Chapter3.1.1.2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 导入pandas并且更名为pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "titanic = pd.read_csv('http://biostat.mc.vanderbilt.edu/wiki/pub/Main/DataSets/titanic.txt')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "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>row.names</th>\n",
       "      <th>pclass</th>\n",
       "      <th>survived</th>\n",
       "      <th>name</th>\n",
       "      <th>age</th>\n",
       "      <th>embarked</th>\n",
       "      <th>home.dest</th>\n",
       "      <th>room</th>\n",
       "      <th>ticket</th>\n",
       "      <th>boat</th>\n",
       "      <th>sex</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1st</td>\n",
       "      <td>1</td>\n",
       "      <td>Allen, Miss Elisabeth Walton</td>\n",
       "      <td>29.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>St Louis, MO</td>\n",
       "      <td>B-5</td>\n",
       "      <td>24160 L221</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1st</td>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Miss Helen Loraine</td>\n",
       "      <td>2.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1st</td>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Mr Hudson Joshua Creighton</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(135)</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1st</td>\n",
       "      <td>0</td>\n",
       "      <td>Allison, Mrs Hudson J.C. (Bessie Waldo Daniels)</td>\n",
       "      <td>25.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>1st</td>\n",
       "      <td>1</td>\n",
       "      <td>Allison, Master Hudson Trevor</td>\n",
       "      <td>0.9167</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   row.names pclass  survived  \\\n",
       "0          1    1st         1   \n",
       "1          2    1st         0   \n",
       "2          3    1st         0   \n",
       "3          4    1st         0   \n",
       "4          5    1st         1   \n",
       "\n",
       "                                              name      age     embarked  \\\n",
       "0                     Allen, Miss Elisabeth Walton  29.0000  Southampton   \n",
       "1                      Allison, Miss Helen Loraine   2.0000  Southampton   \n",
       "2              Allison, Mr Hudson Joshua Creighton  30.0000  Southampton   \n",
       "3  Allison, Mrs Hudson J.C. (Bessie Waldo Daniels)  25.0000  Southampton   \n",
       "4                    Allison, Master Hudson Trevor   0.9167  Southampton   \n",
       "\n",
       "                         home.dest room      ticket   boat     sex  \n",
       "0                     St Louis, MO  B-5  24160 L221      2  female  \n",
       "1  Montreal, PQ / Chesterville, ON  C26         NaN    NaN  female  \n",
       "2  Montreal, PQ / Chesterville, ON  C26         NaN  (135)    male  \n",
       "3  Montreal, PQ / Chesterville, ON  C26         NaN    NaN  female  \n",
       "4  Montreal, PQ / Chesterville, ON  C22         NaN     11    male  "
      ]
     },
     "execution_count": 115,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "titanic.head()\n",
    "# 分离数据特征和预测目标"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y = titanic['survived']\n",
    "X = titanic.drop(['row.names','name','survived'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 对缺失数据进行补充"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "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>pclass</th>\n",
       "      <th>age</th>\n",
       "      <th>embarked</th>\n",
       "      <th>home.dest</th>\n",
       "      <th>room</th>\n",
       "      <th>ticket</th>\n",
       "      <th>boat</th>\n",
       "      <th>sex</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1st</td>\n",
       "      <td>29.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>St Louis, MO</td>\n",
