Re: Selecting unique count on non-unique index

Try: select a_field1, a_field2, a_field3, count(*) from a_table group by 1,2,3 having count(*) > 1 This will give you number of records that has same keys (a_field1, a_field2, and a_field3). Hope this helps. Susik slee@gus.net > > Hi All, > > I've spotted a table in our database that has no unique key but is > nevertheless updated using a WHERE clause that identifies a > uniqueness. > > That is: > > UPDATE a_table SET a_field = a_value > WHERE a_field1 = a_value1 > AND a_field2 = a_value2 > AND a_field3 = a_value3> > The fields a_field1, a_field2, a_field3 have not been collated to form > a unique key ( which I intend to do! ). > Having found the total number of records on the table, how do I now > determine the number of records based on the uniqueness of the three > fields sited above? > > Cheers, Ian. > Tedious 'office speak' part 3: "Let's go scuba in the think tank." >