Bruce, That's what the HAVING clause is for in the SELECT syntax.
Presuming you want to look at the details, but group by and count by the plate number: SELECT columnlist FROM tablename + WHERE mvplateno IN + (SELECT myplateno FROM sametablename t2 GROUP BY myplatenum HAVING COUNT (*) > 1) Bill On Thu, Aug 12, 2010 at 3:41 PM, Bruce Chitiea <[email protected]>wrote: > All: > > After importing Excel vehicle records into a receiving table, I ran a > "SELECT DISTINCT ..." against all necessary columns in a working copy of > the receiving table. > > This revealed 'semi-distinct' records such as these: > > mvplateno mvplatestateid mvcolorid mvmfgrid mvmodel ... > (Indexed) (Indexed) > > 112 BLG CA WHT CHV MALIBU > 112 BLG CA WHT -0- -0- > 112 BLG CA -0- CHV IMPALA > 112 BLG FL -0- -0- -0- > > 123 ABC CA RED HON CIVIC > 123 ABC CA HON RED ACCORD > 123 ABC CA MAR TOY ACCORD > > I'm thinking a form-hosted, cursor-stepped editing process; producing a > list of truly distinct records keyed off the license plate number > (mvplateno). > > Where I'm hung up is how to select out only those records where > (count(mvplateno) > 1). > > I'd sure appreciate a pointer in the correct use of SELECT to generate a > subset based solely on record-count. > > I'm working at the crawl level, and every bit of working code is a > victory; so I'm game to work out the coding myself. But I'd sure > appreciate direction. > > bruce chitiea > safesectors, inc. > > Turbo V-8 Win 8.0.23.30809 > Windows 7 > > >
