使用 case 语句时的 Group By 子句问题

Question

I am trying to sum up volumes across the year for every distinct sender and have the years as fields. I am using a case statement to do this.

select distinct sender,
  case
    when extract(month from (month)) = 1
      then sum(volume)
  end as January,
  case
    when  extract(month from (month)) = 2
      then sum(volume)
  end as February,
  case
    when extract(month from (month)) = 3
      then sum(volume)
  end as March
from [lifetime_data]
group by 1

When i run this i get an error that asks me to group by month, Month is not one of the columns i am selecting, i am only using it in my condition... how do i solve this?

Answer 1

I think you want conditional aggregation -- the case expression is an argument to the sum() :

select sender,
       sum(case when extract(month from (month)) = 1 then volume end) as January,
       sum(case when extract(month from (month)) = 2 then volume end) as February,
       sum(case when extract(month from (month)) = 3 then volume end) as March
from [lifetime_data]
group by 1

Answer 2

You got this error because your case statement is running row by row hence o/p for each row doesnt match the group of each row. Try the below

   Select sender, 
  extract(month from (month)), 
    SUM(Volume) from table group by
     sender, 
  extract(month from (month)) having
    extract(month from (month)) IN 
      (1,2,3)