use fillna(0) with condition

Question

I need to replace NaN values with 0, but only when Type = u. How can I do it?

index Landsize    Type    
1       NaN         u      
2       NaN         m

Joe · Accepted Answer · 2019-09-02T14:19:39.180

1

You can use:

df['Landsize'] = np.where(df['Type'] =='u', 0, np.nan)

or if not all values in df['Landsize'] are nan:

df['Landsize'] = np.where(df['Type'] =='u', 0, df['Landsize'])

edited Sep 02 '19 at 14:19

answered Aug 30 '19 at 11:05

Joe

1

I've tried this and now almost all the values are 'NaN'. – Iga Golinska Sep 02 '19 at 11:01
Try this: `df['Landsize'] = np.where(df['Type'] =='u', 0, df['Landsize'])` – Joe Sep 02 '19 at 12:32

score -1 · Answer 2 · answered Aug 30 '19 at 11:04

-1

try this:-

df[df['type']=='u']['Landsize'] = 0

answered Aug 30 '19 at 11:04

Aniket Dixit

https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#evaluation-order-matters – jezrael Aug 30 '19 at 11:09
Solution is correct but not recommended that's different thing – Aniket Dixit Aug 30 '19 at 11:13
ok, agree. Not recommended. – jezrael Aug 30 '19 at 11:14
btw, why not change it? – jezrael Aug 30 '19 at 11:15
can do it from next problems – Aniket Dixit Aug 30 '19 at 11:17
@AniketDixit: why would you share a non-recommended solution? This can only generate trouble. – Willem Van Onsem Aug 30 '19 at 12:49

2 Answers2