I'm sorry you missed my point. The question has sat unanswered for months even though two different people asked it two different times before me. Both of those prior posts appear to have become lost and buried.
I guess you missed the original duplicate post.
So I'm bringing the question back to the front of the line on behalf of myself and two other people and consolidating the open and unanswered variants of the question here.
Any chance you're able to help solve the question that's been presented by 3 fellow users?
It looks like one possible cause of this is when a set is being created from the group of a field. The solution was to use a field that is not using a group or Combine Fields to create the set.
Hope that helps at least a little bit!
Thanks Tracy! That actually might be what's happening. I may have done that, or more likely I created a calculated dimension on a text dimension to group fields automatically instead of using a manual grouping. Will have to dig through the workbook to see if I can deconstruct it to see what I did.
Let me me make sure I understand the caveat correctly. If I do I'll post your reply it in the other two open threads w/ a cross-ref here to close the loop on all three threads.
Do I understand correctly that
1. we should not use a Grouped dimension to create a Set. If we do that can mess things up under Tableau's hood.
2. And instead of using a Grouped Dimension, either
2.1 use a dimension that isn't grouped; or
2.2 if this applies create a combined dimension and use that to create a set.
And 2 last questions,
1. is there any other cause of this error message of which you're aware?
2. is there any way you're aware of to correct the error other than duplicating the data source and then switch to it in the workbook? That's the only solution mentioned in one of the two other threads.
I did NOT miss the point. Once a post is responded to it floats to the top of the forum where it is no longer burried. My response of standard forum etiquette still stands, it's a double-posting, pure and simple. You can still make your point in the original thread.
I don't want to make a general statement that a grouped dimension should NEVER be used in a set, however, that has been a cause of the error in the past and could be a good rule of thumb.
There were no other causes of this error message that I could find. Besides recreating the extract itself, the supplied workaround is the only workaround I am aware of.
Thanks Tracy. BTW, is there a way for me to unbranch this branch. I was trying to put the parts of the thread about posting practices here into a separate branch to keep things on-topic.
Unfortunately, there's no way to undo a branch.