[Minor bug] Unpivot voluntary fields population

I have a situation then input table to unpivot has different number of fields. Unpivot node automatically adds extra fields to the value filter. It is destroying the logic and making solution unstable.
Is it possible to add extra fields to retained columns?

Hi @izaychik63 -

Can you provide a toy example of what you mean? I tried playing with the Unpivot node, but I guess I’m not approximating your input data correctly, because I’m having a hard time identifying the problem. Thanks!

Education Unpivot.knwf (17.9 KB)
Scot, I attached WF with example. To see the issue, look inside unpivot value section
then connect second data set and open unpivot again. The value section will be filled with columns not presented in the first case.
The same effect I’ve got then just change source file in excel reader.

OK - I see what you mean. Two potential solutions:

  1. Check the Enforce Inclusion radio button in the Include box for Value columns. Then the columns stay the same regardless of which Excel Reader node is connected.

  2. In this particular case, you could also fix the issue by switching to Wildcard/Regex Selection with a pattern of *Completion Date, but that solution doesn’t generalize well.

Does that help?

Thank you for recommendation. I hope it could be addressed in a future release as module supposed to do what is on a spec and not supposed to do what is not on a spec (my case).