Forum Discussion

admin11's avatar
admin11
Memorable Member
5 years ago
Solved

My Vari % return wrong Total summary % , How to correct it ?

Hi All

 

_Vari_%_NP = if(isblank( divide(
[_YTD_NP_SGD]
-
[_LYTD_NP_SGD]
,
[_LYTD_NP_SGD]
)),1, divide(
[_YTD_NP_SGD]
-
[_LYTD_NP_SGD]
,
[_LYTD_NP_SGD]
))
 
 

My PBI file :-

https://www.dropbox.com/s/0ln5k19oj4i8cfp/PBT_V2021_400%20GL_TI.pbix?dl=0

 

Paul

  • Anonymous's avatar
    Anonymous
    5 years ago

    HI admin11,

    It seems like a common calculation issue when you try to apply multiple aggregates on measure expression.

    For this scenario, you can take a look at the below blog about measure calculation on the total level.

    Measure Totals, The Final Word 

    BTW, we do not so recommend you to nested multiple measures in one calculation, they may conflict with the calculation when different formulas calculate on current row contents.

    Optimizing DAX expressions involving multiple measures - SQLBI
    Regards,
    Xiaoxin Sheng

2 Replies

  • Hi admin11 ,

     

    First of all you can use the COALESCE to make the value for the non blank values:

    In second place if you look at the totals you have the correct percentage calculated:

     

    369.94 - (-155.336,66) = 155.706,60

     % =  155.706,60 / -155.336,66 = -1,0023 * 100 = -100.23%

     

    Since you are rounding it gives -100%.

     

    What is the value you need to have on the totals is it the sumx of the percentages?

  • Anonymous's avatar
    Anonymous
    Not applicable

    HI admin11,

    It seems like a common calculation issue when you try to apply multiple aggregates on measure expression.

    For this scenario, you can take a look at the below blog about measure calculation on the total level.

    Measure Totals, The Final Word 

    BTW, we do not so recommend you to nested multiple measures in one calculation, they may conflict with the calculation when different formulas calculate on current row contents.

    Optimizing DAX expressions involving multiple measures - SQLBI
    Regards,
    Xiaoxin Sheng