Create boolean columns in DataFrame pandas for date and year-CodePudding

Let it be the following Python Panda DataFrame:

code	date	color	counter
TTT	2021-02-01 00:30:11	blue	2
AVC	2022-02-01 10:00:01	red	12
BCH	2022-02-02 22:00:01	red	12
POD	2022-03-01 01:00:03	blue	1
UDY	2022-05-01 09:02:01	red	3
YUD	2022-12-01 09:02:01	green	0

I need to create a column by year and month that includes whether or not that row belongs (bool) to the specific month indicated in the column. An example of the resulting DataFrame is as follows:

code	date	color	counter	2021-02	2022-02	2022-03	2022-05	2022-12
TTT	2021-02-01 00:30:11	blue	2	True	False	False	False	False
AVC	2022-02-01 10:00:01	red	12	False	True	False	False	False
BCH	2022-02-02 22:00:01	red	12	False	True	False	False	False
POD	2022-03-01 01:00:03	blue	1	False	False	True	False	False
UDY	2022-05-01 09:02:01	red	3	False	False	False	True	False
YUD	2022-12-01 09:02:01	green	0	False	False	False	False	True

To solve the problem only the information in the date column is used, but I include the complete DataFrame to avoid incompatibilities.

CodePudding user response：

Use get_dummies with Series.dt.strftime for YYYY-MM format and parameter dtype for booleans:

df = df.join(pd.get_dummies(df['date'].dt.strftime('%Y-%m'), dtype=bool))
print (df)
  code                date  color  counter  2021-02  2022-02  2022-03  \
0  TTT 2021-02-01 00:30:11   blue        2     True    False    False   
1  AVC 2022-02-01 10:00:01    red       12    False     True    False   
2  BCH 2022-02-02 22:00:01    red       12    False     True    False   
3  POD 2022-03-01 01:00:03   blue        1    False    False     True   
4  UDY 2022-05-01 09:02:01    red        3    False    False    False   
5  YUD 2022-12-01 09:02:01  green        0    False    False    False   

   2022-05  2022-12  
0    False    False  
1    False    False  
2    False    False  
3    False    False  
4     True    False  
5    False     True

CodePudding user response：

You can try df.to_period with str.get_dummies

out = df.join(pd.to_datetime(df['date']).dt.to_period("M")
              .astype(str).str.get_dummies()
              .astype(bool))

print(out)

  code                 date  color  counter  2021-02  2022-02  2022-03  2022-05  2022-12
0  TTT  2021-02-01 00:30:11   blue        2     True    False    False    False    False
1  AVC  2022-02-01 10:00:01    red       12    False     True    False    False    False
2  BCH  2022-02-02 22:00:01    red       12    False     True    False    False    False
3  POD  2022-03-01 01:00:03   blue        1    False    False     True    False    False
4  UDY  2022-05-01 09:02:01    red        3    False    False    False     True    False
5  YUD  2022-12-01 09:02:01  green        0    False    False    False    False     True