I created a matrix with client names and colums of different policy types. the measure in the matrix, is the count of distinct plicy numbers. I want to be able to only filter the # of clients who had nonblanks in atleast 2 of my column categories ( the type of policy), in other words they have more than one policy type.
I cant because the policy categories always has a value, i want to be able to count for each row, if there is more than 1 policy type. which means i need to filter my policy categories (columns) to show me if there is atleast 2 non blanks.
So it looks like in your example data, each person has 1 policy number, but can have different types of policy under that same policy number? What happens when you use COUNT of policy number instead of DISTINCTCOUNT?
I would use a count of the category column insetad of distinct policy numbers, so that the total comes out correct. You may have to use a measure like this to create a visual level filter or slicer:
NumberOfCategories = COUNT(Policies[Type])
and then filter on that where it's greater or equal to 2.
Could you share how you have your table(s) structured? It would help us give you a better answer