Partitioning by Table Calculations

Wednesday, July 20th 2016. | Others

Here’s an issue on the Tableau discussions that appears in any event once every month or something like that: somebody has an information set that they need to process a measure over, and the measure requires table calculations. At that point they need to utilize the consequences of those calculations as a discrete quality for dividing different calculations. For instance, given an arrangement of understudy evaluations, figure every understudy’s percentile, then demonstrate the normal of every percentile in the same perspective. Then again a comparable thought, just the percentile depends on entirety of offers, and you need to demonstrate the main 10% and base 10% of offers.

The issue here is that Tableau does not bolster apportioning table calculations by the consequences of other table calculations. You can segment table calculations by measurements, and even by discrete totals in the event that you kill Ignore Table Calculations utilizing the pill’s menus, yet not by table calculations. An option workaround is to do the percentile calc in one worksheet, send out the information, then utilize that as a datasource for an extra worksheet, however that has the confinements of a) not being dynamic and b) inclined to disappointment for datasets in the a large number of lines.

In any case, table calculations are supernatural things that can cut, dice, modify, re-total, and yes, Martha, even parcel by the aftereffects of other table calculations, and you’re going to figure out how. Because of Joe Mako for his spearheading work, it drove me to a Tableau epiphany that I’ll be partaking in a future post.

The Method

Here’s the fundamental procedure for parceling the consequences of a table estimation by other table calculations:

  1. Begin with a content table. Put every one of the measurements and discrete totals you’ll requirement for the last view on the Rows Shelf, with the tending to pill(s) on the right, apportioning pills on the left.
  2. Assemble the underlying table calc and check that it’s working legitimately.
  3. Assemble the parceling calc(s), include those and check them.
  4. Take the necessary steps to lessen covering content/extra stamps.
  5. Copy the worksheet and improve pills to get to the last view.

The dubious piece is step 3, since each calc that we need to use in the last result needs to fittingly reference the calc we are dividing over. For instance, for doing a normal of test scores inside a percentile, we have to register the aggregate of the scores inside every percentile for the numerator, and the aggregate number of tests inside every percentile for the denominator, and after that in a third calc create the normal. On the off chance that we need the normal of scores inside a scope of percentiles, then we’ll have to give back the scores for the extent and process the normal over that. Visit calcsoft.com to calculations for all occasions!

Related For Partitioning by Table Calculations