
        
        
            
                
                
                    df = pd.DataFrame({"A":[np.nan,3,np.nan],"B":[4,np.nan,5],"C":[np.nan,7,8]}, index=["a","b","c"])
df
                
            
               A    B    C
a  NaN  4.0  NaN
b  3.0  NaN  7
c  NaN  5.0  8

Of each row

To count the number of NaN values in each row of df:


        
        
            
                
                
                    df.isna().sum(axis=1)
                
            
            a    2
b    1
c    1
dtype: int64

Explanation

Here, the df.isna() returns a DataFrame of booleans where True indicates entries that are NaN:


        
        
            
                
                
                    df.isna()
                
            
               A      B      C
a  True   False  True
b  False  True   False
c  True   False  False

Internally, True is represented by 1 while a False is represented by 0. Therefore, summing up the booleans for each row is equivalent to counting the number of True (NaN values) per row:


        
        
            
                
                
                    df.isna().sum(axis=1)
                
            
            a    2
b    1
c    1
dtype: int64

Here, we must specify axis=1 so that we are summing each row, and not each column.

Of a particular row

Consider the same df as before:


        
        
            
                
                
                    df = pd.DataFrame({"A":[np.nan,3,np.nan],"B":[4,np.nan,5],"C":[np.nan,7,8]}, index=["a","b","c"])
df
                
            
               A    B    C
a  NaN  4.0  NaN
b  3.0  NaN  7
c  NaN  5.0  8

To count the number of NaN value of just row a:


        
        
            
                
                
                    df.loc["a"].isna().sum()
                
            
            2

Explanation

Here, the DataFrame's loc property is first used to extract the row a:


        
        
            
                
                
                    df.loc["a"]
                
            
            A    NaN
B    4.0
C    NaN
Name: a, dtype: float64

We then use the same tactic as described above to count the number of NaNs in this row.

Published by Isshin Inada

Edited by 0 others

Did you find this page useful?

thumb_up

thumb_down

Comment

Citation

Ask a question or leave a feedback...

thumb_up

thumb_down

chat_bubble_outline

settings

Enjoy our search

Hit / to insta-search docs and recipes!