In [3]:

    

import pandas as pd
import numpy as np
import os
import math
import graphlab
import graphlab as gl
import graphlab.aggregate as agg


    

In [4]:

    

'''钢炮'''
path = '/home/zongyi/bimbo_data/'
sf = gl.SFrame.read_csv(path + 'train.csv', verbose=False)
town = gl.SFrame.read_csv(path + 'town_state.csv', verbose=False)


    

In [2]:

    

'''MAC'''
path = '/Users/zonemercy/jupyter_notebook/bimbo_data/'
sf = gl.SFrame.read_csv(path + 'train.csv', verbose=False)
town = gl.SFrame.read_csv(path + 'town_state.csv', verbose=False)


    

In [5]:

    

sf['return_rate'] = sf['Dev_uni_proxima'] / ( sf['Dev_uni_proxima'] + sf['Demanda_uni_equil'] )
re_lag = sf.groupby(key_columns=['Semana','Cliente_ID','Producto_ID'], operations={'re_lag':agg.MEAN('return_rate')})

# re_lag['Semana'] = re_lag['Semana'] + 1


    

In [6]:

    

lag = re_lag.copy()
re_lag.remove_column('re_lag')

lag['Semana'] = lag['Semana'].apply(lambda x: x+1)
re_lag = re_lag.join(lag,on=['Cliente_ID','Producto_ID','Semana'],how='outer')
lag['Semana'] = lag['Semana'].apply(lambda x: x+1)
re_lag = re_lag.join(lag,on=['Cliente_ID','Producto_ID','Semana'],how='outer')
lag['Semana'] = lag['Semana'].apply(lambda x: x+1)
re_lag = re_lag.join(lag,on=['Cliente_ID','Producto_ID','Semana'],how='outer')
lag['Semana'] = lag['Semana'].apply(lambda x: x+1)
re_lag = re_lag.join(lag,on=['Cliente_ID','Producto_ID','Semana'],how='outer')
lag['Semana'] = lag['Semana'].apply(lambda x: x+1)
re_lag = re_lag.join(lag,on=['Cliente_ID','Producto_ID','Semana'],how='outer')
# re_lag.rename({'re_lag':'re_lag1','re_lag.1':'re_lag2','re_lag.2':'re_lag3','re_lag.3':'re_lag4','re_lag.4':'re_lag5'})


    

In [13]:

    

re_lag.rename({'re_lag':'re_lag1','re_lag.1':'re_lag2','re_lag.2':'re_lag3','re_lag.3':'re_lag4','re_lag.4':'re_lag5'})


    

    
Out[13]:





    

        
Cliente_ID
        
Producto_ID
        
Semana
        
re_lag1
        
re_lag2
        
re_lag3
        
re_lag4
        
re_lag5
    
    

        
18648
        
2233
        
5
        
0.0
        
0.0
        
None
        
None
        
None
    
    

        
2323868
        
48417
        
4
        
0.0
        
None
        
None
        
None
        
None
    
    

        
538561
        
32819
        
4
        
0.0
        
None
        
None
        
None
        
None
    
    

        
2102008
        
34054
        
4
        
0.142857142857
        
None
        
None
        
None
        
None
    
    

        
503038
        
3270
        
5
        
0.0
        
0.0
        
None
        
None
        
None
    
    

        
1460981
        
2233
        
4
        
0.0
        
None
        
None
        
None
        
None
    
    

        
4669623
        
43207
        
4
        
0.0
        
None
        
None
        
None
        
None
    
    

        
181950
        
30575
        
4
        
0.0
        
None
        
None
        
None
        
None
    
    

        
2497667
        
42434
        
5
        
0.0
        
None
        
None
        
None
        
None
    
    

        
2209976
        
41938
        
4
        
0.0
        
None
        
None
        
None
        
None
    

[217171542 rows x 8 columns]
Note: Only the head of the SFrame is printed.
You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.

In [14]:

    

re_train=re_lag[(re_lag['Semana']>5)&(re_lag['Semana']<10)]


    

In [15]:

    

re_train.save(path+'re_lag_train.csv',format='csv')


    

In [16]:

    

re_test=re_lag[(re_lag['Semana']>9)&(re_lag['Semana']<12)]


    

In [17]:

    

re_test.save(path+'re_lag_test.csv',format='csv')