Nested Sql select statement - sql

Question

Can anyone tell me what is wrong with the following sql query ?
Select *,
(SELECT [DiseaseID], COUNT(*) AS [Rank] FROM [DiseaseSymptom] WHERE
([SymptomID] IN(1, 5)) GROUP BY [DiseaseID] ORDER BY [Rank] DESC)
FROM Disease WHERE GenderID in (1, 3)
I have 2 tables one contains disease and the gender it is associated with
Disease
+-----------+-------------------+----------+
| DiseaseID | DiseaseName | GenderID |
+-----------+-------------------+----------+
| 1 | Fever | 3 |
| 2 | Flu | 3 |
| 3 | Lady Disease | 2 |
| 4 | Gentlemen Disease | 1 |
+-----------+-------------------+----------+
Gender 1 = Male, 2 = Female, 3 = Common
And a Symptom Disease Matrix like this
DiseaseSymptom
+-----------+-----------+----------+
| DiseaseID | SymptomID | DissymID |
+-----------+-----------+----------+
| 1 | 1 | 1 |
| 1 | 2 | 3 |
| 1 | 4 | 4 |
| 2 | 1 | 5 |
| 2 | 3 | 9 |
| 2 | 4 | 6 |
| 2 | 5 | 7 |
+-----------+-----------+----------+
I get symptoms from user and match it in the DiseaseSymptom table and rank it according to the number of symptoms matched (inner sql statement)
In the outer statement I simply want get the result from inner statement and evaluate whether it belongs to specific gender. The error I get when I try to run the above query is
The ORDER BY clause is invalid in views, inline functions, derived tables, subqueries, and common table expressions, unless TOP or FOR XML is also specified.

Answer 1

Subqueries in select clause must only generate a scalar value, not a resultset with multiple columns or rows. if you want both then put the subquery in the from clause (properly correlated), and refer to the two different vqlues in the select clause
Select d.*, z.DeseaseId, z.Rank
FROM Disease d
join (SELECT DiseaseID, COUNT(*) Rank
FROM DiseaseSymptom
WHERE SymptomID IN(1, 5)
GROUP BY DiseaseID) Z
On z.DeseaseId = d.DeseaseId
WHERE GenderID in (1, 3)
Order By z.Rank

Answer 2

You are using a subquery with group by. Your intention is to have a correlated subquery. The problem is that the subquery is returning more than one row. I think this is what you want:
Select d.*,
(SELECT COUNT(*) AS [Rank]
FROM [DiseaseSymptom] ds
WHERE [SymptomID] IN (1, 5)) AND ds.DiseaseId = d.DiseaseId
)
FROM Disease d
WHERE GenderID in (1, 3);

Answer 3

You should use Common Table Expression (cte) like this:
with cte as (SELECT [DiseaseID], GenderID, COUNT(*) AS [Rank] FROM [DiseaseSymptom] WHERE
([SymptomID] IN(1, 5)) GROUP BY [DiseaseID],GenderID ORDER BY [Rank] DESC)
select * FROM cte WHERE GenderID in (1, 3)
Hope this help ;)

Answer 4

There is really no need to have a nested query, just join and filter
SELECT d.DiseaseID, d.DiseaseName, d.GenderID
, Symptoms = Count(ds.SymptomID)
FROM Disease d
INNER JOIN DiseaseSymptom ds ON d.DiseaseID = ds.DiseaseID
WHERE ds.SymptomID IN (1, 5)
AND d.GenderID IN (1, 3)
GROUP BY d.DiseaseID, d.DiseaseName, d.GenderID
ORDER BY Count(SymptomID) Desc
SQLFiddle Demo