|
Hi all SQL Anywhere 8.0 I am using the following SQL successfully SELECT cr_no, dated, height, weight FROM malaviya_vitals WHERE cr_no IN ( SELECT cr_no FROM mal_diagnosis WHERE mal_diagnosis.diag_code LIKE M05.%' ) order by cr_no, dated; with a result set of 6605 rows However I want to refine the result set a little more to get only the weight and height data of this group of patients only the first time the patient was seen. [dated = MIN (dated)] Need help to construct the exact syntax (should lead to a total of around 1215 rows = no. of patients with diag_code M05.---) Any help will be appreciated Regards |
|
Try adding the following subquery to the outer AND dated = ( SELECT MIN( t.dated ) FROM malaviya_vitals AS t WHERE t.cr_no = malaviya_vitals.cr_no ) |
|
You could also try: SELECT cr_no, min(dated) as theDated, height, weight FROM malaviya_vitals WHERE cr_no IN ( SELECT cr_no FROM mal_diagnosis WHERE mal_diagnosis.diag_code LIKE ''%M05.%'' ) group by cr_no, height, weight order by cr_no, theDated This will create a worktable as part of the query optimisation, but if your caching is sufficient and fast enough, your performance should be good too. 1
I think that gives a row for the earliest date of each unique (cr_no, height, weight) combination, rather than only a row for the first time each patient was seen.
(19 Feb '13, 10:56)
Graham Hurst
|
