使用前面的行创建一个新列，pandas

Question

I have a pandas dataframe as shown below

CID RefID   Date        Group   MID 
100     1   1/01/2021       A   101                     
100     2   3/01/2021       A   nan                   
100     3   4/01/2021       A   nan            
100     4   15/01/2021      A   nan 

I want to transform the 'MID' column in a way that : MID + RefID = MID and so on for all the other values. I have used .shift() method but didn't get the desired output

Expected output:

CID RefID   Date        Group   MID 
100     1   1/01/2021       A   101                     
100     2   3/01/2021       A   102                   
100     3   4/01/2021       A   104            
100     4   15/01/2021      A   107

Answer 1

Try cumsum and shift :

>>> df['MID'] = df['MID'].ffill() + df['RefID'].shift(fill_value=0).cumsum()
>>> df
   CID  RefID        Date Group    MID
0  100      1   1/01/2021     A  101.0
1  100      2   3/01/2021     A  102.0
2  100      3   4/01/2021     A  104.0
3  100      4  15/01/2021     A  107.0
>>>