Some experiments for kaggle titanic survivors machine learning competition (https://www.kaggle.com/c/titanic)

titanic.py

import pandas
from sklearn import linear_model, svm, tree, naive_bayes
from sklearn.model_selection import cross_val_score
import numpy as np

data = pandas.read_csv('train.csv')


def preprocess(data):
    data['Fare'] = data['Fare'].fillna(data['Fare'].mean())
    data['Age'] = data['Age'].fillna(data['Age'].mean())

    titles = {
        "Capt": "Officer",
        "Col": "Officer",
        "Major": "Officer",
        "Jonkheer": "Royalty",
        "Don": "Royalty",
        "Sir": "Royalty",
        "Dr": "Officer",
        "Rev": "Officer",
        "the Countess":"Royalty",
        "Dona": "Royalty",
        "Mme": "Mrs",
        "Mlle": "Miss",
        "Ms": "Mrs",
        "Mr": "Mr",
        "Mrs": "Mrs",
        "Miss": "Miss",
        "Master": "Master",
        "Lady": "Royalty",
    }

    data['Title'] = data['Name'].map(lambda name: name.split(',')[1].split('.')[0].strip()).map(titles).astype('category')

    # Do the one-hot encoding
    data = pandas.get_dummies(data, columns=['Sex', 'Title', 'Pclass'])

    return data

data = preprocess(data)


"""
print(data['Sex_male'].astype('category').describe())
print()
print(data['Sex_female'].astype('category').describe())
print()
print(data['Survived'].describe())
print()
print(data['Age'].describe())
print()
"""

def getX(data):
    return data.as_matrix(['Pclass_1', 'Pclass_2', 'Pclass_3',
                           'Sex_male', 'Sex_female',
                           'Age', 'Parch', 'SibSp', 'Fare',
                           'Title_Officer', 'Title_Royalty', 'Title_Master',
                           'Title_Mr', 'Title_Mrs', 'Title_Miss'])



classifiers = [
    ('Logistic regression', linear_model.LogisticRegression()),
    ('SVM classifier', svm.SVC(kernel='linear')),
    ('Decision tree classifier', tree.DecisionTreeClassifier()),
    ('Naive Bayes classifier', naive_bayes.GaussianNB()),
]

for name, candidate in classifiers:
    scores = cross_val_score(candidate, getX(data), data['Survived'], cv=5, scoring='accuracy')
    print(name, np.mean(scores))


"""
model = linear_model.LogisticRegression()
model.fit(getX(data), data['Survived'])

data = pandas.read_csv('test.csv')

data = preprocess(data)

data['Survived'] = model.predict(getX(data))

data.to_csv('submission.csv', index=False, columns=['PassengerId', 'Survived'])
"""