CODEWITHSUNDEEP
t-sql
Cute Bear
Cute Bear

Reputation: 3293

Joining query results into one

This is my table to track down the employees off days. This example is only for one person.

ID   PID     Year       OffDays       DayTypeNumber
------------------------------------------
1     1      2011       10            1
2     1      2011       5             2
3     1      2012       20            1
4     1      2012       3             2

I would like to write such a query that should only show one result for each year with additional column

Year       OffDays(1)    OffDays(2)
------------------------------------------
2011       10            5
2012       20            3

Upvotes: 1

Views: 53

Answers (1)

Taryn
Taryn

Reputation: 247620

You can use the PIVOT function for this:

select year,
  [1] [OffDays(1)],
  [2] [OffDays(2)]
from
(
  select year, offdays, daytypenumber
  from yourtable
) src
pivot
(
  sum(offdays)
  for daytypenumber in([1], [2])
) piv

See SQL Fiddle with Demo

Result:

| YEAR | OFFDAYS(1) | OFFDAYS(2) |
----------------------------------
| 2011 |         10 |          5 |
| 2012 |         20 |          3 |

Or you can use an aggregate function with a CASE statement:

select year,
  sum(case when daytypenumber = 1 then offdays end) [OffDays(1)],
  sum(case when daytypenumber = 2 then offdays end) [OffDays(2)]
from yourtable
group by year

See SQL Fiddle with Demo

If you only have two types that you are comparing, then you can use subqueries:

select t1.year, 
  [OffDays(1)],
  [OffDays(2)]
from
(
  select sum(offdays) [OffDays(1)], year
  from yourtable
  where daytypenumber = 1
  group by year
) t1
left join
(
  select sum(offdays) [OffDays(2)], year
  from yourtable
  where daytypenumber = 2
  group by year
) t2
  on t1.year = t2.year

See SQL Fiddle with Demo

The above answers will work great, if you have a known number of values for DayTypeNumber, but if those are unknown then you can use dynamic SQL to generate the PIVOT:

DECLARE @cols AS NVARCHAR(MAX),
    @colNames AS NVARCHAR(MAX),
    @query  AS NVARCHAR(MAX)

select @cols = STUFF((SELECT distinct ',' + QUOTENAME(DayTypeNumber) 
                    from yourtable
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)') 
        ,1,1,'')

select @colNames = STUFF((SELECT distinct ', ' + QUOTENAME(DayTypeNumber) 
                      +' as [OffDays('+cast(DayTypeNumber as varchar(10))+')]'
                    from yourtable
            FOR XML PATH(''), TYPE
            ).value('.', 'NVARCHAR(MAX)') 
        ,1,1,'')


set @query = 'SELECT year,' + @colNames + ' from 
             (
                select year, offdays, daytypenumber
                from yourtable
            ) x
            pivot 
            (
                sum(offdays)
                for daytypenumber in (' + @cols + ')
            ) p '

execute(@query)

See SQL Fiddle with Demo

All of these will produce the same results:

| YEAR | OFFDAYS(1) | OFFDAYS(2) |
----------------------------------
| 2011 |         10 |          5 |
| 2012 |         20 |          3 |

Upvotes: 3

Related Questions