Separate values of a column as diferrent columns as same query - sql

Question

I have a table UTENSILS with 3 columns like this:
CLASS_NAME RANGE COUNT
---------------------------
pens 0-0.5 200
pencil 0-0.5 50
pens 0.5-1.0 300
pencil 0.5-1.0 40
pens 1.0-1.5 150
pencil 1.0-1.5 45
I want a query that displays the above table result as below:
RANGE Pens Pencils
------------------------------
0-0.5 200 50
0.5-1.0 300 40
1.0-1.5 150 45
Any ideas about this? Thanks in advance!

Answer 1

What you are trying to do is known as a PIVOT. This is when you transform data from rows into columns. Some databases have a PIVOT function that you can take advantage of but you did not specify which RDBMS.
If you do not have a PIVOT function then you can replicate the functionality using an aggregate function along with a CASE statement:
select `range`,
sum(case when class_name = 'pens' then `count` end) pens,
sum(case when class_name = 'pencil' then `count` end) pencils
from yourtable
group by `range`
See SQL Fiddle with Demo
Note: the backticks are for MySQL, if SQL Server then use a square bracket around range and count. These are used to escape the reserved words.
If you are working in an RDBMS that has a PIVOT function, then you can use the following:
select *
from
(
select class_name, [range], [count]
from yourtable
) src
pivot
(
sum([count])
for class_name in ([pens], [pencil])
) piv
See SQL Fiddle with Demo
Both will produce the same result:
| RANGE | PENS | PENCIL |
---------------------------
| 0-0.5 | 200 | 50 |
| 0.5-1.0 | 300 | 40 |
| 1.0-1.5 | 150 | 45 |
The above will work great if you have a known number of values for class_name, if you do not then, depending on your RDBMS there are ways to generate a dynamic version of this query.
In SQL Server a dynamic version will be similar to this:
DECLARE #cols AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX)
select #cols = STUFF((SELECT distinct ',' + QUOTENAME(CLASS_NAME)
from yourtable
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set #query = 'SELECT [range], ' + #cols + ' from
(
select CLASS_NAME, [RANGE], [COUNT]
from yourtable
) x
pivot
(
sum([COUNT])
for CLASS_NAME in (' + #cols + ')
) p '
execute(#query)
See SQL Fiddle with Demo

Answer 2

This pivot query can be used in all major DBMS. The trick is getting the bad column names 'range' and 'count' quoted property.
SQL Server (below) uses [], MySQL uses backticks (`), Oracle uses double quotes, SQLite can use any of the preceding.
select [range],
sum(case when class_name='pens' then [count] else 0 end) Pens,
sum(case when class_name='pencil' then [count] else 0 end) Pencils
from tbl
group by [range]
order by [range];

Answer 3

here is the dynamic version of the pivot:
IF OBJECT_ID('tempdb..#T') IS NOT NULL
DROP TABLE #T
CREATE TABLE #T(
[CLASS_NAME] VARCHAR(20) NOT NULL
, [range] VARCHAR(10) NOT NULL
, [count] INT NOT NULL
)
INSERT INTO #T([CLASS_NAME], [range], [count]) VALUES ('pens', '0-0.5', '200')
INSERT INTO #T([CLASS_NAME], [range], [count]) VALUES ('pencil', '0-0.5', '50')
INSERT INTO #T([CLASS_NAME], [range], [count]) VALUES ('pens', '0.5-1.0', '300')
INSERT INTO #T([CLASS_NAME], [range], [count]) VALUES ('pencil', '0.5-1.0', '40')
INSERT INTO #T([CLASS_NAME], [range], [count]) VALUES ('pens', '1.0-1.5', '150')
INSERT INTO #T([CLASS_NAME], [range], [count]) VALUES ('pencil', '1.0-1.5', '45')
DECLARE #PivotColumnHeaders VARCHAR(MAX)
SELECT #PivotColumnHeaders = STUFF((
--SELECT DISTINCT TOP 100 PERCENT
SELECT DISTINCT '],[' + [CLASS_NAME]
FROM #T
ORDER BY '],[' + [CLASS_NAME]
FOR XML PATH('')
), 1, 2, '') + ']';
DECLARE #PivotTableSQL NVARCHAR(MAX)
SET #PivotTableSQL = N'
SELECT *
FROM (
SELECT *
FROM #T
) AS PivotData
PIVOT (
SUM([count])
FOR [CLASS_NAME] IN (
' + #PivotColumnHeaders + '
)
) AS PivotTable
'
EXECUTE(#PivotTableSQL)