Calculate week/month-to-date change in Pandas (with grouped data)

Question

I have panel data of stock tickers over time and the price of the stock each day.

I have seen solutions using pandas shift (code attempts to implement below) to calculate the change over a period (assuming a fixed number of days) but I need to calculate the change month-to-date, year-to-date, versus the last Friday (i.e week-to-date) etc. (I.e. the calculation is not over a fixed number of periods).

Code attempts:

df['Price_chg'] = df.groupby('Ticker')['Price'].apply(pd.Series.diff)

df['Price_chg'] = df.groupby('Ticker')['Price'].apply(lambda x: x - x.shift())

Example data:

Ticker	Date	Price
AAUK	2022-02-28	100
AAUK	2022-03-01	105
AAUK	2022-03-02	107
AAUK	2022-03-03	110
AAUK	2022-03-04	105
AAUK	2022-03-07	112
AAUK	2022-03-08	120
AAUK	2022-03-09	117
AAUK	2022-03-10	112
AAUK	2022-03-11	125
ABX	2022-02-28	100
ABX	2022-03-01	100
ABX	2022-03-02	110
ABX	2022-03-03	115
ABX	2022-03-04	117
ABX	2022-03-07	120
ABX	2022-03-08	122
ABX	2022-03-09	125
ABX	2022-03-10	127
ABX	2022-03-11	130

Desired output:

Ticker	Date	Price	WTD	MTD
AAUK	2022-02-28	100	N/A	N/A
AAUK	2022-03-01	105	5	5
AAUK	2022-03-02	107	7	7
AAUK	2022-03-03	110	10	10
AAUK	2022-03-04	105	5	5
AAUK	2022-03-07	112	7	12
AAUK	2022-03-08	120	15	20
AAUK	2022-03-09	117	12	17
AAUK	2022-03-10	112	7	12
AAUK	2022-03-11	125	20	25
ABX	2022-02-28	100	N/A	N/A
ABX	2022-03-01	100	0	0
ABX	2022-03-02	110	10	10
ABX	2022-03-03	115	15	15
ABX	2022-03-04	117	17	17
ABX	2022-03-07	120	3	20
ABX	2022-03-08	122	5	22
ABX	2022-03-09	125	8	25
ABX	2022-03-10	127	10	27
ABX	2022-03-11	130	13	30

Thanks

Answer 1

Have figured a solution by first extracting the week/month/quarter etc.

df['WEEK'] = pd.to_datetime(df['Date']).dt.to_period('W')

Then apply a broader grouping when applying the lambda function:

df['1WEEK_Price_chg'] = df.groupby(['Ticker', 'WEEK'])['Price'].apply(lambda x: x - x.shift())