Forum Discussion

CJBoelt's avatar
CJBoelt
Frequent Visitor
3 years ago
Solved

How do I create a DAX calculating the change per sequence?

Here is a Example of what the new graph would be.   Essentially subtracting the actual_hours with all the values in the previous rows. This will provide a measure to show the difference of ch...
CALCULATE
DAX
  • wdx223_Daniel's avatar
    wdx223_Daniel
    3 years ago

    =SUM(Table[[Total Actual hours])-CALCULATE(SUM(Table[[Total Actual hours]),TOPN(1,FILTER(ALLSELECTED(Table[SEQ]),Table[SEQ]<MAX(Table[SEQ])),Table[SEQ]))

wdx223_Daniel's avatar
wdx223_Daniel
Icon for Community Champion rankCommunity Champion
3 years ago

=SUM(Table[[Total Actual hours])-CALCULATE(SUM(Table[[Total Actual hours]),TOPN(1,FILTER(ALLSELECTED(Table[SEQ]),Table[SEQ]<MAX(Table[SEQ])),Table[SEQ]))

CJBoelt's avatar
CJBoelt
Frequent Visitor
3 years ago

wdx223_Daniel 
This works! Thank you!

 

I have one additional question if you are able,
I have a function [Gaussian_Hours] which calculates a forecast of values which by a normal Distribution, though for some reason this new measure is incredible show (to measure the increase by SEQ. 


 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Do you have any ideas for why the performance is slow?

Thank you!

  • wdx223_Daniel's avatar
    wdx223_Daniel
    Icon for Community Champion rankCommunity Champion
    3 years ago

    performance optimization is  hard work, it involves many aspects. you can try this code, maybe it won't work or not.

    SeqChange=[Gaussian_Hours]-CALCULATE([Gaussian_Hours],OFFSET(-1,,ORDERBY(PROJ_FRIDAY_SEQ[SEQ#])))