Forum Discussion

jav226's avatar
jav226
Frequent Visitor
7 years ago
Solved

Computing rollup weighted sum

Seeking assistance from Power BI experts.   I'm trying to compute the overall progress of Program 1, which is sum of each individual Project Progress, based on the weightage assigned for each Proje...
Need Help
  • d_gosbell's avatar
    d_gosbell
    7 years ago

    So based on your new requirement to apply the weightings differently at different levels you might be able to do something like the following which just dynamically changes the factor inside the SUMX based on which columns are filtered

     

    SUMX( table1, IF(ISFILTERED( table1[Project]), 1,  -- at the project level multiply by 1
        IF( ISFILTERED(Program], [Project Weightage],  -- at the program level apply the project weight
        [Program Weightage] * [Project Weightage])   -- at the total level apply program and project weighting
    * [Progress])

jav226's avatar
jav226
Frequent Visitor
7 years ago

Hi,

 

I hit another roadblock, hope you can assist again. 

 

Your DAX works perfectly fine when I display the records in a table.

 

When I display a chart showing the overall All Programs progress, and start to slice to only Project A and Project B, it sums up the Progress of Project A and Project B which is not correct. Because I'm at overall All Programs level, the expectation is to sum up the progress of Project A and Project B with consideration of the Project and Program weightage. 

d_gosbell's avatar
d_gosbell
Super User
7 years ago

It should not matter if the data is visualized in a table or a chart.

 

But I don't really understand what your issue is - I can't understand how you can be both at the total level and slicing by a project. Can you maybe share a screenshot or something? 

  • jav226's avatar
    jav226
    Frequent Visitor
    7 years ago

    In Pic01 below, when I'm not slicing any Project, the overall progress of 0.82% at the bar chart is correct.

    Pic01

    But at Pic02 below, when I slice to Project A and Project E, the Overall Progress in bar chart becomes 14.31%. The expected progress should be 0.64% [ Project A (8.14% * 8.80% * 36%) + Project E (6.17% * 9.60% * 64%)].

     

    Pic02

    My DAX script as following:-
     
    Cumm Planned (%) =
    SUMX(All_DATA, IF(ISFILTERED( 'ALL_DATA'[L4 - Discipline]), [Planned (%)],
    IF( ISFILTERED('ALL_DATA'[L3 - Phase]), [L4 Weightage (%)], IF( ISFILTERED('ALL_DATA'[L2 - Package]), [L3 Weightage (%)] * [L4 Weightage (%)],
    [L2 EPCC Weightage (%)] * [L3 Weightage (%)] *[L4 Weightage (%)])) * [Planned (%)])
    )
  • d_gosbell's avatar
    d_gosbell
    Super User
    7 years ago

    Hmm, I was just about to reply back saying that you might just have to create 3 separate measures, then I remembered that PowerBI added a new ISINSCOPE function in Nov last year.

     

    You'll have to adjust the column names in the code below as I used the sample table you posted earlier to test this code, but I think this might work for both your tables and charts.

     

    Cumm Planned (%) 2 = SUMX( ALL_DATA, IF(ISINSCOPE( ALL_DATA[Project]) , 1,  -- at the project level multiply by 1
    IF( ISINSCOPE(ALL_DATA[Program]) , [Project Weightage (%)] ,  -- at the program level apply the project weight
    [Program Weightage (%)] * [Project Weightage (%)]))   -- at the total level apply program and project weighting
* [Project Progress (%)])
  • jav226's avatar
    jav226
    Frequent Visitor
    7 years ago

    It works! Thanks again. 

  • jav226's avatar
    jav226
    Frequent Visitor
    7 years ago

    Hi Gospel,

     

    Seems like IsInscope is computing the % correctly when I filter at L3 - Phase. 

    When I filter at L2 - Package level, IsFiltered computation returns the correct value.

     

    Is there a way to combine both IsFiltered and IsInScope to cater both filtering at L3 - Phase and L2 - Package as one single measure?

     

    Thanks again.

     

     

     

    Cumm Planned IsFiltered(%) =
    SUMX(All_DATA, IF(ISFILTERED( 'ALL_DATA'[L4 - Discipline]), [Planned (%)],
    IF( ISFILTERED('ALL_DATA'[L3 - Phase]), [L4 Weightage (%)], IF( ISFILTERED('ALL_DATA'[L2 - Package]), [L3 Weightage (%)] * [L4 Weightage (%)], [L2 EPCC Weightage (%)] * [L3 Weightage (%)] *[L4 Weightage (%)])) * [Planned (%)])
    )
     
    Cumm Planned InScope(%) =
    SUMX(All_DATA, IF(ISINSCOPE( 'ALL_DATA'[L4 - Discipline]), [Planned (%)],
    IF( ISINSCOPE('ALL_DATA'[L3 - Phase]), [L4 Weightage (%)],
    IF( ISINSCOPE('ALL_DATA'[L2 - Package]), [L3 Weightage (%)] * [L4 Weightage (%)],
    [L2 EPCC Weightage (%)] * [L3 Weightage (%)] *[L4 Weightage (%)])) * [Planned (%)])
    )
  • d_gosbell's avatar
    d_gosbell
    Super User
    7 years ago
    jav226 wrote:

     

    Seems like IsInscope is computing the % correctly when I filter at L3 - Phase. 

    When I filter at L2 - Package level, IsFiltered computation returns the correct value.

     

    Is there a way to combine both IsFiltered and IsInScope to cater both filtering at L3 - Phase and L2 - Package as one single measure?

    Hi jav226 - I'm seeing the same figures returned by the IsFiltered and IsInScope versions of the calc. It's hard to fix a problem I can't see. Are you able to share the details (or screenshots) of how you have your visuals configured when they return the "incorrect" results?