add column dynamically and assign calculated value from another column-CodePudding

I have some prepaid sales orders and need to assign the sales value to the subsequent months according to the number of months the sales order cover (paid/month). the dataframe looks like this:

order  number of months    year    start month start year  paid       
1        2                2021         10         2021      300
2        3                2021         10         2021      300
3        1                2021         11         2021      50

.... and it should look like

order months year start_month start_year paid  2021_10  2021_11  2021_12 
  1      2    2021     10         2021    300      150     150             
  2      3    2021     10         2021    300      100     100     100 
  3      1    2021     11         2021     50               50

....

Thank you!

CodePudding user response：

I am assuming you are using pandas.

First, you will need to get all the combinations of months and years in which there were sales. This depends on your edge cases and data. but for the sake of the example:

date_combinations = [(10,2021), (11,2021), (12,2021)]

Now, the way to dynamically add columns and fill them based on multiple other rows is to use the 'apply' function on the entire data frame and get the rows like this:

def fill_date_col(row):
    if date[0] < row["start_month"]   row["months"] and date[0] >= row["start_month"] and row["start_year"] == date[1] :
        return row["paid"]/row["months"]
    else:
        return None
for date in date_combinations:
    df[f"{date[0]}_{date[1]}"] = df.apply(lambda row: fill_date_col(row), axis=1)

and this is the result

order   months  start_month start_year  paid    10_2021 11_2021 12_2021
0   1   2   10  2021    300 150.0   150.0   NaN
1   2   3   10  2021    300 100.0   100.0   100.0
2   3   1   11  2021    50  NaN 50.0    NaN

Of course, you need the adjust this with regard to your span of dates and edge cases, especially if some sales span over a year.