Suggestion

[rparente16]

LOVE this function. Would love even more if there was an option to do nway=True so that it would add all combinations. For example:

CURRENT:
df.groupby(['sex', 'class','city']).agg({'fare': ['sum']}).stb.subtotal(nway=True)

sex	class	city	Count
Female	First	Melbourne	5
Female	First	Burbank	10
Female	First	Female/First Subtotal	15
Female	Second	Melbourne	3
Female	Second	Burbank	6
Female	Second	Female/Second Subtotal	9
Female	Female Subtotal		24
Male	First	Melbourne	10
Male	First	Burbank	15
Male	First	Male/First Subtotal	25
Male	Second	Melbourne	9
Male	Second	Burbank	3
Male	Second	Male/Second Subtotal	12
Male	Male Subtotal		37

NEW ROWS:

sex	class	city	Count
Female	Female/City Subtotal	Melbourne	8
Female	Female/City Subtotal	Burbank	16
Male	Male/City Subtotal	Melbourne	19
Male	Male/City Subtotal	Burbank	18
First/Burbank Subtotal	First	Burbank	25
First/Burbank Subtotal	Second	Burbank	9

and so on ...

Proc Means in SAS does this and it's such a headache to code in Python. WOULD love if you added it!!!