I used left joins with for instance fields like gemeente and jaar in my views in BigQuery. I am thinking about using a generated ID that we then can use in PowerBI.
My BI / DataScience at Bank team are using multiple primary / foreign fields in relationships between tables/views, but an other team is using Generated Hash ID's (based on multiple primary keys). My team decided not using that in Databricks, because of performance issues.
I used left joins with for instance fields like gemeente and jaar in my views in BigQuery. I am thinking about using a generated ID that we then can use in PowerBI.
My BI / DataScience at Bank team are using multiple primary / foreign fields in relationships between tables/views, but an other team is using Generated Hash ID's (based on multiple primary keys). My team decided not using that in Databricks, because of performance issues.
https://showmethedata.blog/generating-unique-keys-in-bigquery