I have been reading about Star schema lately and its importance. I want to know if its really important to have it implemented everytime you develop a BI App? I have situation where I want to implement it but the problem is I have plenty of different types of facts e.g. violations, fundings, budgets, spendings, projects, events, visits etc, some of them have amounts, some of have other KPI's. I cannot convert them into single fact table. I can though try to combine a few of them into one but not all of them can be combined into a single fact table. How can I convert such a data set into star schema?
So far I have been following relational model by bringing in different data tables and removing their redundancies(if any) and keeping them connected with single key.
Usually the solution for multiple facts is concatenation in a fact table, using a field to identify the fact and use it in set analysis to create the expressions, also trying to reuse field names to not make it too wider.
Thanks for your reply. I am, not so sure myself either. Google comes out mostly with one example of Sales where it fits in the best but not every data set is related to sales. I am, not so sure how to implement a star on something where data is coming from different modules like sales, inventory, accounts, hr, crm .. Is an integrated star schema possible in that case or even snowflake?
Usually the solution for multiple facts is concatenation in a fact table, using a field to identify the fact and use it in set analysis to create the expressions, also trying to reuse field names to not make it too wider.
