Very interesting approach. Is this a hack?! Or considered best practice? I am curious learn more about these hidden Tableau tricks.
Thank you for all your help.
Would you rather use this approach OR change the dataset in excel and arrange the data blending connections?
1 of 1 people found this helpful
That's really depends on the whole picture of your data connections.
You connected to several data and it's difficult to judge immediately.
But talking about Union which I use, the risk is it duplicates all the data set, so without specifying the "table name" all the measure become double.
Th image is all the line item becomes two and recognized uniqueness with "table name"
Other option is to pivot the data. Obviously has implications if you're blending this with other data sets.
Pivoted 4 measure fields, then stripped the pivoted names apart with calculation. See attached.
2 ref lines.twbx 4.6 MB