Question

I have a table

tblDiseaseTrack

TrackID DiseaseID  PostalCode 
1       3          111
2       3          111
3       2          111
4       1          222
5       2          222
6       4          111
7       1          222
8       5          333
9       5          333
10      5          333

I want to write a query to display the disease id and the postal code of the maximum repeated DiseaseId for each postalcode as follows,

DiseaseID  PostalCode 
3          111
1          222
5          333

PLEASE any help would be much appreciated. i tried evrything and couldnt find any help..Thank you again 🙂

Answer 1

select diseaseid, postalcode
from 
(
    select
        postalcode, 
        diseaseid,
        row_number() over (partition by postalcode order by count desc) as row
    from 
    (
        select postalcode, count(postalcode) as count, diseaseid
        from tblDiseaseTrack
        group by postalcode, diseaseid
    ) as T1
) as T2
where row = 1
order by postalcode

This returns exactly what you wanted:

DiseaseID  PostalCode 
3          111
1          222
5          333

---

EDIT:

Same query like above, but with a JOIN to get the city name from a second table:

select diseaseid, T2.postalcode, city
from 
(
    select
        postalcode, 
        diseaseid,
        row_number() over (partition by postalcode order by count desc) as row
    from 
    (
        select postalcode, count(postalcode) as count, diseaseid
        from @tblDiseaseTrack
        group by postalcode, diseaseid
    ) as T1
) as T2
inner join tblcity on T2.postalcode = tblcity.postalcode
where row = 1
order by postalcode

Note that I used the same column name PostalCode in the city table as well, so I have to prefix the PostalCode in the outmost query: T2.postalcode instead of postalcode.

Otherwise the query would crash with the message:

Ambiguous column name ‘postalcode’

…because SQL Server wouldn’t know which of the two PostalCodes I want.