Count the number of rows produced by a query

You're going to smile when you see this, because it's so simple:

select count(*) as rows
  from ( 
       select ID, type
         from personal_time_event
        where logdate = '2005-04-15'
       group by ID, type
       ) as dt 

Notice how the subquery no longer includes either the count for each individual ID and type combination, or the ORDER BY clause. They aren't necessary in the subquery, because all you really want is a count of the rows produced by the subquery.

There's also another way to get the same answer:

select count(distinct ID, type) as rows
  from personal_time_event
 where logdate = '2005-04-15'

You don't often see two columns inside a COUNT(DISTINCT...) expression, but it gives exactly what you expect -- at least, it does in those databases that support it. In those that don't, you'll have to concatenate the values, and it's perfectly okay to do so, because all you want is to count the number of unique combinations:

select count(distinct ID || type) as rows
  from personal_time_event
 where logdate = '2005-04-15'

Depending on your database, you might need a different concatenation operator, and you may or may not need to worry about implicit data conversion if ID is numeric and type isn't. Test it and see.