       "      <td>B-5</td>\n",
       "      <td>24160 L221</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1st</td>\n",
       "      <td>2.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1st</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>(135)</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1st</td>\n",
       "      <td>25.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1st</td>\n",
       "      <td>0.9167</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C22</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  pclass      age     embarked                        home.dest room  \\\n",
       "0    1st  29.0000  Southampton                     St Louis, MO  B-5   \n",
       "1    1st   2.0000  Southampton  Montreal, PQ / Chesterville, ON  C26   \n",
       "2    1st  30.0000  Southampton  Montreal, PQ / Chesterville, ON  C26   \n",
       "3    1st  25.0000  Southampton  Montreal, PQ / Chesterville, ON  C26   \n",
       "4    1st   0.9167  Southampton  Montreal, PQ / Chesterville, ON  C22   \n",
       "\n",
       "       ticket   boat     sex  \n",
       "0  24160 L221      2  female  \n",
       "1         NaN    NaN  female  \n",
       "2         NaN  (135)    male  \n",
       "3         NaN    NaN  female  \n",
       "4         NaN     11    male  "
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X['age'].fillna(X['age'].mean(),inplace = True)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 119,
   "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>pclass</th>\n",
       "      <th>age</th>\n",
       "      <th>embarked</th>\n",
       "      <th>home.dest</th>\n",
       "      <th>room</th>\n",
       "      <th>ticket</th>\n",
       "      <th>boat</th>\n",
       "      <th>sex</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1st</td>\n",
       "      <td>29.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>St Louis, MO</td>\n",
       "      <td>B-5</td>\n",
       "      <td>24160 L221</td>\n",
       "      <td>2</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1st</td>\n",
       "      <td>2.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1st</td>\n",
       "      <td>30.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>(135)</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1st</td>\n",
       "      <td>25.0000</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C26</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>female</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1st</td>\n",
       "      <td>0.9167</td>\n",
       "      <td>Southampton</td>\n",
       "      <td>Montreal, PQ / Chesterville, ON</td>\n",
       "      <td>C22</td>\n",
       "      <td>UNKNOWN</td>\n",
       "      <td>11</td>\n",
       "      <td>male</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  pclass      age     embarked                        home.dest room  \\\n",
       "0    1st  29.0000  Southampton                     St Louis, MO  B-5   \n",
       "1    1st   2.0000  Southampton  Montreal, PQ / Chesterville, ON  C26   \n",
       "2    1st  30.0000  Southampton  Montreal, PQ / Chesterville, ON  C26   \n",
       "3    1st  25.0000  Southampton  Montreal, PQ / Chesterville, ON  C26   \n",
       "4    1st   0.9167  Southampton  Montreal, PQ / Chesterville, ON  C22   \n",
       "\n",
       "       ticket     boat     sex  \n",
       "0  24160 L221        2  female  \n",
       "1     UNKNOWN  UNKNOWN  female  \n",
       "2     UNKNOWN    (135)    male  \n",
       "3     UNKNOWN  UNKNOWN  female  \n",
       "4     UNKNOWN       11    male  "
      ]
     },
     "execution_count": 119,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.fillna(\"UNKNOWN\",inplace=True)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 分割数据，依然采用25%用于测试"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import  train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 122,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train,X_test,y_train,y_test = train_test_split(X,y,test_size=0.25,random_state = 33)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 类别型特征向量化"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction import DictVectorizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "vec = DictVectorizer()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = vec.fit_transform(X_train.to_dict(orient = 'record'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_test = vec.transform(X_test.to_dict(orient='record'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "474\n"
     ]
    }
   ],
   "source": [
    "print len(vec.feature_names_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[ 31.19418104   0.           0.         ...,   0.           0.           1.        ]\n",
      " [ 31.19418104   0.           0.         ...,   0.           0.           0.        ]\n",
      " [ 31.19418104   0.           0.         ...,   0.           0.           1.        ]\n",
      " ..., \n",
      " [ 12.           0.           0.         ...,   0.           0.           1.        ]\n",
      " [ 18.           0.           0.         ...,   0.           0.           1.        ]\n",
      " [ 31.19418104   0.           0.         ...,   0.           0.           1.        ]]\n"
     ]
    }
   ],
   "source": [
    "print X_train.toarray()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['age', 'boat=(103)', 'boat=(108)', 'boat=(109)', 'boat=(110)', 'boat=(122)', 'boat=(124)', 'boat=(126)', 'boat=(135)', 'boat=(140)', 'boat=(142)', 'boat=(148)', 'boat=(149)', 'boat=(15)', 'boat=(165)', 'boat=(171)', 'boat=(172)', 'boat=(174)', 'boat=(18)', 'boat=(189)', 'boat=(19)', 'boat=(190)', 'boat=(209)', 'boat=(22)', 'boat=(230)', 'boat=(232)', 'boat=(234)', 'boat=(236)', 'boat=(241)', 'boat=(245)', 'boat=(249)', 'boat=(258)', 'boat=(259)', 'boat=(263)', 'boat=(269)', 'boat=(275)', 'boat=(283)', 'boat=(286)', 'boat=(287)', 'boat=(292)', 'boat=(293)', 'boat=(307)', 'boat=(322)', 'boat=(38)', 'boat=(45)', 'boat=(46)', 'boat=(52)', 'boat=(62)', 'boat=(72)', 'boat=(75)', 'boat=(79)', 'boat=(80)', 'boat=(81)', 'boat=(89)', 'boat=(96)', 'boat=1', 'boat=10', 'boat=11', 'boat=12', 'boat=13', 'boat=14', 'boat=14/12', 'boat=14/D', 'boat=15', 'boat=16', 'boat=2', 'boat=3', 'boat=4', 'boat=5', 'boat=5/7', 'boat=6', 'boat=7', 'boat=8', 'boat=9', 'boat=A', 'boat=B', 'boat=C', 'boat=D', 'boat=UNKNOWN', 'embarked=Cherbourg', 'embarked=Queenstown', 'embarked=Southampton', 'embarked=UNKNOWN', 'home.dest=?Havana, Cuba', 'home.dest=Aberdeen / Portland, OR', 'home.dest=Albany, NY', 'home.dest=Altdorf, Switzerland', 'home.dest=Amenia, ND', 'home.dest=Antwerp, Belgium / Stanton, OH', 'home.dest=Argentina', 'home.dest=Asarum, Sweden Brooklyn, NY', 'home.dest=Ascot, Berkshire / Rochester, NY', 'home.dest=Auburn, NY', 'home.dest=Aughnacliff, Co Longford, Ireland New York, NY', 'home.dest=Australia Fingal, ND', 'home.dest=Austria', 'home.dest=Austria Niagara Falls, NY', 'home.dest=Austria-Hungary', 'home.dest=Austria-Hungary / Germantown, Philadelphia, PA', 'home.dest=Ballydehob, Co Cork, Ireland New York, NY', 'home.dest=Bangkok, Thailand / Roseville, IL', 'home.dest=Barcelona, Spain / Havana, Cuba', 'home.dest=Barre, Co Washington, VT', 'home.dest=Bayside, Queens, NY', 'home.dest=Belfast', 'home.dest=Belfast, NI', 'home.dest=Belgium  Montreal, PQ', 'home.dest=Belmont, MA', 'home.dest=Bennington, VT', 'home.dest=Berne, Switzerland / Central City, IA', 'home.dest=Birkdale, England Cleveland, Ohio', 'home.dest=Bishopstoke, Hants / Fayette Valley, ID', 'home.dest=Boston, MA', 'home.dest=Bournemouth, England', 'home.dest=Bournemouth, England Newark, NJ', 'home.dest=Bournmouth, England', 'home.dest=Brennes, Norway New York', 