Forum Discussion
Weird Behavior in YOY analysis?
Hi PBI Gurus,
I'm requesting your help in trying to figure out some weird behavior of Power BI.
Take a look at my screenshot. The top graph is displaying correct data. The following fields are being used:
Total Shipped Orders = DISTINCTCOUNT (Orders)
Axis is Month ( from date table )
Legend is Year ( from date table )
The bottom graph is showing incorrect data for THIS YEAR. ( It seems to be adding this year and last year together )
Total Shipped Orders LY = CALCULATE( [Total Shipped Orders] , SAMEPERIODLASTYEAR( [Date] ))
You can see the correct, verified values in the table below both graphs.
The reason this is causing me trouble, is because I want to display a bar chart below the top graph that is % difference from last year. However, because it is not calculating current year correctly, the % change is obviously different.
My % change measure is, % Change = DIVIDE ( [Total Shipped Orders LY] - [Total Shipped Orders] , [Total Shipped Orders LY], 0 )
For example, June % change is showing me an INCREASE of 81%, when in actuality, it should be a DECREASE of -18.5%.
Any idea of why this is happening? Or more importantly, how to fix it?
Thank you so much!
Nevermind,
I just fixed it. It looks like I just had to apply a YEAR filter set to 2019.
Thanks for your help!
3 Replies
- amitchandak
Super User
You seem to be using [Total Shipped Orders], which seems like a measure.
Can you share its calculation? In case it is a measure, is possible to have the same calculation in last year's formula itself.
Thanks
- CoreyP
Solution Sage
Yes, it is a measure, I thought I specified it's calculation in my OP.
Here are my measures.
Total Shipped Orders = DISTINCTCOUNT( 'Orders'[Order Number] )
Total Shipped Orders LY = CALCULATE( [Total Shipped Orders] , SAMEPERIODLASTYEAR( 'Date'[Date] ) )
% Change from Last Year = DIVIDE( [Total Shipped Orders] - [Total Shipped Orders LY] , [Total Shipped Orders LY] , 0 )
My dataset is only 2018/2019.
- CoreyP
Solution Sage
Nevermind,
I just fixed it. It looks like I just had to apply a YEAR filter set to 2019.
Thanks for your help!