Microsoft Fabric Community Conference 2025, March 31 - April 2, Las Vegas, Nevada. Use code FABINSIDER for a $400 discount.
Register nowGet inspired! Check out the entries from the Power BI DataViz World Championships preliminary rounds and give kudos to your favorites. View the vizzies.
Hello Community,
I have a table like this:
Id, Key, Value
In Power Query I pivot the key and value pairs to create columns from them.
My issue is, every time a change happens in the table, the pivoted table fails in the service. I can then open up the .pbix file, refresh it locally without issue, re-upload and it can refresh in the service again until this table is changed.
If there was an error that error should also happen locally, so not sure how i can avoid this or what to look for.
Solved! Go to Solution.
The M code doesn't have hard coded coulmn names
That's pretty much your problem. The output of a M query (what gets loaded into a Vertipaq partition) is expected to have exactly that, hardcoded column names.
Why?
Because that's how Power Query is designed. It does not tolerate dynamic query meta data changes.
All right thank you for explaining this
I would like to understand this limitation more in depth but am unsure where I can read more about it. I tried searching the words you used but could not find a match in the documentation
Don't pivot. Load the data as is into Power BI and let the visuals do the pivoting for you.
Not an option in my case.
Then you will have to live with these issues. Every time you change the Power Query Meta Data that will require a re-publish.
What metadata?
The M code doesn't have hard coded coulmn names
Is this what you are referring to when you say metadata?
https://bengribaudo.com/blog/2021/03/17/5523/power-query-m-primer-part20-metadata
Is it not possible AT ALL to use pivot operations and push to service?
The M code doesn't have hard coded coulmn names
That's pretty much your problem. The output of a M query (what gets loaded into a Vertipaq partition) is expected to have exactly that, hardcoded column names.
all right thanks
still lots to learn it seems
User | Count |
---|---|
54 | |
31 | |
31 | |
30 | |
24 |
User | Count |
---|---|
46 | |
43 | |
16 | |
16 | |
14 |