Zeroout Process

[Abinaya]

Hi
I want to ZeroOut a cube based on 4 parameters. The parameters are fields in the excel data source.
What is the best way to do it?

If I include the ZeroOut process in the prolog then I can't retrive the parameter values ?


regards,

Abi

[Michel Zijlema]

Hi
I want to ZeroOut a cube based on 4 parameters. The parameters are fields in the excel data source.
What is the best way to do it?

If I include the ZeroOut process in the prolog then I can't retrive the parameter values ?


regards,

Abi

I assume you're referring to variables (data in the data source) instead of parameters.
Another assumptions is that the value of the regarding 4 variables is the same on each row.
What you could do in that case is:
- setup the ZeroOut view on the Prolog tab, without assigning the element selection subsets (for the 4 variables) to the view,
- initialize a flag variable ZeroOut = 1 on the Prolog tab
- on the MetaData (or Data if no metadata manipulation is required) you can have a conditional statement like:
IF(ZeroOut = 1);
assign variable values to ZeroOut view selection subsets
ZeroOut = 0;
ENDIF;
The above is partially pseudo code and will only work when my two assumptions are correct...

Michel

[tomok]

Hi
I want to ZeroOut a cube based on 4 parameters. The parameters are fields in the excel data source.
What is the best way to do it?

If I include the ZeroOut process in the prolog then I can't retrive the parameter values ?


regards,

Abi

What I do in the scenario where the client wants to load only certain versions, or months, or whatever, and those are defined in the dimensionality, and they don't want to have to answer a parameter dialog box is split it into two processes. The first process reads the source and builds subsets for each of the affected dimensions based on what is found in the data. The second process builds the view, based on the previously built subsets, does a ViewZeroOut, and then loads the data. The only caveat here is that the source data has to be complete. What this means is that if you erase a portion of the cube, based on what is found in the source, then you need to make sure the source has a compete replacement of what you erased. This is a business issue, not a technical one, but needs to be understood and handled properly.

[Abinaya]

Hi
I want to ZeroOut a cube based on 4 parameters. The parameters are fields in the excel data source.
What is the best way to do it?

If I include the ZeroOut process in the prolog then I can't retrive the parameter values ?


regards,

Abi

What I do in the scenario where the client wants to load only certain versions, or months, or whatever, and those are defined in the dimensionality, and they don't want to have to answer a parameter dialog box is split it into two processes. The first process reads the source and builds subsets for each of the affected dimensions based on what is found in the data. The second process builds the view, based on the previously built subsets, does a ViewZeroOut, and then loads the data. The only caveat here is that the source data has to be complete. What this means is that if you erase a portion of the cube, based on what is found in the source, then you need to make sure the source has a compete replacement of what you erased. This is a business issue, not a technical one, but needs to be understood and handled properly.

Thanks.I did like what you mentioned here. I have few problems.

When I built a subset for the one of the dimension I am having duplicated elements? How do I create subset without duplicated elements?

Where should I call the second process? Epilog of the first process passing those 4 variables?

regards,

Abi

[lotsaram]

Usually you build a temporary dimension (when using dimensionelementinsert you can't add an element to a dimension more than once). Then on the epilogue you loop through the temp dims to build subsets in the original dims and lastly delete the temp dims.

[Abinaya]

Usually you build a temporary dimension (when using dimensionelementinsert you can't add an element to a dimension more than once). Then on the epilogue you loop through the temp dims to build subsets in the original dims and lastly delete the temp dims.

Thanks. I will try that.