SQL COUNT(*) 与 GROUP BY 和最后一行与总计数

Question

I want to use COUNT(*) on a group and then one more row at the end to give the full count. I have used UNION to run the COUNT(*) without GROUP BY in a separate sub-query to give the total count, but that approach seems inefficient. Any ways to combine both?

SELECT PCUST, COUNT(*) FROM PAYMENT 
WHERE PAYMENT_DATE = '2020-12-30'
GROUP BY PCUST
UNION
SELECT 'TOTAL', COUNT(*) FROM PAYMENT
WHERE PAYMENT_DATE = '2020-12-30'

As you can see from above query, I am repeating the same condition to get the partial and total counts. Plz suggest any way to optimize this.

Answer 1

DB2 supports GROUPING SETS . Assuming no PCUST values are NULL and it is a string, the simplest method is:

SELECT COALESCE(PCUST, 'TOTAL'), COUNT(*)
FROM PAYMENT 
WHERE PAYMENT_DATE = '2020-12-30'
GROUP BY GROUPING_SETS ( (PCUST), () )