Filtering Matrix Hierarchy by Parent's Measure Value

[nbs33]

Hi,

I've been stuck on this problem for a few weeks. You can find a sample of the data and the model here. 

I have a Orders, Companies, and Dates table. The Companies table has a Parent/Child relationship that forms a ragged hierachy. 

I have visualized it in a matrix to show YTD Revenue, PYT Revenue, YTD vs PYTD and YTD vs PYTD %.

I would like to be able to filter the visual for Parents YTD vs PYTD % that fall in a 10% range.  For exmaple, I'd like a slicer that showed 100% to 90%, 89% to 70%...etc. and to be able to select one of these ranges to filter Parents whos YTD vs PYTD % fall in this range along with their childern, even if the childern have a different value. 

I have tried using dynamic segemnations patterens, also tried createing a calculated table to compute the YTD vs PYTD % and use as a filter. I think my DAX skills aren't advanced enough, as I struggle conceputally how to tackle this problem. Any guidence is greatly apprciated.

Thanks!

  

[AlexisOlson · Accepted solution]

Try defining a new parameter that has values from 0% - 100% and create a measure to filter based on those parameter values in a slicer.

I defined a new table named Range and defined this as a filter measure.

InSelectedRange =
VAR Level1Growth =
    CALCULATE (
        [YTD vs PYTD Revenue %],
        ALLSELECTED (),
        VALUES ( Companies[Level 1] )
    )
VAR RangeBuckets =
    FILTER (
        Range,
        Range[Range] <= Level1Growth
            && ( Range[Range] = 1
            || Range[Range] + 0.1 > Level1Growth )
    )
RETURN
    IF ( ISEMPTY ( RangeBuckets ), 0, 1 )

See attached file.

[nbs33]

Thank you @AlexisOlson, I am going to try and implement into my model. Question is there a way to filter a whole page for just the customers that fall in the ranage?

[AlexisOlson]

You could define a calculated column on the Companies table that buckets each [Level 1] into the corresponding range and then use that column for your slicer. This isn't dynamic though. Each [Level 1] item can only have a single value regardless of what report filters you have.

[nbs33]

Ok I'll play with that and see if the end result gets me close.

I added your measure filter to my model and it works except if I have a year selected in my date slicer. Do I need to add some type of date filter in the Level1Growth variable?

InSelectedRange = 
VAR Level1Growth = CALCULATE ( [YTD vs PYTD Revenue %], ALLSELECTED (), VALUES ( Companies[Level 1] ) )
VAR RangeBuckets = FILTER ( Range, Range[Range] <= Level1Growth && ( Range[Range] = 1 || Range[Range] + 0.1 > Level1Growth ) )
RETURN
    IF ( ISEMPTY ( RangeBuckets ), 0, 1 )

[AlexisOlson]

I'm not sure what you're saying doesn't work.

This is an example of what I see with a year selected:

Is this not the expected output?

[nbs33]

Strange, if you look at the pictures below both range filters are the same but populate the matrix differently. The only modification I made was to extended the range down to -1. 

I also have one additional page level filter to filter out inactive customers but removing this didn't change anything. 

[AlexisOlson]

Maybe because 2019 doesn't have a previous year?

[nbs33]

Maybe you're right, I expanded my date table to include the full range of dates in the Orders table and I still have a blank in the slicer. It's a head scratcher....the only other thing that is different is I user Company Name instead Company Id for the levels of the hierarchy. 

When I change the range slicer to less than zero the behavior is even stranger. The matrix shows positive percentages. 

[nbs33]

Hi @AlexisOlson, I've been trying to break you solution apart to see where it fails in my model. I think I found where the issue is coming from. 

When I test the effect of selecting the year on the RangeBuckets varible I get the results below. When I run this test in your file, the RangeBuckets remain -100% no matter if a year is selected or not. 

I tried using a REMOVEFILTER for the year but no luck. Any idea how to solve this?

[nbs33]

I figured out the issue, it is two fold. First, I had a relationship in my model that created a circular dependancy. Second, the filter mease is calculating in 10% increments so it includes values greater than the max or min value in the slicer. I change the slicer to a drop down so the user can select a bucket of value in 10% increments which makes more sense from a usability perspective.

[AlexisOlson]

I can't access the file you linked to.

[nbs33]

@AlexisOlson sorry, put the wrong type of permision in google drive. It works now, just checked.