Forum Discussion
Post ProdigyHow to display zero using Measure or column value (when there is no value)?
I would prefer you don't do that. Instead, enable "Show items with no data".
Super UserYou need to do a lot of work on your data model before you can think of that
You need to create a Facilities dimension table and then merge it with the trend and occupancy tables. Then you need another PayerType dimension , and then youcan connect your dimensions to your fact without the need for a joinkey table.
Post ProdigyCan you please take a look at my updated PBI file here that has two additional dim tables (tblFacility & tblPayerType)? I deleted all relationships from my previous "JoinKey" table.
I am not sure how to connect tblFacility with 3 fact tables.
Actually, I am lost in many ways 🙂
Also, I am not sure why on the "Actual" tab, it shows fewer columns (4 instead of 6) than others.
Appreciated.
- lbendlin
Merge these tables
Bidirectional 1:1 between two tables usually indicates that this should be a single table.
This is not yet a dimension table. It is missing a primary key. Figure out what the primary key is and kick out/downstream anything that prevents it from working.
Avoid bidirectional *:* relationships wherever possible. They are expensive and can lead to wrong results.
- JustinDoh1
Ok. Thank you so much for your feedback.
I will work on it again tomorrow and let you know 🙂
- JustinDoh1
I have been modifying relationships between dim table (especially "zzvw_occupancy_trend") with three fact tables for last day or so.
Here is my PBI file.
I am still having a difficulty with using a Filter (Facility Name).
I am not sure whether I need to bring value from dim table or Fact table to be used for filter.
Can you please take a look at the relationship that I created first and give me any feedback where I am wrong?
Appreciated for your help again.
- lbendlin
This looks great. Filters should in the vast majority of scenarios come from the dimension tables. That filter choice then flows down into the fact tables. If you filter on a fact column that choice can not (and should not) propagate back up into the dimension.