Deal with np.where and date condition

Question

I have a variable "date_0" which correspond to date of infection in a certain period, and I have to create another variable "date_1" from the previous variable (date_0) with a date condition (>= 33 days). "date_0" and "date_1" are in format %Y-%m-%d %H:%M:%S. So basically I should do a date condition like this :

if df['date_1] >= 33 days then df['date_1] else df['date_0']

And I do this :

df['date_1']= np.where((df['date_1']>= 33d), df['date_1'],df['date_O'])

And I try also with this :

df['date_1']= np.where((df['date_1'].days>= 33), df['date_1'],df['date_O'])

But I failed can you help me please ?

Answer 1

you can use the pandas equivalent to np.where. Ex:

import pandas as pd

df = pd.DataFrame({'date_0': [pd.Timestamp('2022-01-01'), pd.Timestamp('2022-02-01')],
                   'date_1': [pd.Timestamp('2022-01-29'), pd.Timestamp('2022-03-15')]})

df['days_diff'] = (df['date_1']-df['date_0']).dt.days

# date_select:
# use date 1 if the difference is greater equal 33 days else use date 0
df['date_select'] = df['date_1'].where(df['days_diff'] >= 33, df['date_0'])

df
      date_0     date_1  days_diff date_select
0 2022-01-01 2022-01-29         28  2022-01-01
1 2022-02-01 2022-03-15         42  2022-03-15