Forum Discussion
Data Modeling: Connecting to a Dimension Table Twice
Hello,
I have a question about data modeling in Power BI. This is a simplified example of the data but it will suffice:
I have a main shipment fact table which includes keys to origin ports and destination ports:
| BookingID | OriginPortID | DestinationPortID |
| 111 | 1 | 3 |
| 112 | 2 | 1 |
| 113 | 3 | 2 |
I also have a Locations table which contains information about the ports that both of these keys link to:
| LocationID | Locations |
| 1 | Shanghai |
| 2 | Antwerp |
| 3 | Rotterdam |
These tables contain far more columns and information which is used, however, my question is what is the best way to model this?
Option 1:
In power query: merge queries to pull needed data (Port Name, Region, Country, etc.) into the main bookings table for both the origin and destination. Don't load the location table.
Option 2:
Load the location table twice. Create a relationships from the booking table to each copy of the location table.
Option 3:
Load the location table once. Use lookup to add DAX columns to my main bookings table.
Option 4:
??
Does anyone have feedback on the best way to handle situations like this? This isn't the only instance of this type of relationship in my data so I'll need to apply the solution elsewhere as well.
sem1 you will be able to slice on original port (assuming it is active relationship) but if you also want to slice on destination port, in that case recommendation would be to have two table one for origin and one for destination and then set relatnship with respective table with shipment table.
Above will be more flexibile and will meet most/all the requirements.
3 Replies
sem1 the best option is option 4, set relationship between location table and shipment table on both origin and destination port id. One relation will be active and other inactive coz there can be only one active relationship between tables. In mesure use USERELATIONSHIP function for inactive relationship and you will able to achieve desired results.
Would that enable me to use both dimensions in a chart? Or would I only be able to use one.
I.e.:OriginPort DestinationPort Shipments Shanghai Rotterdam 10 Antwerp Shanghai 5 Rotterdam Antwerp 13 I'd also like to be able to use slicers/filters to filter them. So for example if I look at the origin of Shanghai, I want to see the destinations I'm shipping to.
sem1 you will be able to slice on original port (assuming it is active relationship) but if you also want to slice on destination port, in that case recommendation would be to have two table one for origin and one for destination and then set relatnship with respective table with shipment table.
Above will be more flexibile and will meet most/all the requirements.
