Question

In SQL Server 2008 R2, I’m trying to work out the rank of an individual item, as determined by the number of occurances. I can work out the rank of a number of items, i.e. the top ten instances of a colour, but I can’t get the rank of a specific colour.

I.E I have a table of Vehicles, which has a Vehicles.ColourID column.

For a specific ColourID I want to determine the rank of it.

i.e.

• ColourID 7 has 10 instances
• ColourID 10 has 5 instances
• ColourID 2 has 1 instance

In that result-set, ColourID 7 has a rank of 1, ColourID 10 has a rank of 2 and so on.

I would like to query for ColourID 10 and find it has a rank of 2. This is what I am struggling with.

I’ve tried using derived tables and common table expressions, but in both cases they bring back a rank of 1.

Does anyone know how I could query for the rank of a single item?

Answer 1

To rank all the colours:

;with cteCount as (
    select ColourID, count(*) as ColourCount
        from Vehicles
        group by ColourID
)
select ColourID, ColourCount, 
       row_number() over (order by ColourCount desc) as Rank
    from cteCount

You can then extend this technique to query for the rank of a single item:

;with cteCount as (
    select ColourID, count(*) as ColourCount
        from Vehicles
        group by ColourID
),
cteRank as (
    select ColourID, ColourCount, 
           row_number() over (order by ColourCount desc) as Rank
        from cteCount
)
select ColourID, ColourCount, Rank
    from cteRank
    where ColourID = 10