If you are unable to create a new account, please email support@bspsoftware.com

 

News:

MetaManager - Administrative Tools for IBM Cognos
Pricing starting at $2,100
Download Now    Learn More

Main Menu

Error using same dimension on both axes of a crosstab

Started by casey.smith, 18 Jun 2007 05:07:26 PM

Previous topic - Next topic

casey.smith

The documentation states that you cannot use levels from multiple hierarchies of the same dimension in a report, but does not say whether the same limitation should apply to multiple levels of the same hierarchy. I am experiencing errors in a few scenarios, and I hope I can determine whether they are flaws in my model or intrinsic limitations that I must work around.

First of all, the error manifests at report runtime as "GEN-ERR-0001    Expected query result 'R0h0_1' is not found in the response of 'OlapQueryProvider' ".

Consider a scenario not unlike the following: I want a table of differences between sales in each month throughout the year, where by each axis is shows the span of months and each intersection shows the difference in sales between those 2 months. In this hypothetical scenario, it doesn't matter what fact(s) I try to put into the query: having the same dimension on both axes (one aliased of course) throws an error.

The same problem occurs when using different levels of the same hierarchy. Non-unique levels in a hierarchy can produce useful reports when cross-tabulated, like, say, sales broken down by month and day of the week. But this fails as well.

The strangest thing about all of this, is that the exact same queries seem to work fine in list format, when pivoting a list to crosstab in Query Studio, or when certain summaries are added in report studio. But about 90% of the permutations I have tried in Report Studio, the crosstab fails to run properly.

For reference, I'm using Cognos BI version 8.1.209, with a DMR model over a SQL Server 2000 database. If anyone has experienced something like this and knows whether it is a known issue with Cognos, or a product of something wrong in the model (i.e. uniqueness of dimension levels, relationship scope, determinant) please let me know.