{ "cells": [ { "cell_type": "markdown", "metadata": { "_cell_guid": "754603d1-6faa-8c5f-4fed-2e8881c92171" }, "source": [ "**

First Bit of Code

**" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "49b13e2c-3c7d-2050-05c4-6848de1d8064" }, "outputs": [], "source": [ "# This Python 3 environment comes with many helpful analytics libraries installed\n", "# It is defined by the kaggle/python docker image: https://github.com/kaggle/docker-python\n", "# For example, here's several helpful packages to load in \n", "\n", "import numpy as np # linear algebra\n", "import pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n", "\n", "# Input data files are available in the \"../input/\" directory.\n", "# For example, running this (by clicking run or pressing Shift+Enter) will list the files in the input directory\n", "\n", "from subprocess import check_output\n", "print(check_output([\"ls\", \"../input\"]).decode(\"utf8\"))\n", "\n", "# Any results you write to the current directory are saved as output." ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "bc914698-5891-800b-028c-7d5da66b1aa1" }, "outputs": [], "source": [ "# Ignore warnings\n", "import warnings\n", "warnings.filterwarnings('ignore')\n", "\n", "# Handle table-like data and matrices\n", "import numpy as np\n", "import pandas as pd\n", "\n", "# Modelling Algorithms\n", "from sklearn.tree import DecisionTreeClassifier\n", "from sklearn.linear_model import LogisticRegression\n", "from sklearn.neighbors import KNeighborsClassifier\n", "from sklearn.naive_bayes import GaussianNB\n", "from sklearn.svm import SVC, LinearSVC\n", "from sklearn.ensemble import RandomForestClassifier , GradientBoostingClassifier\n", "\n", "# Modelling Helpers\n", "from sklearn.preprocessing import Imputer , Normalizer , scale\n", "from sklearn.cross_validation import train_test_split , StratifiedKFold\n", "from sklearn.feature_selection import RFECV\n", "\n", "# Visualisation\n", "import matplotlib as mpl\n", "import matplotlib.pyplot as plt\n", "import matplotlib.pylab as pylab\n", "import seaborn as sns\n", "\n", "# Configure visualisations\n", "%matplotlib inline\n", "mpl.style.use( 'ggplot' )\n", "sns.set_style( 'white' )\n", "pylab.rcParams[ 'figure.figsize' ] = 8 , 6" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "fef8e781-76b3-4b58-0bf8-c01238327695" }, "outputs": [], "source": [ "def plot_histograms( df , variables , n_rows , n_cols ):\n", " fig = plt.figure( figsize = ( 16 , 12 ) )\n", " for i, var_name in enumerate( variables ):\n", " ax=fig.add_subplot( n_rows , n_cols , i+1 )\n", " df[ var_name ].hist( bins=10 , ax=ax )\n", " ax.set_title( 'Skew: ' + str( round( float( df[ var_name ].skew() ) , ) ) ) # + ' ' + var_name ) #var_name+\" Distribution\")\n", " ax.set_xticklabels( [] , visible=False )\n", " ax.set_yticklabels( [] , visible=False )\n", " fig.tight_layout() # Improves appearance a bit.\n", " plt.show()\n", "\n", "def plot_distribution( df , var , target , **kwargs ):\n", " row = kwargs.get( 'row' , None )\n", " col = kwargs.get( 'col' , None )\n", " facet = sns.FacetGrid( df , hue=target , aspect=4 , row = row , col = col )\n", " facet.map( sns.kdeplot , var , shade= True )\n", " facet.set( xlim=( 0 , df[ var ].max() ) )\n", " facet.add_legend()\n", "\n", "def plot_categories( df , cat , target , **kwargs ):\n", " row = kwargs.get( 'row' , None )\n", " col = kwargs.get( 'col' , None )\n", " facet = sns.FacetGrid( df , row = row , col = col )\n", " facet.map( sns.barplot , cat , target )\n", " facet.add_legend()\n", "\n", "def plot_correlation_map( df ):\n", " corr = titanic.corr()\n", " _ , ax = plt.subplots( figsize =( 12 , 10 ) )\n", " cmap = sns.diverging_palette( 220 , 10 , as_cmap = True )\n", " _ = sns.heatmap(\n", " corr, \n", " cmap = cmap,\n", " square=True, \n", " cbar_kws={ 'shrink' : .9 }, \n", " ax=ax, \n", " annot = True, \n", " annot_kws = { 'fontsize' : 12 }\n", " )\n", "\n", "def describe_more( df ):\n", " var = [] ; l = [] ; t = []\n", " for x in df:\n", " var.append( x )\n", " l.append( len( pd.value_counts( df[ x ] ) ) )\n", " t.append( df[ x ].dtypes )\n", " levels = pd.DataFrame( { 'Variable' : var , 'Levels' : l , 'Datatype' : t } )\n", " levels.sort_values( by = 'Levels' , inplace = True )\n", " return levels\n", "\n", "def plot_variable_importance( X , y ):\n", " tree = DecisionTreeClassifier( random_state = 99 )\n", " tree.fit( X , y )\n", " plot_model_var_imp( tree , X , y )\n", " \n", "def plot_model_var_imp( model , X , y ):\n", " imp = pd.DataFrame( \n", " model.feature_importances_ , \n", " columns = [ 'Importance' ] , \n", " index = X.columns \n", " )\n", " imp = imp.sort_values( [ 'Importance' ] , ascending = True )\n", " imp[ : 10 ].plot( kind = 'barh' )\n", " print (model.score( X , y ))\n", " " ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "ddb17a98-cbb8-1e80-94fa-02b9651b732f" }, "outputs": [], "source": [ "# get titanic & test csv files as a DataFrame\n", "train = pd.read_csv(\"../input/train.csv\")\n", "test = pd.read_csv(\"../input/test.csv\")\n", "\n", "full = train.append( test , ignore_index = True )\n", "titanic = full[ :891 ]\n", "\n", "del train , test\n", "\n", "print ('Datasets:' , 'full:' , full.shape , 'titanic:' , titanic.shape)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "00a6c5bb-b090-4600-cf7e-1bd0bd66b0d2" }, "outputs": [], "source": [ "titanic.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "8038f04c-80a4-5305-15a9-fb2e5b36b652" }, "outputs": [], "source": [ "titanic.describe()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "a23b63eb-527c-4681-8cf4-83cf4644d58e" }, "outputs": [], "source": [ "plot_correlation_map( titanic )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "3dc51c1e-d808-6258-7744-7a3093b09265" }, "outputs": [], "source": [ "# Plot distributions of Age of passangers who survived or did not survive\n", "plot_distribution( titanic , var = 'Age' , target = 'Survived' , row = 'Sex' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "a5017228-b0bf-731b-a07b-fed74a290d82" }, "outputs": [], "source": [ "# Plot distributions of Age of passangers who survived or did not survive\n", "plot_distribution( titanic , var = 'Fare' , target = 'Survived' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "bd70be88-e0ad-e5a6-3aa2-ba6d34b61b32" }, "outputs": [], "source": [ "# Plot survival rate by Embarked\n", "plot_categories( titanic , cat = 'Embarked' , target = 'Survived' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "1929b6e4-c79b-2a94-7270-cabf182018c1" }, "outputs": [], "source": [ "# Excersise 2\n", "# Plot survival rate by Sex\n", "plot_categories( titanic , cat = 'Sex' , target = 'Survived' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "a3c731e7-5c7e-6a72-ab34-08cd00f86b1a" }, "outputs": [], "source": [ "# Excersise 3\n", "# Plot survival rate by Pclass\n", "plot_categories( titanic , cat = 'Pclass' , target = 'Survived' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "d95cbb52-8ae0-af11-a784-b4ef0dbb68b4" }, "outputs": [], "source": [ "# Excersise 4\n", "# Plot survival rate by SibSp\n", "plot_categories( titanic , cat = 'SibSp' , target = 'Survived' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "41b5caed-cca8-398a-3fa2-30789b6e2c8a" }, "outputs": [], "source": [ "# Excersise 5\n", "# Plot survival rate by Parch\n", "plot_categories( titanic , cat = 'Parch' , target = 'Survived' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "9144f6f1-9db1-c16d-ad4e-30199756d3d7" }, "outputs": [], "source": [ "# Transform Sex into binary values 0 and 1\n", "sex = pd.Series( np.where( full.Sex == 'male' , 1 , 0 ) , name = 'Sex' )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "a9dd9909-cbc2-cefd-443d-f6e59d627e32" }, "outputs": [], "source": [ "# Create a new variable for every unique value of Embarked\n", "embarked = pd.get_dummies( full.Embarked , prefix='Embarked' )\n", "embarked.