别（Sex）也提高了不少准确度。现在该你来做预测了：找到一系列的特征和条件来对数据进行划分，使得预测结果提高到80%以上。这可能需要多个特性和多个层次的条件语句才会成功。你可以在不同的条件下多次使用相同的特征。**Pclass**，**Sex**，**Age**，**SibSp** 和 **Parch** 是建议尝试使用的特征。   \n",
    "\n",
    "使用 `survival_stats` 函数来观测泰坦尼克号上乘客存活的数据统计。  \n",
    "**提示:** 要使用多个过滤条件，把每一个条件放在一个列表里作为最后一个参数传递进去。例如: `[\"Sex == 'male'\", \"Age < 18\"]`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfgAAAGDCAYAAADHzQJ9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/X9AvlP3A6gewgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xb3fe1d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "survival_stats(data, outcomes, 'SibSp',[])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "当查看和研究了图形化的泰坦尼克号上乘客的数据统计后，请补全下面这段代码中缺失的部分，使得函数可以返回你的预测。   \n",
    "在到达最终的预测模型前请确保记录你尝试过的各种特征和条件。   \n",
    "**提示:** 您可以用之前 `predictions_2` 的代码作为开始来修改代码，实现新的预测函数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "def predictions_3(data):\n",
    "    \"\"\" 考虑多个特征，准确率至少达到80% \"\"\"\n",
    "    \n",
    "    predictions = []\n",
    "    for _, passenger in data.iterrows():\n",
    "        \n",
    "        # TODO 3\n",
    "        # 移除下方的 'pass' 声明\n",
    "        # 输入你自己的预测条件\n",
    "        if passenger['SibSp']> 2:\n",
    "            predictions.append(0)\n",
    "        else:\n",
    "            if passenger['Sex']=='female':\n",
    "                predictions.append(1)\n",
    "            else:\n",
    "                if passenger['Age']< 10:\n",
    "                    predictions.append(1)\n",
    "                else:\n",
    "                    predictions.append(0)\n",
    "\n",
    "    # 返回预测结果\n",
    "    return pd.Series(predictions)\n",
    "\n",
    "# 进行预测\n",
    "predictions = predictions_3(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**问题4**：请描述你实现80%准确度的预测模型所经历的步骤。您观察过哪些特征？某些特性是否比其他特征更有帮助？你用了什么条件来预测生还结果？你最终的预测的准确率是多少？\n",
    "\n",
    "**回答**：使用了不同的特征绘图来展示出数据的分布特征，并找出约束条件，不断尝试后发现，属性SibSp>2时几乎没有幸存者，于是使用此约束条件，再加上Sex、Age属性约束后，准确率即可达到81.71%。\n",
    "\n",
    "**提示**：你需要在下面添加一个代码区域，实现代码并运行来计算准确率。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Predictions have an accuracy of 81.71%.\n"
     ]
    }
   ],
   "source": [
    "print accuracy_score(outcomes, predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 结论\n",
    "\n",
    "经过了数次对数据的探索和分类，你创建了一个预测泰坦尼克号乘客存活率的有用的算法。在这个项目中你手动地实现了一个简单的机器学习模型——决策树（*decision tree*）。决策树每次按照一个特征把数据分割成越来越小的群组（被称为 *nodes*）。每次数据的一个子集被分出来，如果分割后新子集之间的相似度比分割前更高（包含近似的标签），我们的预测也就更加准确。电脑来帮助我们做这件事会比手动做更彻底，更精确。[这个链接](http://www.r2d3.us/visual-intro-to-machine-learning-part-1/)提供了另一个使用决策树做机器学习入门的例子。  \n",
    "\n",
    "决策树是许多**监督学习**算法中的一种。在监督学习中，我们关心的是使用数据的特征并根据数据的结果标签进行预测或建模。也就是说，每一组数据都有一个真正的结果值，不论是像泰坦尼克号生存数据集一样的标签，或者是连续的房价预测。\n",
    "\n",
    "**问题5**：想象一个真实世界中应用监督学习的场景，你期望预测的结果是什么？举出两个在这个场景中能够帮助你进行预测的数据集中的特征。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "**回答**: \n",
    "机票价格的预测问题就属于应用监督学习的情形，可以使用机票价格的历史数据对模型进行训练，预测的结果是当前的机票价格。数据集的特征：飞行的里程，距起飞的时间，座位等级等。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> **注意**: 当你写完了所有**5个问题，3个TODO**。你就可以把你的 iPython Notebook 导出成 HTML 文件。你可以在菜单栏，这样导出**File -> Download as -> HTML (.html)** 把这个 HTML 和这个 iPython notebook 一起做为你的作业提交。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "翻译：毛礼建 ｜ 校译：黄强 ｜ 审译：曹晨巍"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}