我有一个以下格式的日期框架：
CUST_ID PRODUCT MONTH DAY WEEK_DAY HOUR
1 A JUNE 17 0 23
1 B JULY 20 2 15
2 B JUNE 5 4 9
2 A JULY 9 1 2
3 C JUNE 22 2 10
3 C JULY 30 3 18
3 A JUNE 17 0 23
8 C JULY 30 3 18
15 C JULY 30 3 18
我想对其进行汇总，以便对每个产品和每小时的Cust_id进行计数：
PRODUCT MONTH DAY HOUR WEEK_DAY COUNT_CUST_ID
A JUNE 17 23 0 2
A JULY 9 2 1 1
B JUNE 5 9 4 1
B JULY 20 15 2 1
C JUNE 22 10 2 1
C JULY 30 18 3 3
我尝试了这个，但是它没有给我一个“经典的”数据框：
pd.DataFrame(df.groupby(['PRODUCT','MONTH','DAY','HOUR','WEEK_DAY']).count()).groupby('CUST_ID')