Pandas: Insert a zero in a DataFrame cell

June 26, 2023

I have a Dataframe with time series, whose values are presented below:

01/05/2023      25.1  25.9  25.1  
01/05/2023   1  25.1  25.2  25.0 
01/05/2023   2  24.7  25.1  24.7  
01/05/2023   3  24.7  24.8  NaN

In the cell above number 1, located at [0,1], should have a zero. How is the best way to do it?

>Solution :

That depends on what pandas considers you have there:

If you have a blank space, i. e. ' ', then you should probably try pd.replace()

import pandas as pd
import numpy as np
df = pd.DataFrame({'col1':['',1,2,3],
                   'col2':[25.9,25.2,25.1,24.8],
                   'col3':[25.1, 25.0, 24.7, np.nan]})
print(df)

returns:

  col1  col2  col3
0       25.9  25.1
1    1  25.2  25.0
2    2  25.1  24.7
3    3  24.8   NaN

If you then do

df.replace('', 0, inplace=True)
print(df)

you get

   col1  col2  col3
0     0  25.9  25.1
1     1  25.2  25.0
2     2  25.1  24.7
3     3  24.8   NaN

Another method of doing the same thing is using pd.replace without the inplace argument (be careful to assign the df again then df = df.replace('',0).

If it’s a NaN (which I doubt seeing the other NaN in your dataframe) you could try with pd.fillna()