Shift months/years in Python by integer input-CodePudding

I have a dataset where I would like to shift the month and years in a given column based on an integer value. a2 column is effected when we input the integer: 9 We are incrementing the month by the value 9. We start with month 4 and add 9 more months = 1/1/2023

Data

start       a1          a2          stat        exit
1/1/2018    1/1/2022    4/1/2022    5/1/2022    6/1/2022
1/1/2018    1/1/2022    4/1/2022    5/1/2022    6/1/2022
1/1/2018    1/1/2022    4/1/2022    5/1/2022    6/1/2022

Desired output

start       a1          a2          stat        exit
1/1/2018    1/1/2022    1/1/2023    5/1/2022    6/1/2022
1/1/2018    1/1/2022    1/1/2023    5/1/2022    6/1/2022
1/1/2018    1/1/2022    1/1/2023    5/1/2022    6/1/2022

Doing

   d = {
             'm1': pd.DateOffset(months=3),
             'de': pd.DateOffset(months=5),
             're': pd.DateOffset(months=2),
             }
         s = pd.Series(d).rsub(datevalue)

monthvalue = pd.to_datetime(input("Enter month value: "))   #enter 9

new = pd.Series(d).rsub(datevalue) \
                     .append(pd.Series({'a2': monthvalue}))

I am still researching this. Any suggestions is helpful. I am not sure if we can mix datetime w./ integer input.

CodePudding user response：

Use pd.to_datetime with pd.DateOffset:

# Convert all columns to pandas datetime
In [558]: df = pd.DataFrame([pd.to_datetime(df[i]) for i in df.columns]).T

# Take user input for monthvalue
In [559]: monthvalue = int(input("Enter month value: "))
Enter month value: 9

# If monthvalue entered by user is 9, then add 9 months to column 'a2'
In [560]: if monthvalue == 9:
     ...:     df['a2'] = df['a2']    pd.DateOffset(months=monthvalue)
     ...: 

In [561]: df
Out[561]: 
       start         a1         a2       stat       exit
0 2018-01-01 2022-01-01 2023-01-01 2022-05-01 2022-06-01
1 2018-01-01 2022-01-01 2023-01-01 2022-05-01 2022-06-01
2 2018-01-01 2022-01-01 2023-01-01 2022-05-01 2022-06-01