Forum Discussion
XM_TYPE_STRING error?
My norm of "complex" is very much different from other's norm of "complex". I'd say it is easy peasy. It's just a minx which uses a VAR of a minx and the main minx uses 3 variables. Barely anything. I have expressions that are 100 lines long on large switch statement, which I would call mediocre. I also have same amount of lines in a measure this time that does a lot with dynamic tables. Which I would call complex. At least, that's my norm. But those are not in play.
However, I saw one of the comments in your thread propose a desync between the column it is versus what the code chooses. I tried initially through applying a filter ( 'table , 1 = 1 ), which didn't help. Nor adding/ removing a comment.
What did solve my problem was assign a different table reference and then come back to it.
Origin : Calculatetable ( 'table 1' )
Step1 : Calculatetable ( 'table 2' )
Step2 : Calculatetable ( 'table 1' )
Done
I'm just afraid now my visuals are **bleep**ed and are something I have to redo. I have dozens of pages in this model and reports referencing the semantic model. Just like if you first build your visuals on a data field, then create a separate date table with a link onto that field. At least the relationships broke, so I have to redo those.
But it solves it, thanks for referencing the thread!
Hi DouweMeer
I found an old thread that maybe you have already read. Does the "commenting out the lines one at a time and then uncommenting them back" practice works for you?
DAX Error with new table (urgent !) - Microsoft Fabric Community
If the solutions in above link don't work, is your DAX for the calculated table is very complex? You may try first implementing part of it to get an intermediate result, then get the final result from the intermediate result. Perhaps the engine in the back end is not able to understand and process all at a time.
Best Regards,
Jing
If this post helps, please Accept it as Solution to help other members find it. Appreciate your Kudos!
My norm of "complex" is very much different from other's norm of "complex". I'd say it is easy peasy. It's just a minx which uses a VAR of a minx and the main minx uses 3 variables. Barely anything. I have expressions that are 100 lines long on large switch statement, which I would call mediocre. I also have same amount of lines in a measure this time that does a lot with dynamic tables. Which I would call complex. At least, that's my norm. But those are not in play.
However, I saw one of the comments in your thread propose a desync between the column it is versus what the code chooses. I tried initially through applying a filter ( 'table , 1 = 1 ), which didn't help. Nor adding/ removing a comment.
What did solve my problem was assign a different table reference and then come back to it.
Origin : Calculatetable ( 'table 1' )
Step1 : Calculatetable ( 'table 2' )
Step2 : Calculatetable ( 'table 1' )
Done
I'm just afraid now my visuals are **bleep**ed and are something I have to redo. I have dozens of pages in this model and reports referencing the semantic model. Just like if you first build your visuals on a data field, then create a separate date table with a link onto that field. At least the relationships broke, so I have to redo those.
But it solves it, thanks for referencing the thread!