"Mark C. Stock" <mcstockX_at_Xenquery .com> wrote:
>
>"Turkbear" <john.g_at_dot.spamfree.com> wrote in message
>news:v5kovv4ifip3c7v8teaojjv2fooe3si28l_at_4ax.com...
>| harve_projects_at_yahoo.com (HRR) wrote:
>|
>| >Hi,
>| >
>| >I have the following table
>| >
>| >1001 2 3
>| >1001 3 5
>| >1001 5 7
>| >1002 5 4
>| >1002 6 2
>| >
>| >I need to get a result
>| >
>| >1001 2 3
>| >1002 6 2
>| >
>| >How is it possible ?
>| Column names would have helped..assume a, b and c
>|
>| Select a,min(b),min(c) from sometable
>| group by a;
>|
>|
>
>not exactly -- your query returns
>
>1001 2 3
>1002 5 2
>
>either there's a typo by the OP, or, if OP really wants the row with the
>minimum value of 'c', you need to do a pairwise compare with a subquery (see
>my reply to the same OP in comp.databases.oracle)
>
>-- mcs
>
Yep..
( The results from my query are, in fact, the MIns for both b and c for each a, but not the row that has the min c value -
because it does not have the min b)
I'm not sure if it is a typo, but if not, then the logic is a little strange:
Get the minimum b and minimum c when a = 1001 but Max(b),Min(c) when 1002????
I assumed ( yes, I know) then the OP only cared about the values not the rows the values were in.
Received on Wed Jan 07 2004 - 14:04:53 CST
