How to display pre-calculated non-aggregated value in decomposition tree or matrix?

[Michelle113]

I am facing the issue when use decomposition tree to analysis non-aggregation measures. (Percentile 75 of duration)

My dataset like:

level1	level2	level3	level4	Percentile-75
Homepage				3
Homepage	Workbenchpage			2
Homepage	Workbenchpage	Policies		5
Homepage	Workbenchpage	Alert		1

The analyze is target at Percentile-75, explain by level1,level2,level3,level4. 

Composition Tree analyze measure has default aggregation function. 

Our expecting analysis is based on the pre-calculated percentile-75 in the dataset. 

Explain by each level display the specific value without any aggregation.

Could anyone share if there's any method to realize this requirement? Thanks!

[Michelle113]

Thank you so much, Kelly!  It's a wonderful solution. I am unblocked with it.

[v-kelly-msft · Accepted solution]

Hi  @Michelle113 ,

Create a measure as below:

measure = IF(ISINSCOPE('Table'[level4 ]),
CALCULATE(SUM('Table'[Percentile-75]),FILTER(ALL('Table'),'Table'[level4 ]=MAX('Table'[level4 ])&&'Table'[level4 ]<>BLANK())),
IF(ISINSCOPE('Table'[level3]),
CALCULATE(SUM('Table'[Percentile-75]),FILTER(ALL('Table'),'Table'[level3]=MAX('Table'[level3])&&'Table'[level3]<>BLANK())),
IF(ISINSCOPE('Table'[level2]),
CALCULATE(SUM('Table'[Percentile-75]),FILTER(ALL('Table'),'Table'[level2]=MAX('Table'[level2])&&'Table'[level2]<>BLANK())),
IF(ISINSCOPE('Table'[level1]),
CALCULATE(SUM('Table'[Percentile-75]),FILTER(ALL('Table'),'Table'[level1]=MAX('Table'[level1])&&'Table'[level1]<>BLANK())),
BLANK()))))

And you will see:

For the related .pbix file,pls see attached.

Best Regards,
Kelly

Did I answer your question? Mark my post as a solution!