Get Series with lowest value

Question

I have a dataframe df:

   a      b       c
   d  e   f   g   h   i
a  12 13  23  5   3   13
b  5  43  31  31  41  76
c  14 13  4   24  20  12

I want to get the series that contains the lowest value, so in this case df[('c', 'h')] . How can I code this?

Answer 1

You can try stack with idxmin :

df[df.stack([0, 1]).idxmin()[1:]]

Result:

a     3
b    41
c    20
Name: (c, h), dtype: int64