Project_23/Basics.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "# Creating a Kaggle Workflow\n",
    "\n",
    "In this project, we will create a workflow for the Kaggle Titanic competition.  We will create a pipeline of functions to allow for easier experimentation and multiple iterations.\n",
    "\n",
    "In this dataset, we are given some information about passengers aboard the Titanic, as well as whether or not they survived.  Our goal is to use this data to predict the survival of the test data from Kaggle.  More information is available [here](https://www.kaggle.com/c/titanic/data)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# imports\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import seaborn as sns\n",
    "sns.set_style('white')\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.feature_selection import RFECV\n",
    "from sklearn.model_selection import cross_val_score\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore', category=FutureWarning)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# import our datasets\n",
    "# data for training\n",
    "train = pd.read_csv('train.csv')\n",
    "# data for testing, renamed holdout\n",
    "holdout = pd.read_csv('test.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": 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": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# review first few lines of train\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Processing\n",
    "\n",
    "To perform our data processing, we will begin by loading in all of our previous work to create data cleaning functions.  We will then create a function that combines these functions to process the train and holdout dataframes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# use %load to load the functions we previously built\n",
    "# %load functions.py\n",
    "def process_missing(df):\n",
    "    \"\"\"Handle various missing values from the data set\n",
    "\n",
    "    Usage\n",
    "    ------\n",
    "\n",
    "    holdout = process_missing(holdout)\n",
    "    \"\"\"\n",
    "    df[\"Fare\"] = df[\"Fare\"].fillna(train[\"Fare\"].mean())\n",
    "    df[\"Embarked\"] = df[\"Embarked\"].fillna(\"S\")\n",
    "    return df\n",
    "\n",
    "def process_age(df):\n",
    "    \"\"\"Process the Age column into pre-defined 'bins' \n",
    "\n",
    "    Usage\n",
    "    ------\n",
    "\n",
    "    train = process_age(train)\n",
    "    \"\"\"\n",
    "    df[\"Age\"] = df[\"Age\"].fillna(-0.5)\n",
    "    cut_points = [-1,0,5,12,18,35,60,100]\n",
    "    label_names = [\"Missing\",\"Infant\",\"Child\",\"Teenager\",\"Young Adult\",\"Adult\",\"Senior\"]\n",
    "    df[\"Age_categories\"] = pd.cut(df[\"Age\"],cut_points,labels=label_names)\n",
    "    return df\n",
    "\n",
    "def process_fare(df):\n",
    "    \"\"\"Process the Fare column into pre-defined 'bins' \n",
    "\n",
    "    Usage\n",
    "    ------\n",
    "\n",
    "    train = process_fare(train)\n",
    "    \"\"\"\n",
    "    cut_points = [-1,12,50,100,1000]\n",
    "    label_names = [\"0-12\",\"12-50\",\"50-100\",\"100+\"]\n",
    "    df[\"Fare_categories\"] = pd.cut(df[\"Fare\"],cut_points,labels=label_names)\n",
    "    return df\n",
    "\n",
    "def process_cabin(df):\n",
    "    \"\"\"Process the Cabin column into pre-defined 'bins' \n",
    "\n",
    "    Usage\n",
    "    ------\n",
    "\n",
    "    train process_cabin(train)\n",
    "    \"\"\"\n",
    "    df[\"Cabin_type\"] = df[\"Cabin\"].str[0]\n",
    "    df[\"Cabin_type\"] = df[\"Cabin_type\"].fillna(\"Unknown\")\n",
    "    df = df.drop('Cabin',axis=1)\n",
    "    return df\n",
    "\n",
    "def process_titles(df):\n",
    "    \"\"\"Extract and categorize the title from the name column \n",
    "\n",
    "    Usage\n",
    "    ------\n",
    "\n",
    "    train = process_titles(train)\n",
    "    \"\"\"\n",
    "    titles = {\n",
    "        \"Mr\" :         \"Mr\",\n",
    "        \"Mme\":         \"Mrs\",\n",
    "        \"Ms\":          \"Mrs\",\n",
    "        \"Mrs\" :        \"Mrs\",\n",
    "        \"Master\" :     \"Master\",\n",
    "        \"Mlle\":        \"Miss\",\n",
    "        \"Miss\" :       \"Miss\",\n",
    "        \"Capt\":        \"Officer\",\n",
    "        \"Col\":         \"Officer\",\n",
    "        \"Major\":       \"Officer\",\n",
