I have a datacube where each Associate has 5 strengths, so their name will be repeated 5 times. I am building a chart with the percentage of Associates who have each strength, so I need the percentage over the total number of Associates, not Associate strengths (which would be 5 times the total number of Associates.
It would replicate this, and the calculation is just a distinct count of Associates using a table calculation set to Percent of Total computed using Table (down) in Tableau:
I am unable to replicate this in Dundas. PERCENTTOTAL and PERCENTTOTALGROUP both seem to use the total number of rows, not Associates. I thought perhaps I could trick it into getting the total for the whole dataset by creating a field that is the same for everyone and using SUM() to group the distinct Associate count by that, but no luck. I am able to get close by manually dividing the number by 5, but as some people mis-entered their strengths, it is slightly off, and I’d prefer if I could get the true number of Associates rather than a number calculated based on how many strengths they entered.
How can I do this?