'home.dest=Bridgerule, Devon', 'home.dest=Bridgwater, Somerset, England', 'home.dest=Brighton, MA', 'home.dest=Bristol, Avon / Jacksonville, FL', 'home.dest=Bristol, England / New Britain, CT', 'home.dest=Bristol, England Cleveland, OH', 'home.dest=Brockton, MA', 'home.dest=Bromsgrove, England / Montreal, PQ', 'home.dest=Bronx, NY', 'home.dest=Brooklyn, NY', 'home.dest=Broomfield, Chelmsford, England', 'home.dest=Brunswick, ME', 'home.dest=Bryn Mawr, PA', 'home.dest=Bryn Mawr, PA, USA', 'home.dest=Buffalo, NY', 'home.dest=Bulgaria Chicago, IL', 'home.dest=Calgary, AB', 'home.dest=Cambridge, MA', 'home.dest=Cape Town, South Africa / Seattle, WA', 'home.dest=Catford, Kent / Detroit, MI', 'home.dest=Chelsea, London', 'home.dest=Chicago, IL', 'home.dest=Clevedon, England', 'home.dest=Co Athlone, Ireland New York, NY', 'home.dest=Co Clare, Ireland Washington, DC', 'home.dest=Co Cork, Ireland Charlestown, MA', 'home.dest=Co Limerick, Ireland Sherbrooke, PQ', 'home.dest=Co Longford, Ireland New York, NY', 'home.dest=Co Sligo, Ireland New York, NY', 'home.dest=Cooperstown, NY', 'home.dest=Cornwall', 'home.dest=Cornwall / Akron, OH', 'home.dest=Cornwall / Camden, NJ', 'home.dest=Cornwall / Clear Creek, CO', 'home.dest=Cornwall / Hancock, MI', 'home.dest=Cornwall / Houghton, MI', 'home.dest=Cornwall, England Houghton, MI', 'home.dest=Croatia', 'home.dest=Deephaven, MN / Cedar Rapids, IA', 'home.dest=Deer Lodge, MT', 'home.dest=Denmark / New York, NY', 'home.dest=Denmark Hill, Surrey / Chicago', 'home.dest=Denver, CO', 'home.dest=Detroit, MI', 'home.dest=Devon, England Wichita, KS', 'home.dest=Devonport, England', 'home.dest=Dorking, Surrey, England', 'home.dest=Dowagiac, MI', 'home.dest=Duluth, MN', 'home.dest=East Bridgewater, MA', 'home.dest=East Orange, NJ', 'home.dest=East Providence, RI', 'home.dest=Effington Rut, SD', 'home.dest=Elizabeth, NJ', 'home.dest=Elkins Park, PA', 'home.dest=Elmira, NY / Orange, NJ', 'home.dest=England', 'home.dest=England / Bennington, VT', 'home.dest=England / Detroit, MI', 'home.dest=England / Philadelphia, PA', 'home.dest=England / Sacramento, CA', 'home.dest=England / San Francisco, CA', 'home.dest=England Albion, NY', 'home.dest=England Brooklyn, NY', 'home.dest=England New York, NY', 'home.dest=England Oglesby, IL', 'home.dest=England Salt Lake City, Utah', 'home.dest=Finland / Minneapolis, MN', 'home.dest=Finland / Washington, DC', 'home.dest=Folkstone, Kent / New York, NY', 'home.dest=Fond du Lac, WI', 'home.dest=Foresvik, Norway Portland, ND', 'home.dest=Gallipolis, Ohio / ? Paris / New York', 'home.dest=Geneva, Switzerland / Radnor, PA', 'home.dest=Glasgow', 'home.dest=Glasgow / Bangor, ME', 'home.dest=Goteborg, Sweden / Rockford, IL', 'home.dest=Greece', 'home.dest=Green Bay, WI', 'home.dest=Greenport, NY', 'home.dest=Greenwich, CT', 'home.dest=Guernsey', 'home.dest=Guernsey / Elizabeth, NJ', 'home.dest=Guernsey / Montclair, NJ and/or Toledo, Ohio', 'home.dest=Guernsey, England / Edgewood, RI', 'home.dest=Gunnislake, England / Butte, MT', 'home.dest=Guntur, India / Benton Harbour, MI', 'home.dest=Haddenfield, NJ', 'home.dest=Halifax, NS', 'home.dest=Hamilton, ON', 'home.dest=Harrisburg, PA', 'home.dest=Harrow, England', 'home.dest=Harrow-on-the-Hill, Middlesex', 'home.dest=Hartford, CT', 'home.dest=Haverford, PA', 'home.dest=Haverford, PA / Cooperstown, NY', 'home.dest=Helsinki, Finland Ashtabula, Ohio', 'home.dest=Hessle, Yorks', 'home.dest=Hong Kong New York, NY', 'home.dest=Hornsey, England', 'home.dest=Hudson, NY', 'home.dest=Huntington, WV', 'home.dest=Ilford, Essex / Winnipeg, MB', 'home.dest=Ilfracombe, Devon', 'home.dest=Illinois, USA', 'home.dest=India / Pittsburgh, PA', 'home.dest=India / Rapid City, SD', 'home.dest=Indianapolis, IN', 'home.dest=Ireland', 'home.dest=Ireland Chicago, IL', 'home.dest=Ireland New York, NY', 'home.dest=Ireland Philadelphia, PA', 'home.dest=Isle of Wight, England', 