Search found 9 matches

by Billy
Thu Oct 23, 2003 3:42 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: DS Basic
Replies: 7
Views: 3288

DS Basic

I need to check a range of values for a column and am creating a routine to check for the list of possible values Begin Case Case Arg1 = 'Fiat' or Arg1 = 'FIAT' or Arg1 = 'fiat' THEN .... Case Arg1 = 'Ford' or Arg1 = 'FORD' or Arg1 = 'ford' THEN .... .... End Does D/S have a faster method like the I...
by Billy
Tue Aug 26, 2003 12:16 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: DS Best Practices CDD question
Replies: 9
Views: 5601

Hi Micheal,

I definitely welcome any material regarding CDD, tips and tricks relating to ETL. But let's start with the lighter materials, fairly new to this area [:)]


cheers
Billy
by Billy
Tue Aug 26, 2003 12:02 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: JOIN tables in Datastage
Replies: 14
Views: 5872

Hi Stephan & Rasi, Thanks for the advise, I am extracting from the source into the a temp DBMS which supports sub-select statements and then I extract it out into the staging area again for JOINs I'm new to the ETL and perhaps this may sound like a naive question - from your experiences is it no...
by Billy
Fri Aug 22, 2003 9:49 pm
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: DS Best Practices CDD question
Replies: 9
Views: 5601

DS Best Practices CDD question

Hi All, I've been going thru the course materials and I noticed that the CDD (Change Data Detection) method requires the extraction of the entire table into bkt, src and hash files. Based on all your vast and many years of experience, is this the best CDD method for a DBMS w/out a timestamp feature?...
by Billy
Fri Aug 22, 2003 9:01 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: JOIN tables in Datastage
Replies: 14
Views: 5872

Stephan Yes, this works. We also have another similiar problem but this is even more complex. TranNo TranCnt TranCo Amt TranDt 000001 0000000 ABCCom 100 1/1/01 000002 0000000 DEFCom 100 1/1/01 000003 0000000 GHICom 100 1/1/01 000001 0000001 ABCCom 100 3/1/01 000002 0000001 DEFCom 100 3/1/01 000003 0...
by Billy
Fri Aug 22, 2003 6:10 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: JOIN tables in Datastage
Replies: 14
Views: 5872

Hi Stephan Looks workable and good. However I will still have to perform an additional SELECT statement to get the cummulative figures from the previous week. All these are part of a pretty complex subset of the final table we wish to get. We need to accumulate based on a certain range and the full ...
by Billy
Fri Aug 22, 2003 5:34 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: JOIN tables in Datastage
Replies: 14
Views: 5872

Hi Stephan The objective is not to get the max date but a cummulative amount. Let me give an example Table A ------- IDNo Tran Date Tran Amount AAA 1/1/01 100 BBB 2/1/01 100 CCC 3/1/01 100 AAA 1/1/02 100 BBB 2/1/02 100 CCC 3/1/02 100 AAA 1/1/03 100 BBB 2/1/03 100 CCC 3/1/03 100 AAA 1/4/03 100 BBB 2/...
by Billy
Fri Aug 22, 2003 4:35 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: JOIN tables in Datastage
Replies: 14
Views: 5872

Hi Rasi,
Yes, I have thot of that option unfortunately the DBMS I am using does not support sub-select statements. FYI, the DBMS is Progress.

cheers
Billy
by Billy
Fri Aug 22, 2003 4:07 am
Forum: IBM<sup>®</sup> Infosphere DataStage Server Edition
Topic: JOIN tables in Datastage
Replies: 14
Views: 5872

JOIN tables in Datastage

Hi, I have a table transaction (IDNo, Date, Tran amount) which I have to UNION back again based on the IDNo to get the cummulative tran amount. Something like : SELECT a.IDNo, a.Date, (SELECT SUM(b.Tran amount) from table b where b.Date < a.Date and b.IDNo = a.IDNo group by b.IDNo) from table a Can ...