Hi Florian,

Have you tried using aggregation block? It allows you to combine list and measure (data point):

First, as a general concept, the PQL and MDX languages are very flexible in achieving almost any logical formulation needed. Sometimes, the graphical tools don't facilitate this flexibility so you need to use the scripting mode - which is not bound by the UI's limitations. Other times the path to resolving your analytical problem is not direct, so you need to do a few extra steps. And then, unfortunately, there are rare instances where something is impossible. We like to monitor all three headaches to see if we can make it simpler, easier or possible.

Now on to your question: The 'data point' is a logical element that can only ('legally') refer to a single item from any dimension at a time. To achieve the multiple items outcome you are after, you need to build a set of those items and aggregate them into a custom member (or formula)  - what Uros is suggesting. Then, select this single custom member in your final data point  - in the second calculation - to achieve the outcome you are after. Its indirect, but achievable.

To further enhance your calculation, consider building a parameter of the items you want. Then build the customer member aggregate using the parameter instead. This will allow you to dynamically change the aggregated element in the final data point from the report on-the-fly.