Measure for Subtotals, using lower level measures, as values ​​for calculations

[Anonymous]

Hi All !

I need a measure for subtotals, using lower level measures, as values ​​for calculations. (This is a big problem that I can not solve for a long time.) See this file pls. 

I have a table with sales visits for different dates. 

Date	Superviser	City	TradeAgent
01.01.19	A	ABC	A1
02.01.19	A	ABC	A2
01.01.19	A	ABC	A1
02.01.19	B	ABC	B1
01.01.19	B	ABC	B2
02.01.19	B	ABC	B1
01.01.19	B	ABC	B1
03.01.19	A	ABC	A1
02.01.19	A	ABC	A2
03.01.19	A	ABC	A1
01.01.19	B	ABC	B2
02.01.19	B	ABC	B1
04.01.19	A	ABC	A1
05.01.19	A	ABC	A2
04.01.19	A	ABC	A1
05.01.19	B	ABC	B1
01.01.19	B	ABC	B2
02.01.19	B	ABC	B1
01.01.19	B	ABC	B1
02.01.19	A	ABC	A1
02.01.19	A	ABC	A2
03.01.19	A	ABC	A1
01.01.19	B	ABC	B2
02.01.19	B	ABC	B1

I need to calculate a call rate plan hierarchy:

call rate plan  = isfiltered( [tradeAgent] ,
distinctcount(date visit) * distinctcount(tradeAgent) * 25 ,
isfiltered ( [superViser] , 
sum (values lower lvl measure) and etc

Now measure give correct result only for TradeAgent :

call rate plan  = distinctcount(date visit) * distinctcount(tradeAgent) * 25 

I thought sumx would give the desired result for SuperViser and City, but Sumx gives wrong result (very high results, Why Sumx  dont give disitinct (date) and (tradeAgent) i dont know) :

Call rate plan for SuperViser = sumx (disitinctcount(date)*disitinctcount(tradeAgent)*25)

If you have ideas on how to calculate values for the entire hierarchy up to GrandTotal, please share.

When solving this problem, it will be possible to look average sales per visit (for example, for a period of 3 days, there is point A with 3 visits and an order for $ 300 where the average value is 300/3 = 100 $, and point B with 2 visits and order for $ 300 where the average value is 300/2 = 150$, which are included in the ABC zone for it, the measure will count 600 $ /3 (distinct(date))= $ 200, but it will be correct to take the average of the average (100 + 150) / 2 = 125 $ .  Very bad that can not work with values ​​in the hierarchy

[d_gosbell · Accepted solution]

So basically it sounds like you want the sum of the distinct dates for each agent * 25. The following expression will do that and gives the output you want.

= SUMX(VALUES(Table1[TradeAgent]), CALCULATE(DISTINCTCOUNT(Table1[Date])) * 25)