
        
        
            
                
                
                    df = pd.DataFrame({"A":[3,pd.np.nan],"B":[pd.np.nan,6]})
df
                
            
               A    B
0  3.0  NaN
1  NaN  6.0

To fill all missing values with the value 8:


        
        
            
                
                
                    df.fillna(8)
                
            
               A    B
0  3.0  8.0
1  8.0  6.0

Replacing missing values in certain columns

Consider the same df as above:


        
        
            
                
                
                    df = pd.DataFrame({"A":[3,pd.np.nan],"B":[pd.np.nan,6]})
df
                
            
               A    B
0  3.0  NaN
1  NaN  6.0

To fill missing values in certain columns:


        
        
            
                
                
                    df.fillna({"A":8})
                
            
               A    B
0  3.0  NaN
1  8.0  6.0

Here, we are replacing all missing values in column A with 8.

Replacing missing values with value in preceding row

Consider the following DataFrame:


        
        
            
                
                
                    df = pd.DataFrame({"A":[3,pd.np.nan,5],"B":[pd.np.nan,6,7],"C":[8,pd.np.nan,pd.np.nan]})
df
                
            
               A    B
0  3.0  NaN
1  NaN  6.0
2  5.0  7.0

To fill missing values with value in the preceding row:


        
        
            
                
                
                    df.fillna(method="ffill")
                
            
               A    B    C
0  3.0  NaN  8.0
1  3.0  6.0  8.0
2  5.0  7.0  8.0

Here, note the following:

method="ffill" standard for forward-fill, that is, we replace a missing value with the previous non-NaN value in the same column.
we still have NaN in column B because there is no previous row.

Replacing missing values with value in next row

Consider the following DataFrame:


        
        
            
                
                
                    df = pd.DataFrame({"A":[3,pd.np.nan,5],"B":[6,7,pd.np.nan],"C":[pd.np.nan,pd.np.nan,8]})
df
                
            
               A    B    C
0  3.0  6.0  NaN
1  NaN  7.0  NaN
2  5.0  NaN  8.0

To fill missing values with value in the next row:


        
        
            
                
                
                    df.fillna(method="bfill")
                
            
               A    B    C
0  3.0  6.0  8.0
1  5.0  7.0  8.0
2  5.0  NaN  8.0

Here, note the following:

method="bfill" standard for backward-fill, that is, we replace a missing value with the next non-NaN value in the same column.
we still have NaN in column B because there is no next row.

Pandas DataFrame | fillna method

Fills missing values (NaN) with a specified value or with a filling rule.

chevron_right