Function that compares dates and asks for user input-CodePudding

I have calculated the difference between a date in my pd df and and the date now. I need to write a function that checks if these dates are more than three years apart, and if so, 'block' a contact in by asking confirmation to the user that they want this contact blocked.

#calculate time difference and determine today's date
df_transactions['month'] = df_transactions['date'].str.split('/').str[1]
month = df_transactions['month']
time_difference =  (pd.Timestamp.now().normalize() - pd.to_datetime(df_transactions['date'], errors='coerce'))
three_years = pd.to_datetime('2019-1-01')

#function: compare todays date with date from df and ask to block user
def date_block():
    if time_difference > three_years:
         if df_transactions["access"].any() == True:
             print(df_transactions["sender"]   " "   '')
             input('Would you like to restrict this contact? ')
            # block_contact(select_transaction['sender'])
    else:
        print('else statement triggered')

date_block()

The error I'm getting is the following:

pandas.core.arrays.datetimelike.InvalidComparison: 2019-01-01 00:00:00

How do I compare the two dates in a way that will not give me an error?
Thanks!

CodePudding user response：

The problem might be that you compare timedelta with datetime which does not work. Try the following:

date_range = pd.date_range(start='2018-01-01', end='2022-06-01')
df = pd.DataFrame(date_range, columns=['date'])
df['time_diff'] = (pd.Timestamp.now() - pd.to_datetime(date_range))
df['>3 years'] = df['time_diff']/pd.Timedelta(days=365) > 3

Output:

           date                 time_diff  >3 years
0    2018-01-01 1613 days 09:38:59.903248      True
1    2018-01-02 1612 days 09:38:59.903248      True
2    2018-01-03 1611 days 09:38:59.903248      True
3    2018-01-04 1610 days 09:38:59.903248      True
4    2018-01-05 1609 days 09:38:59.903248      True
...         ...                       ...       ...
1608 2022-05-28    5 days 09:38:59.903248     False
1609 2022-05-29    4 days 09:38:59.903248     False
1610 2022-05-30    3 days 09:38:59.903248     False
1611 2022-05-31    2 days 09:38:59.903248     False
1612 2022-06-01    1 days 09:38:59.903248     False

CodePudding user response：

I think firstly, you should use your date as int and substract them and convert the result in date then compare it