Join us at FabCon Atlanta from March 16 - 20, 2026, for the ultimate Fabric, Power BI, AI and SQL community-led event. Save $200 with code FABCOMM.
Register now!To celebrate FabCon Vienna, we are offering 50% off select exams. Ends October 3rd. Request your discount now.
I've read entries like this about sort-order changing when you merge two tables and then expand one, but that's not what's happening here. I have two tables; one contains company subsidiaries ('Assignee Cleanup [Ext]'), and another (shown below) with patent records with owners (or 'assignees'). I'm merging them together, with the understanding that sometimes the owner won't be on the other list and I'll just have null entries. Here's the table at the step just after merging queries on the 'Merged Assignee' field, which has the company names:
I now expand the 'Assignee Cleanup' column to give me only the parent company name. In this case, none match so this gives me a column of nulls. This is totally fine and expected. However, note the 'FullNum' column:
The order has changed, and row 4 is now row 1. That's totally fine. What's not fine is that the FullNum value has totally changed. I can't work out where this new number comes from. Does anyone know if this is a bug? I feel that conceptually expanding a column on a table should absolutely not be able to change existing data within the table.
Edit: I should also add, I've been using this method without any issues as part of a template for years. To my knowledge this has only started in the last few days.
You are saying that the data value in the first column CHANGES when you do a MERGE JOIN. I serioiusly doubt that. If so, can you share your model (removing any PII)?
Have you tried looking for that original record in that column?
What happend if you FILTER for that one single record BEFORE you do the MERGE JOIN and Expand Columns?
I bet you see that it has not changed.
Proud to be a Super User! | |
I can try to share, but it might take a while as it's part of a much larger process and I'll need to extract it. In terms of "looking for that original record in that column": the data I've shared is the entire table, not an excerpt. I already filtered in a prior step to just those five records as part of trying to find the issue.
For what it's worth though, I've just added a filter for that single number, including the formula bar so you can see the step applied. Here it is pre-merge:
and post merge. record's gone. You can see in the formula bar all this step does is expand out the columns in the merged-in table!
What type of JOIN is it? LEFT? INNER? Is that killing that row? I still maintain that it is strictly data related. I know Power BI does wierd stuff sometimes, but actually CHANGING DATA is not one of them.
Proud to be a Super User! | |
I am experiencing the same problem...after the expanding step, the columns on my first table change. I also chose Leftt Outer Joint
Left-Outer. On a whim, I tried sticking a Table.Buffer step in between the merge+expand and... for whatever reason it seems it might have fixed the problem. Currently waiting for the full dataset to update to confirm... Really confused as to what's happened there. As far as I know the replacement number doesn't appear anywhere in the input data.
Thanks for sharing the solution. This resolved my problem!