Re: Star Schema from 'Flattened' CSV
From: Aaron Peters <transreductionist_at_gmail.com>
Date: Thu, 2 Nov 2017 16:36:45 -0700 (PDT)
Message-ID: <d6e17736-2705-4c85-967f-797b46c7c089_at_googlegroups.com>
Date: Thu, 2 Nov 2017 16:36:45 -0700 (PDT)
Message-ID: <d6e17736-2705-4c85-967f-797b46c7c089_at_googlegroups.com>
Not sure what you are suggesting. Is this a correct assessment? I can go one of 2 ways:
- Use 2 fact tables with 1 dimension: One fact table for vendor 1 One fact table for vendor 2
fact_performance1
- unit_price
- tot_sold
- tot_ret
- ven1_price
fact_performance2
- unit_price
- tot_sold
- tot_ret
- ven2_price 2. Use 1 fact table with 2 dimensions A dimension for vendor 1 A dimension for vendor 2
fact_performance
- unit_price
- tot_sold
- tot_ret
- ven1_price
- ven2_price
dim_ven1
- ven_name
dim_ven2
- ven_name