Sample FoodMart Fact Table

[durkine]

Hi,

Would someone be able to explain how the fact tables eg. sales_fact_1997 in the sample Foodmart schema was put together.  Is this table just a combination of primary keys and corresponding sales figures that is already contained in the base tables eg. product, customer_sales, customer ?

Thanks,

 

 

[akrause]

Hi,

fact tables are the core of datamarts and contain all valuable data to fill an analysis cube.

In the first step this data is collected via ETL in datawarehouses from different datasources and put together in a second step from here to the datamart's fact and dimension tables (of course you can fill your datamarts directly too depending on what you are planning to do, data quantity, and so on) .

The dimension tables data and fact tables data are used to build the cube for your OLAP analysis.

Hope this helped a bit....

[durkine]

Hi,

Thanks for the information.   Can I check then, does jasper analysis require a data warehouse and these fact tables already created, or will jasper analysis allow me to carry out data analysis just on the base tables I have. 

I guess you can probably tell I don't have a data warehouse / star schema setup etc..

Regards,

 

[akrause]

If your base tables does contain all the information needed to fill the data cube then you don't have to create a datamart necessarily.

 

But operational databases are optimized for performance and so the database tables aren't usually suitable for OLAP analysis and so the data has to be extracted, transformed and loaded into datawarehouses or datamarts.

 

And it's a performance issue here too: access to datamarts are usually more performant than accessing the operational databases directly, because their structure is optimized for analysis tasks and the operational database is balanced this way also.

 

Cheers...