    "        \"Dr\":          \"Officer\",\n",
    "        \"Rev\":         \"Officer\",\n",
    "        \"Jonkheer\":    \"Royalty\",\n",
    "        \"Don\":         \"Royalty\",\n",
    "        \"Sir\" :        \"Royalty\",\n",
    "        \"Countess\":    \"Royalty\",\n",
    "        \"Dona\":        \"Royalty\",\n",
    "        \"Lady\" :       \"Royalty\"\n",
    "    }\n",
    "    extracted_titles = df[\"Name\"].str.extract(' ([A-Za-z]+)\\.',expand=False)\n",
    "    df[\"Title\"] = extracted_titles.map(titles)\n",
    "    return df\n",
    "\n",
    "def create_dummies(df,column_name):\n",
    "    \"\"\"Create Dummy Columns (One Hot Encoding) from a single Column\n",
    "\n",
    "    Usage\n",
    "    ------\n",
    "\n",
    "    train = create_dummies(train,\"Age\")\n",
    "    \"\"\"\n",
    "    dummies = pd.get_dummies(df[column_name],prefix=column_name)\n",
    "    df = pd.concat([df,dummies],axis=1)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create a function that applies these functions to the dataframe:\n",
    "# process_missing(), process_age(), process_fare(), process_titles(), process_cabin()\n",
    "# also create dummies for the following columns:\n",
    "# Age_categories, Fare_categories, Title, Cabin_type, Sex\n",
    "\n",
    "def preprocessing(df):\n",
    "    df = process_missing(df)\n",
    "    df = process_age(df)\n",
    "    df = process_fare(df)\n",
    "    df = process_titles(df)\n",
    "    df = process_cabin(df)\n",
    "    \n",
    "    cols = ['Age_categories', 'Fare_categories', 'Title', \n",
    "            'Cabin_type', 'Sex']\n",
    "    for col in cols:\n",
    "        df = create_dummies(df, col)\n",
    "    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# apply the preprocessing function on train and holdout\n",
    "train = preprocessing(train)\n",
    "holdout = preprocessing(holdout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": 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>...</th>\n",
       "      <th>Cabin_type_B</th>\n",
       "      <th>Cabin_type_C</th>\n",
       "      <th>Cabin_type_D</th>\n",
       "      <th>Cabin_type_E</th>\n",
       "      <th>Cabin_type_F</th>\n",
       "      <th>Cabin_type_G</th>\n",
       "      <th>Cabin_type_T</th>\n",
       "      <th>Cabin_type_Unknown</th>\n",
       "      <th>Sex_female</th>\n",
       "      <th>Sex_male</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</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>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 43 columns</p>\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_type_B Cabin_type_C  \\\n",
       "0      0         A/5 21171   7.2500    ...               0            0   \n",
       "1      0          PC 17599  71.2833    ...               0            1   \n",
       "2      0  STON/O2. 3101282   7.9250    ...               0            0   \n",
       "3      0            113803  53.1000    ...               0            1   \n",
       "4      0            373450   8.0500    ...               0            0   \n",
       "\n",
       "  Cabin_type_D Cabin_type_E Cabin_type_F  Cabin_type_G  Cabin_type_T  \\\n",
       "0            0            0            0             0             0   \n",
       "1            0            0            0             0             0   \n",
       "2            0            0            0             0             0   \n",
       "3            0            0            0             0             0   \n",
       "4            0            0            0             0             0   \n",
       "\n",
       "   Cabin_type_Unknown  Sex_female  Sex_male  \n",
       "0                   1           0         1  \n",
       "1                   0           1         0  \n",
       "2                   1           1         0  \n",
       "3                   0           1         0  \n",
       "4                   1           0         1  \n",
       "\n",
       "[5 rows x 43 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# review data after preprocessing\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Exploration\n",
    "\n",
    "Let's begin the next step by exploring our data further.  We will begin by looking at `SibSp` and `Parch`, which we haven't yet explored."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    608\n",