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "1f0224df-65c2-7a6e-16c6-ab9cd3c41f2e" }, "outputs": [], "source": [ "# Create a new variable for every unique value of Embarked\n", "pclass = pd.get_dummies( full.Pclass , prefix='Pclass' )\n", "pclass.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "553e3d8d-3fbc-7618-024e-c0b29ae0570d" }, "outputs": [], "source": [ "# Create dataset\n", "imputed = pd.DataFrame()\n", "\n", "# Fill missing values of Age with the average of Age (mean)\n", "imputed[ 'Age' ] = full.Age.fillna( full.Age.mean() )\n", "\n", "# Fill missing values of Fare with the average of Fare (mean)\n", "imputed[ 'Fare' ] = full.Fare.fillna( full.Fare.mean() )\n", "\n", "imputed.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "66878c1d-ab87-951c-b246-ac3c37557096" }, "outputs": [], "source": [ "title = pd.DataFrame()\n", "# we extract the title from each name\n", "title[ 'Title' ] = full[ 'Name' ].map( lambda name: name.split( ',' )[1].split( '.' )[0].strip() )\n", "\n", "# a map of more aggregated titles\n", "Title_Dictionary = {\n", " \"Capt\": \"Officer\",\n", " \"Col\": \"Officer\",\n", " \"Major\": \"Officer\",\n", " \"Jonkheer\": \"Royalty\",\n", " \"Don\": \"Royalty\",\n", " \"Sir\" : \"Royalty\",\n", " \"Dr\": \"Officer\",\n", " \"Rev\": \"Officer\",\n", " \"the Countess\":\"Royalty\",\n", " \"Dona\": \"Royalty\",\n", " \"Mme\": \"Mrs\",\n", " \"Mlle\": \"Miss\",\n", " \"Ms\": \"Mrs\",\n", " \"Mr\" : \"Mr\",\n", " \"Mrs\" : \"Mrs\",\n", " \"Miss\" : \"Miss\",\n", " \"Master\" : \"Master\",\n", " \"Lady\" : \"Royalty\"\n", "\n", " }\n", "\n", "# we map each title\n", "title[ 'Title' ] = title.Title.map( Title_Dictionary )\n", "title = pd.get_dummies( title.Title )\n", "#title = pd.concat( [ title , titles_dummies ] , axis = 1 )\n", "\n", "title.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "302aa6ad-5fa0-8579-8748-178ee85e26d8" }, "outputs": [], "source": [ "cabin = pd.DataFrame()\n", "\n", "# replacing missing cabins with U (for Uknown)\n", "cabin[ 'Cabin' ] = full.Cabin.fillna( 'U' )\n", "\n", "# mapping each Cabin value with the cabin letter\n", "cabin[ 'Cabin' ] = cabin[ 'Cabin' ].map( lambda c : c[0] )\n", "\n", "# dummy encoding ...\n", "cabin = pd.get_dummies( cabin['Cabin'] , prefix = 'Cabin' )\n", "\n", "cabin.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "ee9d5ae4-5136-a2af-f496-0a9fbbacb9f8" }, "outputs": [], "source": [ "# a function that extracts each prefix of the ticket, returns 'XXX' if no prefix (i.e the ticket is a digit)\n", "def cleanTicket( ticket ):\n", " ticket = ticket.replace( '.' , '' )\n", " ticket = ticket.replace( '/' , '' )\n", " ticket = ticket.split()\n", " ticket = map( lambda t : t.strip() , ticket )\n", " ticket = list(filter( lambda t : not t.isdigit() , ticket ))\n", " if len( ticket ) > 0:\n", " return ticket[0]\n", " else: \n", " return 'XXX'\n", "\n", "ticket = pd.DataFrame()\n", "\n", "# Extracting dummy variables from tickets:\n", "ticket[ 'Ticket' ] = full[ 'Ticket' ].map( cleanTicket )\n", "ticket = pd.get_dummies( ticket[ 'Ticket' ] , prefix = 'Ticket' )\n", "\n", "ticket.shape\n", "ticket.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "65d8835d-331c-b12a-85a3-39efb84d4bc5" }, "outputs": [], "source": [ "family = pd.DataFrame()\n", "\n", "# introducing a new feature : the size of families (including the passenger)\n", "family[ 'FamilySize' ] = full[ 'Parch' ] + full[ 'SibSp' ] + 1\n", "\n", "# introducing other features based on the family size\n", "family[ 'Family_Single' ] = family[ 'FamilySize' ].map( lambda s : 1 if s == 1 else 0 )\n", "family[ 'Family_Small' ] = family[ 'FamilySize' ].map( lambda s : 1 if 2 <= s <= 4 else 0 )\n", "family[ 'Family_Large' ] = family[ 'FamilySize' ].map( lambda s : 1 if 5 <= s else 0 )\n", "\n", "family.