COGNOiSe.com - The IBM Cognos Community

Legacy Business Intelligence => COGNOS PowerPlay => Topic started by: cognostechie on 05 Aug 2005 01:20:41 PM

Title: Powerplay Transformer Book/Training material
Post by: cognostechie on 05 Aug 2005 01:20:41 PM
Hi Guys:

I lost my book that they give at the Cognos training. Is there anything that I can download or anybody can suggest anywhere where I can get that training material. Only for the Transformer. I need to look at some advanced options like what's the difference between 'Precision' - 0 to 9. I know 9 rolls up the data most accurately but what are the diff. roll-up methods for 0 to 9.

Also, the linking of diff. files/tables in the data source based on their keys.

Appreciate your help in advance. 
Title: Re: Powerplay Transformer Book/Training material
Post by: Darek on 05 Aug 2005 03:55:47 PM
Precision is available only if a measure is using floating point storage, in 8 bytes (64 bits). So every number is stored in a scientific form as in 13.1415E0. This means that if you prefer more precision, obviously less digits would be available to the left of the decimal point. Ususally 3 is enough.

Linking is done through source fileds. Each category in a cube contains a source, a category, a label and a description field (plus some others like sort).
Title: Re: Powerplay Transformer Book/Training material
Post by: cognostechie on 05 Aug 2005 05:04:46 PM
Darek, thanks for the answer. Maybe I am not showing any decimals in the cube as they only want to see the dollars, that's why the cube is showing 29 million dollar in sales even when the precision is 9.

For the linking..I know each category contains a source, label etc (in the properties) but I was looking for a document so that I can see how to link 3/4 diff. tables that appear in my data source when they are from different databases and flat files. Anyhow, I figured that one.

Title: Re: Powerplay Transformer Book/Training material
Post by: Darek on 05 Aug 2005 05:23:20 PM
I wasn't probably too specific. What I meant to say is that as long as a source field in one query match another field in another query by name, and they are designated as the source column for a level, those queries will be joined.

Glad that you've figired it out.