
1. Re: Divide a measure by itself with varying only one dimension/column
Jim Dehner Nov 9, 2018 4:21 AM (in response to Harley Chen)Good morning Harley 
Not certain I understand the goal but the formulas that you created are looking at individual records in the data set and identifying the point at which one of the dimensions changes  eg age goes from 39 to 35  It then looks at the current record avg (that is the average of that single record not all the records in the "30 group" and compares it to the first record in the next group "the 35 group"
If you coal is to compare averages in the entire age subgroups you will need to start by creating the subgroup averages using LOD's  (the LOD will create the average for the subgroup and return it as a single value that is NOT an aggregate in itself) then use the LOD expressions in you conditional statement that looks for the divisions between groups
Jim
If this posts assists in resolving the question, please mark it helpful or as the 'correct answer' if it resolves the question. This will help other users find the same answer/resolution. Thank you.

2. Re: Divide a measure by itself with varying only one dimension/column
Harley Chen Nov 9, 2018 8:01 AM (in response to Jim Dehner)Thanks Jim.
The reason that I used AVG is because LOOKUP can't simply take [Rate] as is, even though there is only 1 unique value of [Rate] with the way I setup the grids in the sheets.
For example, in the Age sheet, the cell of Age 30 and Term 12 should be a formula of
([Rate] of the next age in ascending order)/([Rate] of the current age 30), with all other variables/factors being the same between the numerator and denominator. 
3. Re: Divide a measure by itself with varying only one dimension/column
Harley Chen Nov 27, 2018 2:33 AM (in response to Harley Chen)Bump! Please can anyone help

4. Re: Divide a measure by itself with varying only one dimension/column
Yuriy Fal Nov 27, 2018 2:58 AM (in response to Harley Chen)Hi Harvey,
With your existing view setup and calculations,
you may want to take care of the proper Addressing
of your Table Calculations.
For example, your Band calc should be
Compute using > SA Band (only this Dim).
This would create the table calculation windows
with only the [SA Band] values changing (in the sort order) 
and every other Dimensions values to remain the same.
Hope it could help a bit.
Yours,
Yuri

5. Re: Divide a measure by itself with varying only one dimension/column
Harley Chen Dec 6, 2018 6:14 AM (in response to Yuriy Fal)Thanks for replying!
So I think I got it to work. But issue now is that, when I try to apply color to get a heat map visualization, it colors all the blank as the lowest level.
The formula I have for Age measure is now,
IIF(ZN(SUM([Rate])) / LOOKUP(ZN(SUM([Rate])), 1)=0, NULL, ZN(SUM([Rate])) / LOOKUP(ZN(SUM([Rate])), 1))
Please see the screenshot,
I have also attached the Workbook, it's sheet "Age1".

Example1.twbx 201.6 KB


6. Re: Divide a measure by itself with varying only one dimension/column
Yuriy Fal Dec 6, 2018 3:40 PM (in response to Harley Chen)Hi Harley,
You may want to apply the [Age1Rate]
as a Table Calculation Filter,
and select NonNull Values.
This effectively hides the Marks
where [Age1Rate] is Null.
Please find the attached.
Yours,
Yuri

Example1_v10.2_YF.twbx 336.9 KB
