Analysis Menu --> Create a calculated Field --> A Pop UP wil open
1) Provide the name as ND
2) Write the below calc
3) Click Apply
4) Click ok
5) You will find the created calculation in Measures pane
6) Right Click on ND --> Default Properties --> Number format --> Select Percentage
7) Drag the ND after Universe in measure values Section
you will get the final O/P as Required
Hope this helps.
Plz mark this answer as correct or helpful to close the thread
Hi Naveen! Thank you for your attention
There is nuance I've forgot to tell: columns are not a measure but part of a pivot table.
Literally i need to calculate: ND = Value of the Fact named "sellers" / Value of the Fact named "Universe" * 100%
Could you please attach a sample workbook.
1 of 1 people found this helpful
You need to create separate calculation for each fact and use measure names and measure values to plot the data.
I'm sorry Naveen, but I don't get your point. Can you please share an example?
I've made all facts as calculated fields and calculate ND easily:
In table format it looks ok, because I can filter necessary facts to show:
But on a second sheet where facts displayed on a map one-by-one, I need to exclude at least 2 fields from the filter, but I don't know how:
Can you please suggest the workaround?
So, I've hide unnecessary facts and workaround looks fine until I faced with the issue: I'm not able to use Measure Names dimension as a single filter for multiple sheets on a dashboard.
Seems it ruins the idea of using calculated facts
The perceived complexity of this (seemingly trivial) task
is derived from the fact that the data are not in a proper 'shape'.
Namely, the [Fact] (as well as the related [Brand] ) has to be a Dimension (as it is),
but it should include additional Values -- to be used as the 'placeholder' Headers
for the additional Marks on a view populated with the 'derived' (calculated) Measures.
And there is a simple 'rule' in Tableau:
to have Marks on a view one should have Rows in the data.
This brings us to the concept of a so-called 'Scaffold' datasource --
a specially prepared one which is logically structured as a 'virtual cube'
consisting of all the necessary Dimensions with all the necessary Values
(aka domain-complete). The combinations of these Dimensions Values
form the 'Cells' of the cube -- which translates to the Marks on a view --
so they could be populated with the results of (aggregated) Measures,
potentially from other datasources as well (via a Data Blending mechanism).
Please find the attached as an example of such a Scaffold datasource (and a view).
Hope it could help understanding. Please reply with any question that may appear.
BTW, you can drop me a message or email, and we talk about things.
I'm very appreciated for your reply and suggestions, but I already managed to build my report without reshaping the source data.
For future projects I will follow your advice and try to prepare better shape for source.
Can you please share some links where I can learn the best practices?
Thanks once again! Wish you merry holidays and happy New Year!
Thank you for reply, i'm glad you've got it working.
As for the references, i would recommend this: