In [26]:

    

%matplotlib inline
from matplotlib import pylab as pl
import cPickle as pickle
import pandas as pd
import numpy as np
import os


    

In [27]:

    

df = pd.DataFrame()
#0.73188
df['gb'] = pd.read_csv('../submissions/140928-predict.1.csv', index_col='clip', squeeze=True) #64
#0.75037
df['rf'] = pd.read_csv('../submissions/140926-predict.2.csv', index_col='clip', squeeze=True)
#0.72018
df['rf1'] = pd.read_csv('../submissions/141101-predict.14.csv', index_col='clip', squeeze=True)
#0.71358
df['rfpca'] = pd.read_csv('../submissions/141001-predict.1.csv', index_col='clip', squeeze=True)
#0.71603
df['dbn'] = pd.read_csv('../submissions/140930-predict.5.csv', index_col='clip', squeeze=True)
#0.81632
df['best'] = pd.read_csv('../submissions/141101-predict.10.csv', index_col='clip', squeeze=True)


    

In [28]:

    

pd.scatter_matrix(df[['gb','rf','rf1','rfpca','dbn','best']],figsize=(6, 6), diagonal='kde');


    

In [29]:

    

w_gb  = 0.4
w_rf = 0.7
w_rf1 = 0.2
w_rfpca = 0.4
w_dbn = 0.6

s = w_gb + w_rf + w_rf1 + w_rfpca + w_dbn

w_gb /= s
w_rf /= s
w_rf1 /= s
w_rfpca /= s
w_dbn /= s


    

In [30]:

    

w_gb, w_rf, w_rf1, w_rfpca, w_dbn


    

    
Out[30]:





(0.17391304347826086,
 0.3043478260869565,
 0.08695652173913043,
 0.17391304347826086,
 0.26086956521739124)

In [31]:

    

df['preictal'] = w_gb * df['gb'] + w_rf * df['rf'] + w_rf1 * df['rf1'] + w_rfpca * df['rfpca'] + w_dbn * df['dbn']


    

In [32]:

    

df['preictal'].to_csv('../submissions/141101-predict.18.csv', header=True)


    

In [33]:

    

pd.scatter_matrix(df[['best','preictal']])


    

    
Out[33]:





array([[<matplotlib.axes._subplots.AxesSubplot object at 0x113d1c390>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x11afc0090>],
       [<matplotlib.axes._subplots.AxesSubplot object at 0x113693a50>,
        <matplotlib.axes._subplots.AxesSubplot object at 0x119e43290>]], dtype=object)






    

In [33]: