Hi @oshello ,
The answer depends on several factors, such as the size of your data, the complexity of your calculations, and the mode of your data source (import or direct query).
In general, calculated columns are not recommended for large datasets or complex expressions, because they are pre-calculated and stored in memory when you refresh the model. This can increase the size of your model and affect the performance of queries and refreshes. Also, calculated columns cannot be used in slicers or filters.
A better alternative is to use measures, which are dynamic and only calculated when needed. Measures can also be used in slicers and filters. However, measures require a good understanding of DAX and how it works with the filter context.
If you need to create columns based on other columns in the same table, you can also use Power Query to create custom columns before importing the data into the model. Power Query can perform transformations faster and more efficiently than DAX, and it can also push some calculations to the data source if possible. However, Power Query custom columns require a full refresh of the table whenever you change them.
So, to summarize, here are some guidelines to help you choose between calculated columns, measures, and custom columns:
- Use calculated columns when you need to create a column that depends on values from other tables or that is used as a key in a relationship.
- Use measures when you need to create a calculation that depends on the filter context or that is used in visuals, slicers, or filters.
- Use custom columns when you need to create a column that depends on other columns in the same table or that can be pushed to the data source.
Best Regards,
Neeko Tang
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
