Sort dataframe include empty field

I have a dataframe as below:

data = {'Test': [' ', ' ', 'K', ],
        'Name': ['A', 'B', 'B', 'B'],
        'value': ['D1', 'A1', ' ', 'C1'],
        'time': [227, 227, 227, 230]}
df = pd.DataFrame(data)

  Test  Name value  time
0        A    D1    227
1        B    A1    227
2   K    B          227
3        B    C1    230

And I want to make the df to sort as:

  Test  Name value  time
0        A    D1    227
1   K    B          227
2        B    A1    227
3        B    C1    230

I’ve tried using sort_values, but still can’t figure it out. Or should I add more condition for the empty field ‘ ‘(or NA) in sort?

>Solution :

Better not use empty strings/spaces to denote empty cells. Use NAs/NaNs that are directly handled by sort_values:

out = (df.replace(' ', pd.NA)
         .sort_values(by=['Name', 'time', 'Test'])
       )

Which is equivalents to:

out = (df.replace(' ', pd.NA)
         .sort_values(by=['Name', 'time', 'Test'], na_position='last')
       )

Output:

   Test Name value  time
0  <NA>    A    D1   227
2     K    B  <NA>   227
1  <NA>    B    A1   227
3  <NA>    B    C1   230