Titanic code

Files:

• Link to data (required to run)

  • test.csv

  • train.csv

import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

import seaborn as sns

from sklearn.preprocessing import StandardScaler

print("done")

df_train = pd.read_csv('train.csv')

df_train["train_test"] = 1 # This is training data, since 1 = training and 0 = testing

df_train.head()

df_test = pd.read_csv('test.csv')

df_test["train_test"] = 0 # This is testing data, since 1 = training and 0 = testing

df_test["Survived"] = np.NaN

df_test.head()

df_train.info()

df_train.describe()

categorical = df_train[['Survived','Pclass','Sex','Ticket','Cabin','Embarked']]

numeric = df_train[['Age','SibSp','Parch','Fare']]

for i in numeric.columns:

plt.hist(numeric[i])

plt.title(i)

plt.show()

f, ax = plt.subplots(figsize = [25,16])

sns.heatmap(numeric.corr(),linewidths = .5, annot = True, cmap = 'YlGnBu', square = True)

pd.pivot_table(df_train, index='Survived', values = ['Age','SibSp','Parch','Fare'])

for i in categorical.columns:

sns.barplot(categorical[i].value_counts().index, categorical[i].value_counts()).set_title(i)

plt.show()

print(pd.pivot_table(df_train, index = 'Survived', columns = 'Pclass', values = 'Ticket' ,aggfunc ='count'))

print(pd.pivot_table(df_train, index = 'Survived', columns = 'Sex', values = 'Ticket' ,aggfunc ='count'))

print(pd.pivot_table(df_train, index = 'Survived', columns = 'Embarked', values = 'Ticket' ,aggfunc ='count'))

df_train["cabin_letter"] = df_train.Cabin.apply(lambda x : str(x)[0])

df_train['cabin_multiple'] = df_train.Cabin.apply(lambda x: 0 if pd.isna(x) else len(x.split(' ')))

pd.pivot_table(df_train, index = 'Survived', columns = 'cabin_multiple', values = 'Ticket' ,aggfunc ='count')

pd.pivot_table(df_train,index='Survived',columns='cabin_letter', values = 'Name', aggfunc='count')

df_train['numeric_ticket'] = df_train.Ticket.apply(lambda x: 1 if x.isnumeric() else 0)

pd.pivot_table(df_train,index='Survived',columns='numeric_ticket', values = 'Ticket', aggfunc='count')

df_train['name_title'] = df_train.Name.apply(lambda x: x.split(',')[1].split('.')[0].strip())

pd.pivot_table(df_train,index='Survived',columns='name_title', values = 'Ticket', aggfunc='count')

def title_groups(x):

if x == "the Countess" or x == "Sir" or x == "Mlle" or x == "Mme" or x == "Lady":

return "Royalty"

elif x == "Capt" or x == "Don" or x == "Jonkheer" or x == "Rev":

return "AlmostRoyalty"

elif x == "Col" or x == "Major":

return "Military"

else:

return x

df_train['name_title'] = df_train.name_title.apply(title_groups)

pd.pivot_table(df_train,index='Survived',columns='name_title', values = 'Ticket', aggfunc='count')

df_train['ticket_letters'] = df_train.Ticket.apply(lambda x: ''.join(x.split(' ')[:-1]).replace('.','').replace('/','').lower() if len(x.split(' ')[:-1]) >0 else 0)

pd.set_option("max_rows", 6)

df_train['ticket_letters'].value_counts()

all_data = pd.concat([df_train, df_test])

all_data['cabin_multiple'] = all_data.Cabin.apply(lambda x: 0 if pd.isna(x) else len(x.split(' ')))

all_data['cabin_letter'] = all_data.Cabin.apply(lambda x: str(x)[0])

all_data['numeric_ticket'] = all_data.Ticket.apply(lambda x: 1 if x.isnumeric() else 0)

all_data['ticket_letters'] = all_data.Ticket.apply(lambda x: ''.join(x.split(' ')[:-1]).replace('.','').replace('/','').lower() if len(x.split(' ')[:-1]) >0 else 0)

all_data['name_title'] = all_data.Name.apply(lambda x: x.split(',')[1].split('.')[0].strip())

all_data['name_title'] = all_data.name_title.apply(title_groups)

all_data.head()

all_data.Age = all_data.Age.fillna(df_train.Age.median())

all_data.Fare = all_data.Fare.fillna(df_train.Fare.median())

all_data.dropna(subset=['Embarked'],inplace = True)

pd.set_option("max_rows", None)

all_data.isnull().sum()

pd.set_option("max_rows", 6)

all_data['norm_fare'] = np.log(all_data.Fare+1)

all_data['norm_fare'].hist()

all_data.Pclass = all_data.Pclass.astype(str)## Scaling

all_dummies = pd.get_dummies(all_data[['Pclass','Sex','Age','SibSp','Parch','norm_fare','Embarked','cabin_letter','cabin_multiple','numeric_ticket','name_title','train_test']])

X_train = all_dummies[all_dummies.train_test == 1].drop(['train_test'], axis =1)

X_train.shape

X_test = all_dummies[all_dummies.train_test == 0].drop(['train_test'], axis =1)

X_test.shape

y_train = all_data[all_data.train_test==1].Survived

y_train.shape

scale = StandardScaler()

all_dummies_scaled = all_dummies.copy()

all_dummies_scaled[['Age','SibSp','Parch','norm_fare']]= scale.fit_transform(all_dummies_scaled[['Age','SibSp','Parch','norm_fare']])

all_dummies_scaled

X_train_scaled = all_dummies_scaled[all_dummies_scaled.train_test == 1].drop(['train_test'], axis =1)

X_test_scaled = all_dummies_scaled[all_dummies_scaled.train_test == 0].drop(['train_test'], axis =1)

y_train = all_data[all_data.train_test==1].Survived

print(y_train)

from sklearn.model_selection import cross_val_score

from sklearn import tree # DecisionTreeClassifier

from sklearn.neighbors import KNeighborsClassifier

from sklearn.ensemble import RandomForestClassifier

from sklearn import svm

from sklearn.svm import SVC

from sklearn import model_selection

from sklearn import metrics

from sklearn import linear_model # LogisticRegression

from sklearn.linear_model import LogisticRegression

from sklearn.linear_model import Perceptron

from sklearn.linear_model import SGDClassifier

from sklearn.tree import DecisionTreeClassifier

from sklearn.naive_bayes import GaussianNB

from sklearn.ensemble import AdaBoostClassifier

import xgboost as xgb

from sklearn.model_selection import GridSearchCV

from sklearn.model_selection import RandomizedSearchCV

# Winner

svc_clf = SVC(probability = True)

cv = cross_val_score(svc_clf,X_train_scaled,y_train,cv=5)

print(cv)

print(cv.mean())