Forum Discussion
Error: this could be because there is missing intermediate data to connect the two columns
One of the columns participating in relationship must have unique values and no nulls/blanks. You can create the lookup table:
1. Go to the "Edit Queries", duplicate one of your tables, rename it to "MyLookup" and delete all column except key/id column.
2 Remove the duplicate values and filter out the nulls...
3. Create 2 relationships - from each original table to the new lookup table.
This answer seems to depend on each table having a single unique key. My data is related by a set of columns (Segment, Product, Month). How can I implement this relationship?
varakcheev Thanks for the filtering tip. I had created an Excel file filled with dates. For no reason, it had *nulls* in it, once imported in PBI, and creating a relationship resulted in the error mentioned in this thread's title.
After filtering the nulls, no more error.
Today I've got the same error:
I've got a table with hundred of customer names without a customer id and a fact table with the same name.
I've grouped the dimension table by the name and i got the error when I created the relationship.
The reason for the issue was, that the customer names: e.g. Test Company and test company
Power Query is case sensitive and Power Pivot is not. So I make the values in Power Query to uppercase (in fact and dim table). So it works.
Never use text values for key lookups ;-)
