Query of the day: Sum results from multiples select statements and sort them by label with SQL

Hi all!
Today we have to do a clever sum using SQL statements.
We have to sum the number of elements in a table that match four different criteria and we want to aggregate them by year and month of reference. We also want to pick only the elements referring to the last 12 months.
Let us assume our table is in the form:

Table

ID	YEAR	MONTH	A	B
1234	2016	8	foo	bar
5678	2016	3	boo	bar
3412	2015	12	foo	boo
8888	2015	12	foo	bar
3412	2016	3	xxx	yyy

We want our result to count the elements with A=foo and b=boo, A=foo and B=bar, A=xxx and B=yyy and aggregate them by date. So our result will be:

2015, 12, 2
2016, 3,1
2016, 8, 1

The query is:

select yy, mm, SUM(qty) 
from (
  select year(DATA) as yy, month(DATA) as mm ,count(*) as qty 
  from table
  where (DATA)>= '2015-07-01 00:00:00.000' and 
  (DATA)<= '2016-08-01 00:00:00.000' and
  A='foo' and B = 'boo' 
  group by year(DATA),month(DATA)  
  union all 
  select year(DATA) as yy, month(DATA) as mm ,count(*) as qty 
  from table
  where (DATA)>= '2015-07-01 00:00:00.000' and 
  (DATA)<= '2016-08-01 00:00:00.000' and
  A='foo' and B = 'bar' 
  group by year(DATA),month(DATA) 

  union all
  select year(DATA) as yy, month(DATA) as mm ,count(*) as qty 
  from table
  where (DATA)>= '2015-07-01 00:00:00.000' and 
  (DATA)<= '2016-08-01 00:00:00.000' and
  A='xxx' and B = 'yyy' 
  group by year(DATA),month(DATA) 

)t
group by year, month

order by year, month  

The query actually is quite simple yet not so easy to combine. We do three different queries for the three views that we want to sum up, we use the union all statement to put them toghether in the same result set, then we use the outer query to perform the sum. This query has been tested over SQL Server, but it should work fine with any RDBMS. Notice the t is not a typo but an alias for the subquery!

If you have any question feel free to ask in the comments! :)