I am trying to write a validation for the following set of data:
SSYS | Material_Number | Characteristic | Description
001 | 000000000001111 | SH_DESC | TEST
001 | 000000000001111 | DESIGN_TYPE | NULL
001 | 000000000001111 | VOLTAGE | NULL
001 | 000000000009999 | SH_DESC | TEST2
001 | 000000000009999 | OPER_METHOD | LIGHT
001 | 000000000009999 | FILTER_TYPE | Filter element,Air
001 | 000000000014560 | SH_DESC | Horn,Signal
001 | 000000000014560 | DIMENSION_SIZE | NULL
001 | 000000000014560 | FILTER_TYPE | NULL
I would like to group by the Material_Number and count as 1 (ie. true) if within the Material_Number group, the SH_DESC description is NOT NULL and all other characteristics’ descriptions IS NULL. So, in this case my result would be:
SSYS | Material_Number | Characteristic | Description | COUNT
001 | 000000000001111 | SH_DESC | TEST | 1
001 | 000000000009999 | SH_DESC | TEST2 | 0
001 | 000000000014560 | SH_DESC | Horn,Signal | 1
My attempt:
Select COUNT (*), SSYS, Material_Number, Characteristic, Description
From myDB where (Characteristic = 'SH_DESC' AND DESCRIPTION IS NOT NULL) AND (Characteristic NOT IN ('SH_DESC') IS NULL)
GROUP BY SSYS, Material_Number, Characteristic, Description HAVING COUNT (*) < 2
Any help is much appreciated!
Try: