Transforms

[srikanthd]

Hi Folks,
Any one could explain me..what are the built in transforms present in the DataStage Designer.I know about various stages in DS like sequential,ODBC,Transform,Aggregator.... But i found only some Built-ins for Date functions(conversions)..Actually in DS how can we do transforms like Lookup,filter,Sequence Generation,Normalizing,Update Strategy...
It'll be helpful for me
Thanks in Advance
SrikanthD.

 

[JGirl]

Hi, without meaning to sound rude, I really think it would be worthwhile for you to read the doco that comes with datastage to get some sort of idea on how the whole package fits together and to find out how to meet your requirements. Ill give some quick answers anyway though....

Functions:
Read the DataStage server job developers guide - there are heaps of them that can trim strings, convert dates, etc, etc, etc

Lookups:
To achieve a lookup, you will need a transformer stage, and the appropriate stage type for your lookup data (eg. a hash file). Then you need to link the transformer and the hash file stage using a reference (as opposed to stream) link from the hash file to the transformer. Then when you double click on the transformer, you will see the columns in the transformer and the hashed file stage. You will need to associate the columns here (ie. match the data items that you are doing the lookup on). In the next stage, you may want to add some constraints to the stage so that only certain data is passed through (such as all rows that had a successful lookup)

Filters:
to filter the data, based on certain conditions, use the 'constraints' section within the transformer stage window. For example, you may want to include a constraint that only passes transactions / whatever for a particular date to the next stage.

Sequence Number Generation:
DataStage has a whole stack of variables that can be used for sequence number generation. For example, adding @INROWNUM to a field within a transformer will assign the input row number to the field.

Normalising:
This is done within the stage that your data to be normalised is to be coming from (eg. a multivalued files) - do this in the 'columns' tab of the input data source

Update Strategy:
*I have no idea what you mean*.

Thats just a start, I think you will find it beneficial to read the doco though, as there is so much you can do with datastage that I cant explain in a forum posting! Also, if you already have some developed jobs, take a look at them and see how they fit together.

Hope this helps a little
Cheers
J

 

[srikdw]

Hi J,
Thanks a lot for your response.I guess u know that i am new to DataStage.Anyways ur response is quite useful for me.
Srik.