Replace the value in a column with the value of another column Pandas

Question

I have a data frame, with columns col1, col2, and col3. I would like to replace col2 with the value of col3 if col3 is not null; otherwise, skip the row.

input

col1  col2   col3 
   1     2    nan
   4     5    nan
   6     7      8
   1     0      9

output

col1  col2   col3 
   1     2    nan
   4     5    nan
   6     8      8
   1     9      9

This is my code:

for i, row in df.iterrows():
   if row['col3'] != np.nan:
       row['col2'] = row['col3']
   else:
       row['col2'] = row['col2']

But it doesn't work.

From what it sounds like, you should be able to first filter your dataframe by whether col3 is NaN: `df[~df['col3'].isnan()]`, then apply your logic. — ddejohn, Jun 07 '22 at 17:18
@ddejohn intead of `~df['col3'].isnan()` you can use `df['col3'].notna()` — , Jun 07 '22 at 17:19
Note that `!= np.nan` will [never work](https://stackoverflow.com/questions/19322506/pandas-dataframes-with-nans-equality-comparison). — ddejohn, Jun 07 '22 at 17:23

score 2 · Answer 1 · 2022-06-07T17:30:03.147

2

Use .loc and .notna():

df.loc[df['col3'].notna(), 'col2'] = df['col3']

Output:

>>> df
   col1  col2  col3
0     1   2.0   NaN
1     4   5.0   NaN
2     6   8.0   8.0
3     1   9.0   9.0

edited Jun 07 '22 at 17:30

answered Jun 07 '22 at 17:18

Replace the value in a column with the value of another column Pandas

1 Answers1