#!/usr/bin/env python
# coding: utf-8
#
#
# # Divorce Rates
# In[1]:
from lifelines.estimation import AalenAdditiveFitter
import pandas as pd
import numpy as np
import patsy
get_ipython().run_line_magic('pylab', 'inline')
# In[3]:
data = pd.read_csv('../lifelines/datasets/divorce.dat', sep="\s{2,10}")
# In[4]:
data.head(10)
# In[5]:
df = patsy.dmatrix('heduc + heblack + heblack:mixed + years + mixed + div -1 ', data, return_type='dataframe')
# In[6]:
df.head()
# In[130]:
aaf = AalenAdditiveFitter(fit_intercept=True, coef_penalizer=1.)
# In[131]:
df.columns
# In[132]:
timeline = np.linspace(0, 35, 1000)
# In[133]:
aaf.fit(df, 'years', event_col='div[T.Yes]', timeline=timeline)
# In[134]:
figsize(12.5, 8.5)
#aaf.cumulative_hazards_.ix[:][["baseline"]].plot()
#aaf.cumulative_hazards_.ix[:][["bb","bw","wb","ww"]].plot()
aaf.cumulative_hazards_.plot()
plt.legend(loc='upper left')
# In[ ]:
# In[ ]:
# In[ ]: