topic Week over Week difference in DAX Commands and Tips

Week over Week difference

martaTr — Fri, 05 Apr 2024 15:54:46 GMT

hi all, probably this is super easy, but i struggle with it, i need to present on below matrix table difference week over week just for the last two periods (in below between 20 and 27 march), so it basically should show me -22 etc. i did it in excel but we moved to power bi. this is a count of lines. i also have the same with sum of import.

could you please help me how to do it? i tried with calculate and max or with dates in period. but it shows me strange results.

Re: Week over Week difference

adudani — Fri, 05 Apr 2024 16:53:05 GMT

hi martaTr ,

if you have a weeknum column in your dataset, that maybe used. ( for more context: How to Calculate Week over Week Change in Power BI (youtube.com)

OR,

you could explore if the (52) Calculate Growth From Previous Non Consecutive Date - YouTube may be a better fit.

Another alternative maybe to explore DATEADD function (DAX) - DAX | Microsoft Learn

Kindly provide a sample input table masking sensitive data if this doesn't work for you.

Re: Week over Week difference

martaTr — Fri, 19 Apr 2024 11:48:09 GMT

thank you, it seems like it's working, however i am wondering how can i this new metric WoW add to the table.

Here is my formula:

Measure3 = VAR LastDate1 = MAXX ( ALLSELECTED ( Affinity[As of Date] ), Affinity[As of Date] ) VAR LastSales = CALCULATE ( COUNT ( [Importo] ), FILTER ( ALL ( Affinity[As of Date] ), Affinity[As of Date] = LastDate1 ) ) VAR PrevDate = MAXX ( FILTER ( ALL ( Affinity[As of Date] ), 'Affinity'[As of Date] < LastDate1 ), Affinity[As of Date] ) VAR PrevSales = CALCULATE ( COUNT ( [Importo] ), FILTER ( ALL ( Affinity[As of Date] ), Affinity[As of Date] = PrevDate ) ) RETURN IF ( ISBLANK ( PrevDate ), BLANK (), LastSales - PrevSales )

Unfortunately it doesnt work when i normally add it to columns/ rows/values.Now it's in seprate table hidden under a main one which doesnt look good and it 's completely impractical. Any ideas how to present that metric? maybe other chart?