Comments on: Duplicate Records

By: lorit

lorit — Mon, 04 Jan 2010 16:53:07 +0000

Thank you Phil and Tom! I appreciate your help and quick response.

By: jennymack

jennymack — Mon, 04 Jan 2010 15:32:26 +0000

Phil and Tom,

Thanks for contributing so much to this answer! Lorit, let us know if they’ve helped you reach a solution.

Thanks,
Jenny

By: philpl1jb

philpl1jb — Mon, 04 Jan 2010 02:54:44 +0000

Ohh – you are trying to join 5 comments with 4 adjustment records on the taxbill_no – you will get 20 output records — but you might want 4. Ideally the adjustment record and the comment would have a comment number field so you can join these files on the taxbill_no and the comment#. With what you have you could join them on the taxbill_no and the amount giving 3 records or the taxbill_no and the credit giviing 1 more record but this plan will fail if there are two payments for the same amount or two credits for the same amount. So there is no way I can see to get the adjustments linked with the comments without a database change.

By: philpl1jb

philpl1jb — Thu, 31 Dec 2009 16:47:21 +0000

SELECT “T16.GNAM”.N_NAME, “T16.GNAM”.N_CITY,
sum(”T16.CADJ”.AD_TOTAL_ADJUSTMENT)
FROM QS36F.”T16.CPRP” AS “T16.CPRP”,
QS36F.”T16.GNAM” AS “T16.GNAM”,
QS36F.”T16.CADJ” AS “T16.CADJ”,
QS36F.”T16.CRDC” AS “T16.CRDC”

WHERE “T16.CADJ”.AD_TAXBILL_NO = “T16.CPRP”.CP_TAXBILL_NO
AND “T16.CPRP”.CP_PRIMARY_TAXPAYER = “T16.GNAM”.N_ID_NR
AND “T16.CADJ”.AD_ADJUSTMENT_DATE > 20091106 AND “T16.CADJ”.AD_ADJUSTMENT_DATE < 20091205 AND “T16.CADJ”.AD_TOTAL_ADJUSTMENT > -100.00
and “T16.CADJ”.AD_REASON IN (’C ‘, ‘R ‘)
Group by “T16.GNAM”.N_NAME, “T16.GNAM”.N_CITY;

Sorry, I think that’s a little closer. It will sum by name, city which would be correct provided every name, city is unique – not likely. It cannot include the comments and you may want to add an order by clause.
Phil
Phil

By: philpl1jb

philpl1jb — Thu, 31 Dec 2009 16:40:19 +0000

SELECT “T16.GNAM”.N_NAME, “T16.GNAM”.N_CITY,
sum(“T16.CADJ”.AD_TOTAL_ADJUSTMENT) FROM QS36F.”T16.CADJ” AS “T16.CADJ”, QS36F.”T16.CRDC” AS “T16.CRDC”, QS36F.”T16.CPRP” AS “T16.CPRP”, QS36F.”T16.GNAM” AS “T16.GNAM”
WHERE “T16.CADJ”.AD_TAXBILL_NO = “T16.CPRP”.CP_TAXBILL_NO
AND “T16.CPRP”.CP_PRIMARY_TAXPAYER = “T16.GNAM”.N_ID_NR
AND “T16.CADJ”.AD_ADJUSTMENT_DATE > 20091106 AND “T16.CADJ”.AD_ADJUSTMENT_DATE < 20091205 AND “T16.CADJ”.AD_TOTAL_ADJUSTMENT > -100.00
and “T16.CADJ”.AD_REASON IN (‘C ‘, ‘R ‘)
Group by “T16.GNAM”.N_NAME, “T16.GNAM”.N_CITY;

This should give you the totals by person provided every name, city is unique – not likely.
Phil

By: tomliotta

tomliotta — Thu, 31 Dec 2009 09:46:39 +0000

SELECT Distinct
   "T16.GNAM".N_NAME,
   "T16.GNAM".N_CITY,
   "T16.CADJ".AD_TOTAL_ADJUSTMENT,
   "T16.CRDC".RD_COMMENT

  FROM

     QS36F."T16.CPRP" AS "T16.CPRP",
     QS36F."T16.GNAM" AS "T16.GNAM",
     QS36F."T16.CADJ" AS "T16.CADJ",
     QS36F."T16.CRDC" AS "T16.CRDC"

  WHERE

     "T16.CPRP".CP_PRIMARY_TAXPAYER = "T16.GNAM".N_ID_NR AND
     "T16.CPRP".CP_TAXBILL_NO = "T16.CADJ".AD_TAXBILL_NO AND
     "T16.CADJ".AD_TAXBILL_NO = "T16.CRDC".RD_TAXBILL_NO AND
     "T16.CADJ".AD_ADJUSTMENT_DATE > 20091106 AND
     "T16.CADJ".AD_ADJUSTMENT_DATE < 20091205 AND
     "T16.CADJ".AD_TOTAL_ADJUSTMENT > -100.00 and
     "T16.CADJ".AD_REASON IN ('C ', 'R ')

  order BY "T16.GNAM".N_NAME

Phil is probably on the right track. I poked around with the query to see if it could be made to read easier and perhaps make a little more sense. (And I see TAXBILL_NO provides the missing relationship from the previous question.) You might consider re-ordering your query components closer to how I put them here. (I'd be a little concerned about ordering purely by N_NAME since I can guarantee that two taxpayers at the same mailing street address can have the same name. It has happened to me personally with my father.) As part of the re-ordering, I have T16.CRDC as the last file in the chain. That's just from coming to a similar conclusion to Phil's. The query appears to be from a SQL client, possibly iSeries Navigator 'Run SQL Scripts'. If you are intending to print a report, you might use this as the basis for a QM query. It's been a while, but I think the QM form could be adapted to print only the first line of the "unique" values from the first three files while printing each comment on subsequent lines. It would at least make the 'duplicates' read sensibly. Tom