row_number_agg.py

In [2]: weekly_sales_df = sales_data[['Store', 
   ...:                               'Date', 
   ...:                               'Weekly_Sales']].groupby(['Store', 
   ...:                                                         'Date']).agg({'Weekly_Sales':'sum'})                                                                    

In [3]: weekly_sales_df.reset_index(inplace=True)                                                                                                                       

In [4]: weekly_sales_df['Date']= pd.to_datetime(weekly_sales_df['Date'])                                                                                                

In [5]: weekly_sales_df = weekly_sales_df.sort_values(by='Weekly_Sales',ascending=False)                                                                                

In [6]: weekly_sales_df['row num'] = weekly_sales_df.groupby(['Store']).cumcount()+1                                                                                    

In [7]: weekly_sales_df[weekly_sales_df['Store']==1].sort_values(by='Weekly_Sales',ascending=False)                                                                     
Out[7]: 
     Store       Date  Weekly_Sales  row num
112      1 2010-12-24    2387950.20        1
107      1 2011-12-23    2270188.99        2
117      1 2011-11-25    2033320.66        3
122      1 2010-11-26    1955624.11        4
25       1 2012-06-04    1899676.88        5
..     ...        ...           ...      ...
111      1 2010-09-24    1351791.03      139
102      1 2010-10-22    1345454.00      140
95       1 2011-01-21    1327405.42      141
123      1 2012-01-27    1319325.59      142
128      1 2011-01-28    1316899.31      143

[143 rows x 4 columns]