Python pandas: selecting rows whose column value is null / None / nan

Question

How do I select those rows of a DataFrame whose value in a column is none?

I've coded these to np.nan and can't match against this type.

In [1]: import numpy as np

In [2]: import pandas as pd

In [3]: df = pd.DataFrame([[1, 2, 3], [3, 4, None]])

In [4]: df
Out[4]: 
   0  1    2
0  1  2  3.0
1  3  4  NaN

In [5]: df = df.fillna(np.nan)

In [6]: df
Out[6]: 
   0  1    2
0  1  2  3.0
1  3  4  NaN

In [7]: df.iloc[1][2]
Out[7]: nan

In [8]: df.iloc[1][2] == np.nan
Out[8]: False

In [9]: df[df[2] == None]
Out[9]: 
Empty DataFrame
Columns: [0, 1, 2]
Index: []

Thanks. Sorry for the dupe. Wasn't finding what I needed in searches. Why can't I use `df[df[2] == None]` ? Why does np.nan not match against itself after I code NaNs using fillna(np.nan)? — zadrozny, Oct 25 '16 at 17:00
you can find an answer [here](http://stackoverflow.com/questions/10034149/why-is-nan-not-equal-to-nan). And `fillna(np.nan)` will replace NaN's with NaN's ;) — MaxU - stop genocide of UA, Oct 25 '16 at 17:04

MaxU - stop genocide of UA · Answer 1 · 2019-07-15T19:51:00.920

60

you can use .isna() method:

In [48]: df[df[2].isna()]
Out[48]:
   0  1   2
1  3  4 NaN

edited Jul 15 '19 at 19:51

answered Oct 25 '16 at 16:54

MaxU - stop genocide of UA

191,778
30
340
375

Do you know why `df.loc[(df['2'] == nan) & (df['2'] == None)]` not working? – ah bon Oct 30 '19 at 01:43
3

@ahbon, try this: np.nan == np.nan – MaxU - stop genocide of UA Nov 01 '19 at 04:46

Python pandas: selecting rows whose column value is null / None / nan

1 Answers1