

In [ ]:

    
import pandas as pd
%matplotlib inline
import numpy as np



In [ ]:

    
titanic = pd.read_csv("data/titanic.csv")



In [ ]:

    
titanic.columns

Let's do a simple logistic regression to predict survival based on pclass and sex

First we need to prepare our features. Remember we drop one value in each dummy to avoid the dummy variable trap



In [ ]:

    
titanic['sex_female'] = titanic['sex'].apply(lambda x:1 if x=='female' else 0)



In [ ]:

    
dataset = titanic[['survived']].join([pd.get_dummies(titanic['pclass'],prefix="pclass"),titanic.sex_female])



In [ ]:

    
from sklearn.linear_model import LogisticRegression



In [ ]:

    
lm = LogisticRegression()



In [ ]:

    
#drop pclass_1st to avoid dummy variable trap
x = np.asarray(dataset[['pclass_2nd','pclass_3rd','sex_female']])
y = np.asarray(dataset['survived'])



In [ ]:

    
lm = lm.fit(x,y)



In [ ]:

    
lm.score(x,y)



In [ ]:

    
y.mean()



In [ ]:

    
lm.coef_



In [ ]:

    
lm.intercept_



In [ ]: