How to calculate using Total from columns

[ngupta]

I have a very simple problem, but I am not good with DAX or even measures yet, so this may be preventing me from getting the answer.

I have a dataset with the following information.

Region	C Estimate	D Estimate/Low Bid	(D-C) Difference	Variance from Estimate = (D-C)/C
South	790	603	-187	-23.7%
Central	1,988	2,000	12	0.6%
South	9,857	8,515	-1342	-13.6%
South	11,296	40	-11256	-99.6%
South	1,401	1,128	-273	-19.5%
North	273	350	77	28.2%
Total	25,605	350		-50.7%

I want it to calculate the variance between Estimate and Bid for various categories, but I need the Variance of the total.

Question: How do I get -50.7% in the graph, which is coming from “Total” Row.

What I am getting is the average of the variance  -21.3%, this is the average of the individual numbers

Average= (-23.7% +0.6% + -13.6% + -99.6% +-19.5% +28.2%)/6 =-21.3%

Right now I am calculating Variance (column 5) using the following measure: 

Variance = DIVIDE(BI_TENDER_DIMS[CvsD Diff],BI_TENDER_DIMS[C Estimate]) 

Please note numbers in the image below are same as numbers above, but divided by 1000 approx. 

 

Variance calculation only gives me a few options, and none of those include calculation on the total SUm value. These rows od data will change based on several filters, but I always need the variance from the total value, and I cannot calculate that yet.

 

[v-ljerr-msft · Accepted solution]

Hi @ngupta,

According to your description above, you should be able to use the formula below to create a measure(instead of a calculate column) to calculate the Variance in this scenario. 

Variance = DIVIDE(SUM(BI_TENDER_DIMS[CvsD Diff]),SUM(BI_TENDER_DIMS[C Estimate])) 

Regards

[ChrissyG]

Hi, I am having a very similar issue but it seems like I am doing what you suggest and still getting an incorrect subtotal on my matrix. I am also very new to Bi, so I apologize if a super easy question, which I hope it is! I also may have done this completely wrong!  Thanks for your help

I am trying to show the month over month variance for revenue, which I have but the subtotal is giving me the last calculated number, instead of the total variance, which I want to be the total of all variances. In the example below the total should be $3,409,237.76.  also similar, is the additional formula to find the % variance. 

I used the following measures to get to this: 

MTD Rev = TOTALMTD(sum(Revenue[Total Revenue]),'Calendar'[Date])

LY MTD Rev = calculate([MTD Rev],SAMEPERIODLASTYEAR('Calendar'[Date]))

MTD Rev Var = [MTD Rev]-[LY MTD Rev]

 MTD Rev Var % = Divide([MTD Rev Var],[LY MTD Rev])

 

[ngupta]

Yes, this works. Thankyou 

I did try it earlier, but I was getting an error, probably because the syntax was wrong.