gistfile1.py

total = pd.concat([click[['TimeStamp','SesID','ItemID']],buy[['TimeStamp','SesID','ItemID']]],
                  ignore_index=True)

buy_sessions = buy['SesID'].drop_duplicates()
total['IsBuy'] = total['SesID'].isin(buy_sessions)
total_sesID_group = total.groupby('SesID')
click_sesID_group = total[total['IsBuy'] == False].groupby('SesID')
buy_sesID_group = total[total['IsBuy'] == True].groupby('SesID')

item2category = click[['ItemID','Category']].copy()
item2category.drop_duplicates(inplace=True)
print 'length of item2category :',len(item2category)
print 'total num of items: ', len(total['ItemID'].drop_duplicates())

# length of item2category : 100110
# total num of items:  52739

items_group = item2category.groupby('ItemID')
category_group = item2category.groupby('Category')

items_group_count = items_group.count()
category_group_count = category_group.count()