       "1    209\n",
       "2     28\n",
       "4     18\n",
       "3     16\n",
       "8      7\n",
       "5      5\n",
       "Name: SibSp, dtype: int64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# look at the distribution of values in these columns\n",
    "train['SibSp'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    678\n",
       "1    118\n",
       "2     80\n",
       "5      5\n",
       "3      5\n",
       "4      4\n",
       "6      1\n",
       "Name: Parch, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['Parch'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeYAAAFsCAYAAAAZuBpxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxdb9rrJYw8wAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6a51f4bda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# review survival by number of SibSp\n",
    "sibsp_pivot = pd.pivot_table(train, index='SibSp', columns='Survived', \n",
    "                             values='PassengerId', aggfunc='count')\n",
    "sibsp_pivot.plot(kind='bar')\n",
    "plt.title('Counts of Survived by Number of Siblings and Spouses Aboard')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeYAAAFsCAYAAAAZuBpxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYBYRETEQBbOIiIiBKJhFREQM5P8DUySLjo3FJ1cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6a51f835c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# review survival by number of Parch\n",
    "parch_pivot = pd.pivot_table(train, index='Parch', columns='Survived', \n",
    "                             values='PassengerId', aggfunc='count')\n",
    "parch_pivot.plot(kind='bar')\n",
    "plt.title('Counts of Survived by Number of Parents and Children Aboard')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     537\n",
       "1     161\n",
       "2     102\n",
       "3      29\n",
       "5      22\n",
       "4      15\n",
       "6      12\n",
       "10      7\n",
       "7       6\n",
       "Name: family, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create a new column for total family members aboard\n",
    "data_explore = train.copy()\n",
    "data_explore['family'] = data_explore['SibSp'] + data_explore['Parch']\n",
    "data_explore['family'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeYAAAFzCAYAAADrOKo/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f6a4fdf4908>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# review survival rates based on this new feature\n",
    "family_pivot = pd.pivot_table(data_explore, index='family', columns='Survived', \n",
    "                             values='PassengerId', aggfunc='count')\n",
    "family_pivot.plot(kind='bar')\n",
    "plt.title('Counts of Survived by Number of Family Aboard')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above, we can see that for people traveling alone, about twice as many died as survived.  For those will small numbers of family aboard, slightly more survived than died.  For those with larger numbers of family members on board (four or more), however, more died than survived.\n",
    "\n",
    "It appears that we can create some new features for modeling - one to indicate if someone is traveling alone, one to indicate a small number of family on board, and one to indicate a large number of family on board."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Engineer New Features\n",
    "\n",
    "We can now create a function to create these new features that we identified above.  Once we create this function, we can then apply it to both train and holdout."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create function to create three indicator variables:\n",
    "# is_alone - no family on board\n",
    "# small_group - one to three family members on board\n",
    "# large_group - more than three family members on board\n",
    "\n",
    "def process_family(df):\n",
    "    df['family'] = df[['SibSp','Parch']].sum(axis=1)\n",
    "    df['is_alone'] = 0\n",
    "    df.loc[df['family'] == 0, 'is_alone'] = 1\n",
    "    df['small_group'] = 0\n",
    "    df.loc[(df['family'] > 0) & (df['family'] < 4), 'small_group'] = 1\n",
    "    df['large_group'] = 0\n",
    "    df.loc[df['family'] > 3, 'large_group'] = 1\n",
    "    df = df.drop(['family'], axis=1)\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# apply the function to train and holdout\n",
    "train = process_family(train)\n",
    "holdout = process_family(holdout)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": 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>is_alone</th>\n",
       "      <th>small_group</th>\n",
       "      <th>large_group</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   is_alone  small_group  large_group\n",
       "0         0            1            0\n",
       "1         0            1            0\n",
       "2         1            0            0\n",
       "3         0            1            0\n",
       "4         1            0            0"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# confirm the results of our function\n",
    "train[['is_alone', 'small_group', 'large_group']].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Selection\n",
    "\n",
    "The next step in our process is to select the best performing features that we would like to use in our final model.  We will use scikit-learn's RFECV (recursive feature elimination and cross validation) to automate the selection of the best features.  Using this requires an estimator, and since we previously discovered that our Random Forest model performed best (out of a Logistic model, k-Nearest Neighbors model, and Random Forest model), we will use that as the estimator.\n",
    "\n",
    "We will create a function for this feature selection process that will pare down our columns to usable ones and use RFECV to determine the best features out of those that remain."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create a function for feature selection\n",
    "def select_features(df):\n",
    "    # remove columns that are not numeric\n",
    "    df = df.select_dtypes(include='number')\n",
    "    # remove columns with any missing values\n",
    "    df = df.dropna(axis=1)\n",
    "    \n",
    "    # create all_X (our features) and all_y (our target)\n",
    "    all_X = df.drop(['PassengerId', 'Survived'], axis=1)\n",
    "    all_y = df['Survived']\n",
    "    \n",
    "    # perform recursive feature elimination\n",
    "    estimator = RandomForestClassifier(random_state=1)\n",
    "    selector = RFECV(estimator, cv=10)\n",
    "    selector = selector.fit(all_X, all_y)\n",
    "    \n",
    "    # collect the selected features as a list\n",
    "    best_features = list(all_X.columns[selector.support_])\n",
    "    # print out the features\n",
    "    print('Best Features:\\n', best_features)\n",
    "    \n",
    "    return best_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Best Features:\n",
      " ['Pclass', 'Age', 'SibSp', 'Parch', 'Fare', 'Age_categories_Missing', 'Age_categories_Infant', 'Age_categories_Teenager', 'Age_categories_Young Adult', 'Age_categories_Adult', 'Age_categories_Senior', 'Fare_categories_0-12', 'Fare_categories_12-50', 'Fare_categories_50-100', 'Fare_categories_100+', 'Title_Master', 'Title_Miss', 'Title_Mr', 'Title_Mrs', 'Title_Officer', 'Cabin_type_C', 'Cabin_type_D', 'Cabin_type_E', 'Cabin_type_Unknown', 'Sex_female', 'Sex_male', 'is_alone', 'small_group', 'large_group']\n"
     ]
    }
   ],
   "source": [
    "# use the select_features function on train to determine best features\n",
    "cols = select_features(train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Selection and Tuning\n",
    "\n",
    "The next step in the process is to perform model selection and tuning.  We will do this by writing a function to test out three different models and use grid search to find the best parameters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# create a function for model selection and tuning\n",
    "\n",
    "def select_model(df, features):\n",
    "    # create all_X (our features) and all_y (our target)\n",
    "    all_X = df[features]\n",
    "    all_y = df['Survived']\n",
    "    \n",
    "    # create list of dictionaries storing model name, estimator, and hyperparameters\n",
    "    model_list = [\n",
    "        {\n",
    "            'name': 'Logistic Regression',\n",
    "            'estimator': LogisticRegression(),\n",
    "            'hyperparameters': \n",
    "            {\n",
    "                'solver': ['newton-cg', 'lbfgs', 'liblinear']\n",
    "            }\n",
    "        },\n",
    "        {\n",
    "            'name': 'k-Nearest Neighbors Classifier',\n",
    "            'estimator': KNeighborsClassifier(),\n",
    "            'hyperparameters':\n",
    "            {\n",
    "                'n_neighbors': range(1,20,2),\n",
    "                'weights': ['distance', 'uniform'],\n",
    "                'algorithm': ['ball_tree', 'kd_tree', 'brute'],\n",
    "                'p': [1,2]\n",
    "            }\n",
    "        },\n",
    "        {\n",
    "            'name': 'Random Forest Classifier',\n",
    "            'estimator': RandomForestClassifier(random_state=1),\n",
    "            'hyperparameters':\n",
    "            {\n",
    "                'n_estimators': [4, 6, 9],\n",
    "                'criterion': ['entropy', 'gini'],\n",
    "                'max_depth': [2, 5, 10],\n",
    "                'max_features': ['log2', 'sqrt'],\n",
    "                'min_samples_leaf': [1, 5, 8],\n",
    "                'min_samples_split': [2, 3, 5]\n",
    "            }\n",
    "        }\n",
    "    ]\n",
    "    \n",
    "    # iterate over the list of dictionaries of models\n",
    "    for model in model_list:\n",
    "        print('Model: ', model['name'])\n",
    "        # instantiate grid search\n",
    "        grid = GridSearchCV(model['estimator'], \n",
    "                            param_grid = model['hyperparameters'],\n",
    "                            cv = 10)\n",
    "        # fit and store best parameters, score, and estimator\n",
    "        grid.fit(all_X, all_y)\n",
    "        model['best_params'] = grid.best_params_\n",
    "        model['best_score'] = grid.best_score_\n",
    "        model['best_estimator'] = grid.best_estimator_\n",
    "        # print the best parameters and score\n",
    "        print('best parameters: ', model['best_params'])\n",
    "        print('best score: ', model['best_score'])\n",
    "        print('\\n')\n",
    "    \n",
    "    # return list of dictionaries\n",
    "    return model_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model:  Logistic Regression\n",
      "best parameters:  {'solver': 'newton-cg'}\n",
      "best score:  0.8237934904601572\n",
      "\n",
      "\n",
      "Model:  k-Nearest Neighbors Classifier\n",
      "best parameters:  {'p': 1, 'n_neighbors': 3, 'algorithm': 'kd_tree', 'weights': 'uniform'}\n",
      "best score:  0.7890011223344556\n",
      "\n",
      "\n",
      "Model:  Random Forest Classifier\n",
      "best parameters:  {'n_estimators': 9, 'max_depth': 10, 'min_samples_split': 5, 'criterion': 'entropy', 'max_features': 'log2', 'min_samples_leaf': 1}\n",
      "best score:  0.8428731762065096\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# use the above function to determine the best models for train\n",
    "# use the features we identified above in feature selection\n",
    "best_models = select_model(train, cols)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the above, we see that the Random Forest Model is still the model that is scoring highest, with the Logistic Regression as a close second."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Submitting to Kaggle\n",
    "\n",
    "The last step in the workflow is to submit a predictions file to Kaggle.  To do so, we will select our model, make predictions on the holdout data, create a submissions dataframe in the format specified by Kaggle, and save that dataframe as a CSV file.\n",
    "\n",
    "Let's create a function to perform these steps."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create the submission file function\n",
    "\n",
    "def save_submission_file(model, features, filename='submission.csv'):\n",
    "    # make predictions on holdout\n",
    "    holdout_predictions = model.predict(holdout[features])\n",
    "    \n",
    "    # create submissions dataframe\n",
    "    submission = pd.DataFrame({\n",
    "        'PassengerId': holdout['PassengerId'],\n",
    "        'Survived': holdout_predictions\n",
    "    })\n",
    "    \n",
    "    # save as CSV file (with no index so no extra column)\n",
    "    submission.to_csv(filename, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# create submission file using best random forest model\n",
    "best_rf_model = best_models[2]['best_estimator']\n",
    "save_submission_file(best_rf_model, cols)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}