Notice the the dim is on the one side and the Many is on the fact side. Lets have a think about how 1 to many happens using the date dimension as a good example But, how would you change this into a snowflake Schema? Fact table 1 to many At this point it still feels like Snowflake is still a good option. The snowflake seems to be doing everything that it needs to do. There are 12 Agenda ItemsДЄnd again, the new measure shows us correctly the number of agenda items. I created a Measure in the fact table that does a distinct count of the Event Detail IDs # Events = DISTINCTCOUNT('Dim Event Details')ДЄgain, this works a treat. So the next question the user wants is, how many agenda items are there per event? It Sums to the 2 event (Which is correct for the number of events but is not correct for the number of Agenda items) However the metric looks a little confusing if dont understand the dataДЇor example, the table with EventDetail and Event should sum to 12 (the 12 Agenda items) but it doesnвЂ™t. When you bring in data from the Snowflake dimension, it still shows the 2 event metric. This is an interesting look at the results. Seperating the fact and dim when the relationship is a 1 to 1 is knows as a If you however leave the model as one to one, both direction, it makes no difference to the results.

I have changed it to 1 to many to get the single direction filter and the fact that the event table snowflaked off from events creates 1 to many. Moving the facts into a Fact table and the descriptions into a dim table in a 1 to 1 relationship is called creating a Degenerate table the relationships started as both direction. Lets see how it would work in Power BIДЇact to Dim is one to one. Its very easy to create a snowflake schema for this.

We are looking at Events and these events consist of Agenda items. This is very similar to something I am working on at the moment. Also, how can I change my thinking on Snowflake to STAR So i wanted to have a look at this and see why I do it and what it actually means for Power BI. I however, do seem to use snowflake schemas and it seems that even these arenвЂ™t recommended for Power BI. Power BI prefers the STAR Schema model comprising of a fact table and Dimensions providing all the filters for the metrics We have already spoken about Steering clear from flat files. One was from the session related to DAX and Power BI Models. I attended the Power BI conference from and got so many great ideas on the 16th June

Snowflake bi tools