Good day! My Item table is related to the Item_Country table through a column: Item_Country_ID When counting Item (Item_ID) for unique instances of Item_Country_ID, I notice that empty values are not uniquely shown as 1 value = "(empty value)". Can you help me understand this? Thanks!

Hi Hans, could you provide a screenshot of your logical data model (only the relevant part would suffice).

Hi Boris. Is this sufficient, Item is quite wide :-)

I assume that the item Id is the primary key of item dataset, right?

That's correct.

This should work as you expect as long as the primary keys are set correctly and there are no referential integrity issues with data.

Hi Hans, most probably, those "empty value" values are not representing the same value, but each line there represents different one. The reason, why it is displayed like this would be a referential integrity issue as Boris mentioned - basicaly you probably have some values of "country ID" in your "Item" dataset which DO NOT exist in your "Item Country" dataset. If this is the case, I would highly recommend to fix it in the data (eithwr load all Countries into the Item Country dataset or update Items dataset to only cobtain valid countries). If not fixed, it can cause other issues with calculations, filtering etc. If you are using the hosted GoodData platform, you can actually check if some values are missing by running the workspace validation as described here: https://help.gooddata.com/doc/enterprise/en/workspace-and-user-administration/administrating-workspaces-and-workspace-objects/validate-a-workspace