Any good idea to fast filter the following dataframe (the original is big): for rows with same value of 'member', only keep the one with smallest corresponding 'height':
group member height
A m 1
B m 2
C g 3
The result will be:
group member height
A m 1
C g 3