Oracle SQL: Obtain count of distinct column values based on another column

Question

Here is a sample data to explain my case,

CompanyInfo:

Name,   Company, Location,   Completed,  Pending
Andy      AA        Home        4         2
Jim       AA        Office      3         3
Dahli     AA        Home        4         2
Monica    AA        Home        4         2
Chandler  AA    Home-Office     1         0
Ashley    AA    Home-Office     1         0

The last three columns have duplicated information and I am trying to obtain count of location, completed and pending which are bound to each other. So the output would look something like below,

   Company, Count(Locations), Count( Completed+Pending > 0),
    AA            3                   3

Why Count( Completed+Pending > 0) is 3? there are just three unique combinations of Home, Office and home-office columns where sum of completed+pending is > 0.

I did try below, but it gives me (AA, 3, 6) since it is processing all the 6 rows to obtain the count.

select Company, 
       count(distinct Location),
        SUM (
               CASE
                  WHEN (Completed + Pending) > 0 THEN 1
                  ELSE 0
               END)
               AS Total
       From CompanyInfo
       group by Company;

Any pointers?

Answer 1

I think you want a conditional count(distinct):

select Company, 
       count(distinct Location),
       count(distinct case when Completed + Pending > 0 then location end)
from CompanyInfo
group by Company;