Question

I’m trying to find the best way to summarize the data in a table

I have a table Info with fields

id
region_number integer (NOT associated with another table)
member_name string
member_active T/F

Members belong to a region, have a name, and are either active or not.

I’m wondering if there is a single query that will create a table with 3 columns, and as many rows as there are unique region_numbers:

For each unique region_number:
  region_number
  COUNT of members in that region
  COUNT of members in that region with active=TRUE

Suppose I have 50 regions, I see how to do it with 2×50 queries but that surely is not the right approach!

Answer 1

You can always group on several things if you’re prepared to do a tiny bit of post-processing:

SELECT region_number, COUNT(*) AS instances, member_active
  GROUP BY region_number, member_active
  WHERE region_number IN ?

This allows you do to one query for all region numbers at the same time. There will be one row for the T values, one for the F, but only if those are present.

If you see a case where you’re doing a lot of queries that differ only in identifiers, that’s something you can usually execute in one shot like this.