I have a dataframe that contains two columns one is called hits:

0      NaN
1      5.0
2      NaN
3      5.0
4      NaN
      ... 
295    NaN
296    NaN
297    NaN
298    NaN
299    NaN

and another column called converted:

0      0
1      0
2      0
3      0
4      0
      ..
345    0
346    1
347    0
348    1
349    0

what I want to do is to check if the row in the converted column has 0 or 1, if it has 0 the row in hits will have the value 6 ( it will replace the null value ) and if it has 1 it will replace the value with 9

I tried to use lambda but didnt give me a good result:

df['hits'] = df.apply(
    lambda row: 9 if df['converted'] == 1 & df['hits'] == 'nan' else 6
,axis=1) 

CodePudding user response：

You need change df to local variable row

df['hits'] = df.apply(
    lambda row: 9 if row['converted'] == 1 & pd.isnull(row['hits']) else 6
,axis=1) 

Or with boolean masking

df.loc[df['converted'].eq(0) & df['hits'].isna(), 'hits'] = 6
df.loc[df['converted'].eq(1), 'hits'] = 9