Question:
How do I do a “group by” SQL statement? I want to return all rows and columns with ** records. Here is the data sample:
columnA columnB columnCAAA 2334 jan/01/99 **AAA 4 mar/05/99 BBB 32 feb/20/99 BBB 3 feb/23/99BBB 32 feb/24/99 **CCC 32439 jan/23/99 **DDD 666 may/01/99 EEE 444 mar/03/99 **
Conditions:
- SELECT columnA columnB columnC
- GROUP BY ColumnA
- columnC <= mar/04/99 and the max row, which is in the same group
Answer:
What’s required is to treat this problem in two passes.
First isolate the records that are the case by selecting the max of Column B grouped by Column A. That will at least get you the highest value for the group.
Next, you need another pass to eliminate the outlying dates.
Sorry this solution isn’t more elegant.
