Question

First post here! Really need some help.

I have a table with buildings. In this table, there is a number of attributes, of which I am interested in 2:

- roof_material
- additional_roof_material

In the column roof material, there is 10 different roof materials, with the numbers 1-10.

The same type of additional roof materials are listed in additional_roof_material.

How do I count how many times buildings with roof_material 1, has an additional_roof_material 1, for an instance?

This is what i want to do, just for every material:

Select count(*)
From dbo.40000t where roof_material = 1 and additional_roof_material = 1

I want to do it for each of the types, so i get a complete comparision of how many times roof_material 1,2,3,4,5,6,7,8,9,10 has additional_roof_material 1,2,3,4,5,6,7,8,9,10.

Answer 1

You can get permutation counts from a table using sum() with a case statement

select
    sum((case when roof_material = 1 and additional_roof_material = 1 then 1 else 0 end)) as material_1_1,
    sum((case when roof_material = 2 and additional_roof_material = 2 then 1 else 0 end)) as material_2_2
    -- etc...
from
    dbo.40000t

This query has an implicit group by clause - because I am only selecting aggregates I do not need to explicitly specify a group by.

Update

As requested in a comment, if you want each permutation as a distinct record

select
    roof_material,
    additional_roof_material,
    count(1) as num_instances
from
    dbo.40000t
-- Not sure if you want permutations where roof_material != additional_roof_material or not.
-- Uncomment these next 2 lines if you do not want those permutations
--where
--  roof_material = additional_roof_material
group by
    roof_material,
    additional_roof_material

Count how many times a value occurs across tables

1 个答案: