Technical data

Chapter 3 Understanding Analytic Models

Dimensions are separate objects that can be used independently of data cubes. For this reason, even though

you originally created the MONTHS and PRODUCTS dimensions for the SALES cube, you can reuse

these dimensions with other data cubes. For example, imagine taking a new sheet of paper and writing

COST_OF_GOODS at the top of the page, and then attaching the existing dimensions to the new sheet. This

is an example of this image:

SALES

2004/01 2004/02 2004/03

010

020

070

COST_OF_GOODS

2004/01 2004/02 2004/03

010

020

070

Attaching the PRO DUCT_CODES and MONTHS dimensions to the COST_O F _GOODS data cube

It is temp

ting to think that the dimensions attached to the COST_OF_GOODS cube are copies of the

dimensio

ns on the SALES cube. This might be true on paper, but an analytic model works differently. The

dimensi

ons are independent objects that you attach to the data cubes. For this reason, the MONTHS dimension

that is a

ttached to the COST_OF_GOODS cube is the same dimension as the MONTHS dimension t hat is

attache

d to the SALES cube. Therefore, any change that an application makes in a dimension is reflected on

all data

cubes that use that dimension.

For example, suppose the application adds a product code called 090 to the PRODUCT_CODES dimension on

the SALES cube. The analytic model adds 090 to the PR ODUCTS dimension on the COST_OF_GOODS

cube. This is an example of this image: