2

I have a dataframe which has horizontal identifiers (yes and no) and values, and I want to melt it into vertical values into each yes. Here is a snippet of my dataframe:

option        Region                       Store Name       option1 option2   option3 option4    profit
0            Region 1                           Store 1        Y       Y          N        N     48.1575
1            Region 1                           Store 2        N       Y          N        Y     74.7667
2            Region 1                           Store 3        N       Y          N        Y     102.35
3            Region 2                           Store 4        N       Y          N        Y     114.59
4            Region 2                           Store 5        N       Y          N        Y     99.705
5            Region 2                           Store 6        N       Y          N        Y     105.07

The answer is need to get is:

option        Region                       Store Name       options    profit
0            Region 1                           Store 1     option1     48.1575
1            Region 1                           Store 1     option2     48.1575
2            Region 1                           Store 2     option2     74.7667
3            Region 1                           Store 2     option4     74.7667

Essentially, I need to unstack the customer options tables, assign the same profit to everything with a yes, and drop everything with a no.

So far, the function I used is:

e1 = pd.melt(sales_dist_e, id_vars=['Area', 'Store Name'], var_name='option').set_index(['Area', 'Store Name', 'optionx']).squeeze().unstack().reset_index() which was mostly derived from this previous related question, but I can't seem to make it work with my current example.

Aeternis
  • 23
  • 3

1 Answers1

1

IIUC, does this work?

df.melt(['option', 'Region', 'Store Name', 'profit'], var_name='options')\
  .query("value == 'Y'")\
  .drop('value', axis=1)\
  .sort_values('profit')

Output:

    option   Region Store Name    profit  options
0        0  Region1    Store 1   48.1575  option1
6        0  Region1    Store 1   48.1575  option2
7        1  Region1    Store 2   74.7667  option2
19       1  Region1    Store 2   74.7667  option4
10       4  Region2    Store 5   99.7050  option2
22       4  Region2    Store 5   99.7050  option4
8        2  Region1    Store 3  102.3500  option2
20       2  Region1    Store 3  102.3500  option4
11       5  Region2    Store 6  105.0700  option2
23       5  Region2    Store 6  105.0700  option4
9        3  Region2    Store 4  114.5900  option2
21       3  Region2    Store 4  114.5900  option4
Scott Boston
  • 147,308
  • 15
  • 139
  • 187
  • Holy shit, thanks very much! This is very close to what I would need to do. Just some more minor help, the 'Region 1' and 'Store 1' in the table are place holders, since those are actual place names. How would I put them together in the same table? Also, is it possible to drop the 'yes' part and just group them by the options and the profits? Thanks again for a huge help! – Aeternis Nov 20 '22 at 05:13
  • @Aeternis Sure. I had to fix my input dataframe and run that above statement to produce these results. – Scott Boston Nov 20 '22 at 05:26