Joins in Transformer 1

[stigejg]

Hi,

I have a (may be) silly question.

I have two IQD files:
First file contains my product-structure (key field product-number)second file contains my production-data identified by my product-number.

But I can not find out how I can join the this two IQD files in Transformer to be able to make the cube with the proper drill-down on product-category, product-group and finaly product by using this two IQD-files. I have solved this by using MS Access and joined my product-structure and production-data before making one IQD-file containg all data and gives me the desired drill-down facilities.

It si a lot of work to attach tables to Access (from a movexdatabase) via ODBC and make querys there before making IQD-files.

Therefor is there facilities in transformer to join two IQD-files and then make the power-cube with data from both?

Kind regard Stig

 

[stevev]

Hi,
we always use multiple iqds in our transformer model. It's faster for us to let transformer make the join between iqds than create one unique iqd that will contain everything. To make join between iqds you have to use the same column name. Let say you have two iqd :

One for customer containing :
Customer Id
Customer Name
...

One for Orders :
Order Id
Customer Id
Product Id
Amount
...

The join between these iqds will be "Customer Id". You can change the name of the column in the property tab of the field.


Hope this help.
Steve

 

[stigejg]

Hi Steve,

I have done that, but i transformer gives me this errormessage when I am checking the model:

Warning: (TR2706) Level PPC-Prodno can't be reached in drill-down path By Produktgruppe for data source prognose. A level can be the first source level, be directly below a source level, or be designated as a unique level. All measures in data source prognose will have zero values in the PowerCube.

Error: (TR2703) A column in data source 'prognose' matches level 'PPC-Prodno' in dimension 'Produktgruppe' by name. Values from this data source column cannot be related to categories in this level because insufficient context is specified in the data source. To resolve this problem, either associate columns in the data source to higher levels in the dimension or, if appropriate, designate the level as a unique level.

Check model is finished. 1 error(s), 1 warning(s) found.

Stig

 

[n30er]

Just make your product number level unique (properties) and you shouldn't have any more errors. The key here is that product no better be unique, if it isn't you will need to create a unique field through concatenated columns.

 

[stigejg]

Hello again,

It works, I just had forgotten to make my product number unique. I have to take me a litle better time next time and read a litle bit.

Thanks a lot

Stig