Question

Is it possible to count two columns?

I have a table that stores application names and which users have launched over time.

When querying a date range you end up with a result like the below example.

| App   | User   |
| App A | User A |
| App A | User A |
| App A | User B |
| App A | User C |
| App A | User D |
| App A | User D |
| App A | User D |
| App B | User A |
| App B | User E |
| App B | User F |
| App C | User A |
| App C | User C |
| App C | User C |
| App C | User C |
| App C | User F |

What I’m trying to get is one column that shows the number of app launches per app and another showing the number of users launching that app.

I can’t figure out how to do it.

Any one any ideas?

Answer 1

select  App
,       count(*) as Launches
,       count(distinct [User]) as Users
from    Table1
group by
        App

Example at SQL Fiddle.

The square brackets around [User] are for SQL Server. Without them, you’d get your own username instead of the User column. In MySQL, you’d use backticks.