This article describes the use of Expression.Evaluate as a very helpful “swiss-army-knife”-method for your Power BI toolbox as it has many more use cases than the one described below. It lets you perform repeating tasks without using functions and can even replace recursive operations in some cases – but that’s a topic for a later blogpost. Today we start with a simple task:
If you want to filter rows only where none of the fields/columns are empty, you’d either pass not-null filters into every single column (which is tedious and not dynamic), transpose or add an index column, unpivot and filter from there (both might drag performance down on large tables).
Using Expression.Evaluate will create a dynamic approach that doesn’t force you to transform your source table nor to directly specify your column names: It will take in a text string and execute it as if you’ve been written it manually into the formula. If you manage to let this textstring be created dynamically according to the different tables you’re going to pass in – Expression.Evaluate will make sure that these individual statements will be executed accordingly.
So for a table containing Column “A” and Column “B” the filter statement would look like this:
Read more
