0

I work for an OTC company.

Here is the background on the workflow. Taker (Buyer) --> Requests a quote on derivatives --> Maker(Quoter) responds with quotes (bid/ask) ---> Taker either buys/sells ( traded) or takes no action on the quote. The quote can be sent and received from multiple makers.

Let me know if my star schema is correct.

I am thinking of takers, makers as dimension tables. Requests as one fact table and quotes as another fact table. Is this approach correct?

Vinay Emmaadii
  • 125
  • 1
  • 11
  • 1
    A dimensional model design is based on your reporting requirements, not on the structure/workflow of a source system - so without knowing what your reporting requirements are it’s impossible to comment on your design. In any case, just asking if a star schema is correct is too broad a question for this forum: you need to ask a specific, focussed question – NickW Jul 21 '21 at 07:56
  • @NickW Typical reporting questions would be 1) total dollar amount by different derivates by day/month by exchange 2) total number of quotes quoted by makers in last x days and break down by traded/ cancelled/ expired (no action) – Vinay Emmaadii Jul 21 '21 at 21:37

1 Answers1

0

So you’ve started to identify your measures, once you’ve identified them all the next steps are:

  1. Define the grain if each measure; measures with the same grain can exist in the sane fact table; measures with different grains need to be in different fact tables
  2. For each fact table, identify the entities you need to filter or aggregate the facts; this gives you your dimensions
NickW
  • 8,430
  • 2
  • 6
  • 19