>Hi,
>
>Here is another Sunday challenge I am trying to resolve. Here is a simplified case, but close to what I am trying to do.
>
>Table has the following 2 columns/fields, sorted by ID, and the values as shown below:
>
>

>ID                 CATEGORY
>0001             ABC
>0002             NBC
>0002             NBC
>0002             NBC
>0003             CBS
>0004             123
>0004              ABC
>0005             567
>0005             567
>0006             009
>0006             123
>0006             123
>

>I want to delete all records where the ID and CATEGORY are matching, it does not matter 1 records or more. So the recult would be as follows:
>

>ID                 CATEGORY
>0004             123
>0004              ABC
>0006             009
>0006             123
>0006             123
>

>
>Because at least one of the records in the group by ID is not matching column CATEGORY.
>
>Can this be done?
>
>TIA

Here it goes (hopefully):

SET DELETED ON

CREATE CURSOR Original (ID Character(4), Category Character(3))

LOCAL SourceData AS String
LOCAL ARRAY SourceLines(1)
LOCAL LoopIndex AS Integer

TEXT TO m.SourceData NOSHOW
0001             ABC
0002             NBC
0002             NBC
0002             NBC
0003             CBS
0004             123
0004              ABC
0005             567
0005             567
0006             009
0006             123
0006             123
ENDTEXT

FOR m.LoopIndex = 1 TO ALINES(m.SourceLines, m.SourceData)
	INSERT INTO Original VALUES (LEFT(m.SourceLines(m.LoopIndex), 4), RIGHT(m.SourceLines(m.LoopIndex), 3))
ENDFOR

DELETE Original ;
	FROM (SELECT ID, COUNT(DISTINCT Category) AS Categories FROM Original GROUP BY ID HAVING Categories = 1) AS Matcher ;
	WHERE Original.Id = Matcher.Id

BROWSE