Hi,
I wonder what is the best practice to create data for OLAP processing
to achieve comparision of planned sale to sales data.
Sales data are in my facts table and target data are in another table
(data are imported from excel spreedsheet). At the moment I have
created another pchisical table and I do data loading (daily) of sales
data with additional field "Sale" and planned data with additional
field "Planned", it works for me fine but I wish to know how it should
be done in professional way.

regards

This really depends of your OLAP system:
If you are using Microsoft SQL Server 2000 Enterprise Edition. I
would prefer to load data into two exact maching fact tables (or OLAP
partitions), one for sales data and another for the budgets. in SQL
server 2005 I would do it in a similar way but here the layout of the
fact tables can be different, for example - with less dimensions used
for the more simplified budget data. In SQL server 2000 Standard
Edition the only option is to load all data into a single fact table as
there is no partition option here.

You need of course to add a new Dimension to your solution - "scenario"
to differ the sales data from the budget data where the actuals have a
member "Actual" and "Budget" for the budgets in order to not mix up the
data.

Using other OLAP products it is a bit diffecult to give a good advise
as the best solution depends of the architecture behind the OLAP
product used. The most common method is the one described as on SQL
Server Standard Edition, as many OLAP products does not have any
advanced partition options.

RE
Bjørn T




Piotr skrev:

Thank you very much thats what I was really wanted to know