'home.dest=Isleworth, England', 'home.dest=Italy Philadelphia, PA', 'home.dest=Jacksonville, FL', 'home.dest=Janjgir, India / Pennsylvania', 'home.dest=Karberg, Sweden Jerome Junction, AZ', 'home.dest=Krakoryd, Sweden Bloomington, IL', 'home.dest=Krakudden, Sweden Moune, IL', 'home.dest=Lake Arthur, Chavez County, NM', 'home.dest=Lakewood, NJ', 'home.dest=Lexington, MA', 'home.dest=Lima, Peru', 'home.dest=Little Onn Hall, Staffs', 'home.dest=Liverpool', 'home.dest=Liverpool / Montreal, PQ', 'home.dest=Liverpool, England Bedford, OH', 'home.dest=London', 'home.dest=London  Vancouver, BC', 'home.dest=London /  East Orange, NJ', 'home.dest=London / Birmingham', 'home.dest=London / Chicago, IL', 'home.dest=London / Fort Byron, NY', 'home.dest=London / Middlesex', 'home.dest=London / New York, NY', 'home.dest=London / Paris', 'home.dest=London / Winnipeg, MB', 'home.dest=London New York, NY', 'home.dest=London, England', 'home.dest=London, England Norfolk, VA', 'home.dest=Los Angeles, CA', 'home.dest=Lower Clapton, Middlesex or Erdington, Birmingham', 'home.dest=Lucca, Italy / California', 'home.dest=Lyndhurst, England', 'home.dest=Madrid, Spain', 'home.dest=Mamaroneck, NY', 'home.dest=Manchester, England', 'home.dest=Medeltorp, Sweden Chicago, IL', 'home.dest=Mexico City, Mexico', 'home.dest=Middleburg Heights, OH', 'home.dest=Milford, NH', 'home.dest=Milwaukee, WI', 'home.dest=Minneapolis, MN', 'home.dest=Montevideo, Uruguay', 'home.dest=Montreal, PQ', 'home.dest=Montreal, PQ / Chesterville, ON', 'home.dest=Moscow / Bronx, NY', 'home.dest=Mt Airy, Philadelphia, PA', 'home.dest=Myren, Sweden New York, NY', 'home.dest=New Forest, England', 'home.dest=New York, NY', 'home.dest=New York, NY /  Stamford CT', 'home.dest=New York, NY / Briarcliff Manor NY', 'home.dest=New York, NY / Greenwich CT', 'home.dest=New York, NY / Ithaca, NY', 'home.dest=Newark, NJ', 'home.dest=Nice, France', 'home.dest=North Evington, England', 'home.dest=Norway Los Angeles, CA', 'home.dest=Norwich / New York, NY', 'home.dest=Oskarshamn, Sweden Minneapolis, MN', 'home.dest=Oslo, Norway Bayonne, NJ', 'home.dest=Oslo, Norway Cameron, WI', 'home.dest=Ottawa, ON', 'home.dest=Paris', 'home.dest=Paris /  New York, NY', 'home.dest=Paris / Haiti', 'home.dest=Paris / Montreal, PQ', 'home.dest=Paris, France', 'home.dest=Paris, France / New York, NY', 'home.dest=Pennsylvania', 'home.dest=Penzance, Cornwall / Akron, OH', 'home.dest=Perkins County, SD', 'home.dest=Petworth, Sussex', 'home.dest=Philadelphia, PA', 'home.dest=Plymouth, Devon / Detroit, MI', 'home.dest=Plymouth, Dorset / Houghton, MI', 'home.dest=Plymouth, England', 'home.dest=Pomeroy, WA', 'home.dest=Portland, OR', 'home.dest=Portugal', 'home.dest=Portugal / Sau Paulo, Brazil', 'home.dest=Providence, RI', 'home.dest=Roachdale, IN', 'home.dest=Rochester, NY', 'home.dest=Rotherfield, Sussex, England Essex Co, MA', 'home.dest=Rotterdam, Netherlands', 'home.dest=Ruotsinphytaa, Finland New York, NY', 'home.dest=Russia', 'home.dest=Russia New York, NY', 'home.dest=San Francisco, CA', 'home.dest=Sarnia, ON', 'home.dest=Scituate, MA', 'home.dest=Seattle, WA', 'home.dest=Sittingbourne, England / San Diego, CA', 'home.dest=Skara, Sweden / Rockford, IL', 'home.dest=Somerset / Bernardsville, NJ', 'home.dest=Southampton', 'home.dest=Southampton / New York, NY', 'home.dest=Southington / Noank, CT', 'home.dest=Spain', 'home.dest=Spain / Havana, Cuba', 'home.dest=Springfield, MA', 'home.dest=St Andrews, Guernsey', \"home.dest=St Anne's-on-Sea, Lancashire\", 'home.dest=St Austall, Cornwall', 'home.dest=St Ives, Cornwall / Calumet, MI', 'home.dest=St Ives, Cornwall / Hancock, MI', 'home.dest=St Ives, Cornwall / Houghton, MI', 'home.dest=St James, Long Island, NY', 'home.dest=St Leonards-on-Sea, England Ohio', 'home.dest=St Louis, MO', 'home.dest=Stanton, IA', 'home.dest=Stockholm, Sweden', 'home.dest=Stockholm, Sweden / Washington, DC', 'home.dest=Stockholm, Sweden New York', 'home.dest=Strood, Kent, England Detroit, MI', 'home.dest=Sweden  Worcester, MA', 'home.dest=Sweden / Arlington, NJ', 'home.dest=Sweden Akeley, MN', 'home.dest=Sweden Chicago, IL', 'home.dest=Sweden Joliet, IL', 'home.dest=Sweden Winnipeg, MN', 'home.dest=Sweden Worcester, MA', 'home.dest=Swindon, England', 'home.dest=Sydney, Australia', 'home.dest=Syria', 'home.dest=Syria Fredericksburg, VA', 'home.dest=Syria Kent, ON', 'home.dest=Syria New York, NY', 'home.dest=Syria Ottawa, ON', 'home.dest=Syria Youngstown, OH', 'home.dest=Tokyo, Japan', 'home.dest=Toronto, ON', 'home.dest=Tranvik, Finland New York', 'home.dest=Treherbert, Cardiff, Wales', 'home.dest=Trenton, NJ', 'home.dest=Tuxedo Park, NY', 'home.dest=UNKNOWN', 'home.dest=Union Hill, NJ', 'home.dest=Vancouver, BC', 'home.dest=Walthamstow, England', 'home.dest=Warwick, England', 'home.dest=Washington, DC', 'home.dest=Waukegan, Chicago, IL', 'home.dest=West Bromwich, England Pontiac, MI', 'home.dest=West Hampstead, London / Neepawa, MB', 'home.dest=West Haven, CT', 'home.dest=West Hoboken, NJ', 'home.dest=Westcliff-on-Sea, Essex', 'home.dest=Weston-Super-Mare / Moose Jaw, SK', 'home.dest=Wiltshire, England Niagara Falls, NY', 'home.dest=Windsor, England New York, NY', 'home.dest=Winnipeg, MB', 'home.dest=Winnipeg, MN', 'home.dest=Woodford County, KY', 'home.dest=Worcester, England', 'home.dest=Worcester, MA', 'home.dest=Yoevil, England / Cottage Grove, OR', 'home.dest=Youngstown, OH', 'home.dest=Zurich, Switzerland', 'pclass=1st', 'pclass=2nd', 'pclass=3rd', 'room=A-11', 'room=A-23', 'room=A-36', 'room=B-18', 'room=B-3', 'room=B-35', 'room=B-49', 'room=B-5', 'room=B-51/3/5', 'room=B-52/4/6', 'room=B-58/60', 'room=B-82/4', 'room=B-86', 'room=B-?', 'room=C-101', 'room=C-103', 'room=C-104', 'room=C-125', 'room=C-126', 'room=C-22', 'room=C-45', 'room=C-51', 'room=C-55 (?C-95)', 'room=C-6', 'room=C-62', 'room=C-7', 'room=C-83', 'room=C-85', 'room=C-87', 'room=C-91', 'room=C-93', 'room=C-97', 'room=C-99', 'room=C26', 'room=D-21/2', 'room=D-35', 'room=D-7', 'room=D-?', 'room=E-101', 'room=E-12', 'room=E-?', 'room=E77', 'room=F-33', 'room=UNKNOWN', 'sex=female', 'sex=male', 'ticket= ', 'ticket=111361 L57 19s 7d', 'ticket=112058 Complimentary', 'ticket=112059', 'ticket=113780 L28 10s', 'ticket=11755 L39 12s', 'ticket=13502 L77', 'ticket=13529 L26 5s', 'ticket=17477 L69 6s', 'ticket=17483', 'ticket=17483 L221 15s 7d', 'ticket=17485', 'ticket=17485 L56 18s 7d', 'ticket=17582 L153 9s 3d', 'ticket=17593', 'ticket=17593 L56 18s 7d', 'ticket=17604 L39 12s', 'ticket=17608', 'ticket=17608 L262 7s 6d', 'ticket=17610 L27 15s 5d', 'ticket=17613 L27 14s 5d', 'ticket=17754', 'ticket=17754 L224 10s 6d', 'ticket=17755 L512 6s', 'ticket=229236 L13', 'ticket=230080 L26', 'ticket=230136 L39', 'ticket=24160 L221', 'ticket=248744 L13', 'ticket=250647', 'ticket=27849', 'ticket=28220 L32 10s', 'ticket=36973 L83 9s 6d', 'ticket=392091', 'ticket=7076', 'ticket=UNKNOWN']\n"
     ]
    }
   ],
   "source": [
    "print vec.get_feature_names()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 使用决策树模型依靠所有的特征进行预测，并作性能预测"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.tree import DecisionTreeClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 133,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dt = DecisionTreeClassifier(criterion = 'entropy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DecisionTreeClassifier(class_weight=None, criterion='entropy', max_depth=None,\n",
       "            max_features=None, max_leaf_nodes=None,\n",
       "            min_impurity_split=1e-07, min_samples_leaf=1,\n",
       "            min_samples_split=2, min_weight_fraction_leaf=0.0,\n",
       "            presort=False, random_state=None, splitter='best')"
      ]
     },
     "execution_count": 134,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.fit(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.82370820668693012"
      ]
     },
     "execution_count": 135,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 从sklearn中导入特征筛选器"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn import  feature_selection"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 筛选前20%的特征，使用相同配置的决策树进行预测，并且进行性能评估"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fs = feature_selection.SelectPercentile(feature_selection.chi2,percentile=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_train_fs = fs.fit_transform(X_train,y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X_test_fs = fs.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.81458966565349544"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dt.fit(X_train_fs,y_train)\n",
    "dt.score(X_test_fs,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 从sklearn中导入特征筛选器"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.cross_validation import cross_val_score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "percentiles = range(1,100,2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "results = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.85063904  0.85673057  0.87501546  0.88622964  0.86793445  0.86998557\n",
      "  0.86897547  0.87099567  0.87097506  0.86793445  0.86794475  0.86694496\n",
      "  0.86693465  0.86387343  0.86285302  0.86384251  0.86284271  0.86590394\n",
      "  0.86485261  0.86485261  0.8608122   0.86384251  0.87200577  0.86895485\n",
      "  0.86488353  0.86896516  0.86896516  0.87301587  0.87199546  0.87301587\n",
      "  0.86487322  0.86284271  0.86688312  0.86997526  0.86793445  0.86281179\n",
      "  0.87402597  0.86893424  0.87098536  0.86997526  0.86894455  0.86789322\n",
      "  0.86792414  0.86795506  0.86588332  0.86286333  0.85982272  0.85878169\n",
      "  0.86284271  0.85880231]\n"
     ]
    }
   ],
   "source": [
    "for i in percentiles:\n",
    "    fs = feature_selection.SelectPercentile(feature_selection.chi2,percentile=i)\n",
    "    X_train_fs = fs.fit_transform(X_train,y_train)\n",
    "    scores = cross_val_score(dt,X_train_fs,y_train,cv = 5)\n",
    "    results = np.append(results,scores.mean())\n",
    "print results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# 找到最佳性能的特征筛选百分比"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [],
   "source": [
    "opt = np.where(results == results.max())[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "only integer scalar arrays can be converted to a scalar index",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-152-a68562b063cb>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0;32mprint\u001b[0m \u001b[0;34m'Optimal number of features %d'\u001b[0m\u001b[0;34m%\u001b[0m\u001b[0mpercentiles\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mopt\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m: only integer scalar arrays can be converted to a scalar index"
     ]
    }
   ],
   "source": [
    "print 'Optimal number of features %d'%percentiles[opt]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pylab as pl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x110735ad0>]"
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl.plot(percentiles,results)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x11045e9d0>"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl.xlabel(\"percentiles of features\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x1106d2350>"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pl.ylabel(\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEKCAYAAAA4t9PUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nzKwSBxIzM6vEgcTMzCp5ABluXfxzYeHlAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1103c4710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pl.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": 2
}