Multiple counts on a single column SQL - sql

Question

I am currently running a query like the following:
SELECT a.ID, a.ContactID, a.Code,
FROM tableA a
JOIN (SELECT ContactID, Code
FROM tableA
WHERE ContactID IS NOT NULL
GROUP BY Code, ContactID
HAVING COUNT(Code) > 1) b
ON (a.Code = b.Code AND a.ContactID = b.ContactID)
WHERE a.ContactID IS NOT NULL
ORDER BY a.Code
This returns data that looks like the folloing:
table : a
+-------+-----------+-----------+
| ID | ContactID | Code |
+-------+-----------+-----------+
| 1 | 111 | abcd2 |
| 2 | 111 | abcd2 |
| 3 | 222 | abcd1 |
| 4 | 222 | abcd1 |
| 5 | 222 | abcd1 |
| 6 | 222 | abcd1 |
+-------+-----------+-----------+
So as you can see I get ContactID's that have more then one of the same Code.
The problem with this is, is that I don't want all this output (real table is much larger). I want a COUNT to go along side the Code column and just show one row for each iteration of Code. Like the following:
+-------+-----------+-----------+------+
| ID | ContactID | Code |COUNT |
+-------+-----------+-----------+------+
| 1 | 111 | abcd2 | 2 |
| 3 | 222 | abcd1 | 4 |
+-------+-----------+-----------+------+
Any help on this would be great and I hope I have explained my problem well enough. If not please ask for more information and if this has been answered before please point in that direction.
Thanks.

Answer 1

Your solution and other answers are way to complicated, you don't need the self join when you're simply aggregating with HAVING Count(x) > 1:
SELECT MIN(ID), ContactID, Code, COUNT(Code) AS [COUNT]
FROM tableA
WHERE ContactID IS NOT NULL
GROUP BY Code, ContactID
HAVING COUNT(Code) > 1
Full solution:
SQL Fiddle
CREATE TABLE TableA
([ID] int, [ContactID] int, [Code] varchar(5))
;
INSERT INTO TableA
([ID], [ContactID], [Code])
VALUES
(1, 111, 'abcd2'),
(2, 111, 'abcd2'),
(3, 222, 'abcd1'),
(4, 222, 'abcd1'),
(5, 222, 'abcd1'),
(6, 222, 'abcd1')
;
Query 1:
SELECT min(id), ContactID, Code, count(Code) as [COUNT]
FROM tableA
WHERE ContactID IS NOT NULL
GROUP BY Code, ContactID
HAVING COUNT(Code) > 1
Results:
| | ContactID | Code | |
|---|-----------|-------|---|
| 1 | 111 | abcd2 | 2 |
| 3 | 222 | abcd1 | 4 |

Answer 2

I would use exists instead of subquery :
select min(a.id) as id, a.ContactID, a.Code, count(*) as Cnt
from tableA a
where exists (select 1
from tableA a1
where a1.ContactID = a.ContactID and
a1.Code = a.Code and
a1.id <> a.id
)
group by a.ContactID, a.Code;

Answer 3

sub-query
select min(ID) as id, ContactID,Code,count(*) as cnt from
(SELECT a.ID, a.ContactID, a.Code
FROM tableA a
JOIN (SELECT ContactID, Code
FROM tableA
WHERE ContactID IS NOT NULL
GROUP BY Code, ContactID
HAVING COUNT(Code) > 1) b
ON (a.Code = b.Code AND a.ContactID = b.ContactID)
WHERE a.RetailContactID IS NOT NULL
ORDER BY a.Code
) t group ContactID,Code

Answer 4

;WITH CTE AS
(
SELECT a.ID, a.ContactID, a.Code,
FROM tableA a
JOIN (SELECT ContactID, Code
FROM tableA
WHERE ContactID IS NOT NULL
GROUP BY Code, ContactID
HAVING COUNT(Code) > 1) b
ON (a.Code = b.Code AND a.ContactID = b.ContactID)
WHERE a.RetailContactID IS NOT NULL
)
SELECT ID, ContactID, Code, COUNT(*) AS Cnt
FROM CTE
GROUP BY ID, ContactID, Code
ORDER BY 1, 2, 3

Answer 5

Extend your SQL query with one more grouping:
SELECT min(a.ID), a.ContactID, a.Code, count(*)
...
GROUP BY a.ContactID, a.Code
ORDER BY a.Code

Answer 6

Use group by in your select query
select x.ContactID, x.Code, [count] = count(x.id) from (
select id = 1 , ContactID = 111 , Code = 'abcd2' union all
select 2 , 111 , 'abcd2' union all
select 3 , 222 , 'abcd1' union all
select 4 , 222 , 'abcd1' union all
select 5 , 222 , 'abcd1' union all
select 6 , 222 , 'abcd1') x group by x.Code, x.ContactID