DataFrame: Setting all the values to a particular columns as NaN-CodePudding

I have a DF like

A   B   C   D   E   F   G  H   I
=====================================
1   2   3       4   5   6   7   8
1   2   3       4   5   6   7   8
1   2   3       4   5   6   7   8

I need to set all the values for columns except for A and C to be Null

A   B   C   D   E   F   G  H   I
=====================================
1       3
1       3
1       3

If there a way to do this instead of using the below code like using a not condition to check for only columns A & C and set null to every other columns

df[['B','D','E','F','G','H','I']] = ''

CodePudding user response：

An easy way could be to slice and reindex:

target = ['A', 'C']
out = df[target].reindex(df.columns, axis=1)

For in place modification you can take advantage of index difference:

df[df.columns.difference(target)] = float('nan')

output:

   A   B  C   D   E   F   G   H   I
0  1 NaN  3 NaN NaN NaN NaN NaN NaN
1  1 NaN  3 NaN NaN NaN NaN NaN NaN
2  1 NaN  3 NaN NaN NaN NaN NaN NaN

*NB. if you really want spaces as filling values:

# option 1:
out = df[target].reindex(df.columns, axis=1, fill_value='')

# option 2:
df[df.columns.difference(target)] = ''

# output:
   A B  C D E F G H I
0  1    3            
1  1    3            
2  1    3

CodePudding user response：

target = ['A', 'C']
columns = [column for column in df.columns if column not in target]
df[columns] = float('nan')