Utilize Python Input() to perform DateShift on a dataset-CodePudding

I have a dataset where, whenever a date value in the Update column is entered into the input() prompt, Date1, Date2 and Date3 values will be updated according to these rules:

   Date1 is 5 months from the date entered in the input() prompt
   Date2 is 2 months from the date entered in the input() prompt
   Date3 is 1 month from the date entered in the input() prompt

The only data that is changing are the dates, which are essentially getting shifted based upon the user input.

Data

ID   Stat   Date1     Date2     Date3       Update
aa   500    1/1/2021  4/1/2021  5/1/2021    6/1/2021
bb   800    1/1/2021  4/1/2021  5/1/2021    6/1/2021

Desired

Input prompt will ask user which date value they wish to input. User inputs the date '8/1/2021', which updates the remaining column date values. based on rules above.

ID   Stat   Date1     Date2     Date3       Update
aa   500    3/1/2021   6/1/2021 7/1/2021    8/1/2021
bb   800    3/1/2021   6/1/2021 7/1/2021    8/1/2021

Doing

I believe I can use a combination of a function as well as user prompt to approach this problem.

Type input

datevalue = pd.to_datetime(input("Enter date value: "))
print(datevalue)

Use input variable in function or script to create date update for Date1, Date2 and Date3

df[0] = df[1].apply(lambda x: datevalue - pd.DateOffset(months=x))

s = df['Update'].str.replace(r'(\S ) (\S )', r'\2\1')
df['Update'] = (pd.PeriodIndex(s, freq='D')   3).strftime('D%q %Y')

I am looking for some starting point suggestion or a good foundation/documentation on how to best approach this problem. I am still researching. Any suggestion is appreciated.

CodePudding user response：

We can define a mapping dictionary which maps the column names to corresponding dateoffset, then create a series from this dictionary and subtract the datevalue from offsets, finally assign the updated date values back to dataframe

d = {
    'Date1': pd.DateOffset(months=5),
    'Date2': pd.DateOffset(months=2),
    'Date3': pd.DateOffset(months=1),
}

s = pd.Series(d).rsub(datevalue)
df.assign(**{**s, 'Update': datevalue})

   ID  Stat      Date1      Date2      Date3     Update
0  aa   500 2021-03-01 2021-06-01 2021-07-01 2021-08-01
1  bb   800 2021-03-01 2021-06-01 2021-07-01 2021-08-01

CodePudding user response：

I think you're looking for something like this:

cols = df.filter(like='Date').columns.tolist()   ['Update']
df[cols] = df[cols].apply(lambda col: pd.to_datetime(col)   (datevalue - df['Update']))

Output:

>>> df
   ID  Stat      Date1      Date2      Date3     Update
0  aa   500 2021-03-03 2021-06-01 2021-07-01 2021-08-01
1  bb   800 2021-03-03 2021-06-01 2021-07-01 2021-08-01