Forum Discussion
Star Schema Advice
I am currently making a star schema as pictured below:
I want to add a level of detail to each of the dimension tables, which will be a one : many from the detail table to the dimension table. I understand I can merge these detail and dimension tables in power query, but this will result in a larger dataset than just using relationships.
The new schema below looks like this:
Is this against best practice? It functions as expected but wanted to check before taking the report forward further!
Thanks,
Leo
Anonymous
With the mods you now have a snowflake schema structure, which generally is ok (depending as always on the depth/size of the dataset). You can keep a star schema structure (which appears to be the favoured structure in guru land) and still filter other dimensions using measures in the filter pane.
3 Replies
Anonymous
With the mods you now have a snowflake schema structure, which generally is ok (depending as always on the depth/size of the dataset). You can keep a star schema structure (which appears to be the favoured structure in guru land) and still filter other dimensions using measures in the filter pane.
Anonymous , As long Details Dim and Dim is 1-M they should/can be merged. For Week you have check why not join date with Fact. and merge week and date
amitchandak I would do that, but I have to pivot the detailed tables - which when merged go from 1000 rows to 200,000+ !
Is this type of report flawed, or acceptable under these circumstances?