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "fad87cc9-cf99-6389-5eaa-d5cd3ef5aea8" }, "outputs": [], "source": [ "full_X = pd.concat( [ imputed , embarked , cabin , sex ] , axis=1 )\n", "full_X.head()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "737b72c1-b060-3800-c297-0e1c624fc69e" }, "outputs": [], "source": [ "# Create all datasets that are necessary to train, validate and test models\n", "train_valid_X = full_X[ 0:891 ]\n", "train_valid_y = titanic.Survived\n", "test_X = full_X[ 891: ]\n", "train_X , valid_X , train_y , valid_y = train_test_split( train_valid_X , train_valid_y , train_size = .7 )\n", "\n", "print (full_X.shape , train_X.shape , valid_X.shape , train_y.shape , valid_y.shape , test_X.shape)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "3195f4ee-e928-dbf5-7598-1853acb86008" }, "outputs": [], "source": [ "plot_variable_importance(train_X, train_y)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "8d73595d-79dd-c8a7-cc89-df46177f2b8c" }, "outputs": [], "source": [ "model = RandomForestClassifier(n_estimators=100)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "74f928da-4c21-b426-2dc2-c35d76f001a5" }, "outputs": [], "source": [ "model = SVC()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "69e6366c-ad1f-048c-8753-64c069b1e5f3" }, "outputs": [], "source": [ "model = GradientBoostingClassifier()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "6757437e-b1c5-b480-4924-5351d03c016f" }, "outputs": [], "source": [ "model = KNeighborsClassifier(n_neighbors = 3)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "cd9ced88-1075-8ffc-734f-8f13947b468d" }, "outputs": [], "source": [ "model = LogisticRegression()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "d0b46cf6-e849-4f10-38b0-a1d61689c17e" }, "outputs": [], "source": [ "model.fit( train_X , train_y )" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "9853e1c6-70bc-e95c-f94e-b012b6210495" }, "outputs": [], "source": [ "# Score the model\n", "print (model.score( train_X , train_y ) , model.score( valid_X , valid_y ))" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "9e5e2a08-5f89-3178-1358-1ea255a29c7d" }, "outputs": [], "source": [ "#plot_model_var_imp(model, train_X, train_y)" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "3966cfc4-0bd1-b2c2-6112-2bb9bbb5c66c" }, "outputs": [], "source": [ "rfecv = RFECV( estimator = model , step = 1 , cv = StratifiedKFold( train_y , 2 ) , scoring = 'accuracy' )\n", "rfecv.fit( train_X , train_y )\n", "\n", "#print (rfecv.score( train_X , train_y ) , rfecv.score( valid_X , valid_y ))\n", "#print( \"Optimal number of features : %d\" % rfecv.n_features_ )\n", "\n", "# Plot number of features VS. cross-validation scores\n", "#plt.figure()\n", "#plt.xlabel( \"Number of features selected\" )\n", "#plt.ylabel( \"Cross validation score (nb of correct classifications)\" )\n", "#plt.plot( range( 1 , len( rfecv.grid_scores_ ) + 1 ) , rfecv.grid_scores_ )\n", "#plt.show()" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "_cell_guid": "4b29b66b-0183-969e-75b0-34fd6c14a03f" }, "outputs": [], "source": [ "test_Y = model.predict( test_X )\n", "passenger_id = full[891:].PassengerId\n", "test = pd.DataFrame( { 'PassengerId': passenger_id , 'Survived': test_Y } )\n", "test.shape\n", "test.head()\n", "test.to_csv( 'titanic_pred.csv' , index = False )" ] } ], "metadata": { "_change_revision": 0, "_is_fork": false, "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": "3.6.0" } }, "nbformat": 4, "nbformat_minor": 0 }