Question

I have two tables, which share a column, that is not unique. I want all records where table A has more values of the shared column than TABLE B.

TABLE A:
Shared_Column|User_ID|Department
123          |    joe|     sales
123          |    joe|     sales
123          |    joe|     sales
124          |    sam|       ops
124          |    sam|       ops

TABLE B
Shared_Column|Other_Column
123          |           1
123          |           1
124          |           4
124          |           4

From this data, I want joe|sales but not sam|ops. I could also work with this as output:

USER|TABLE_A_COUNT|TABLE_B_COUNT
 joe|            3|            2
 sam|            2|            2

edit: I’ve tried to do a join like this:

select a.user_ID, count(a.shared_column) as 'TABLE_A_COUNT', count(b.shared_column) as 'TABLE_B_COUNT'
from a inner join b on a.shared_column = b.shared_column
group by a.user_ID

but that seems to produce a cross join and I get joe|6|6 instead of 3 and 2

Thanks!

Answer 1

It seems like you want something like this:

select a.user_id,
  count(a.shared_column) TableA,
  TableB
from tablea a
inner join
(
  select count(*) TableB, Shared_column
  from tableb
  group by shared_column
) b
  on a.Shared_Column = b.Shared_Column
group by a.user_id, TableB

See Sql Fiddle with Demo

Result:

| USER_ID | TABLEA | TABLEB |
-----------------------------
|     joe |      3 |      2 |
|     sam |      2 |      2 |