Average Weekly Sales

[DaveyP]

Hi 

I am having trouble working out an average sales DAX expression, it is for a group of stores where the data table (Keys) contains one line per store per week,

The issue i am facing is all stores were not open for all weeks so the average needs to be worked out on a store level based on the weeks they were open

i currently have created a calculated column within a structure table that only has one itterance of each store as below

AWUS (structure) = calculate(sum(Keys[Royalties])/DISTINCTCOUNT(Keys[Week]),RELATEDTABLE(Keys))

This gives me an average sales by store but none of the time filters on the page have any effect ( i assume this is the average sales for all the data relevent to each store)

How can i make the filters made on the date table effect this column?

Thanks in advance!

[v-xjiin-msft]

Hi @DaveyP,

Generally, if we want to filter data based on a Date table. We can

1. Create a relationship on the Date table and source table.

2. Add Filter() inside the expression.

Since you didn't share us your table structure or some sample data. It is hard for us to give a specific solution. Thus, please kindly share us more information which can help us understand your requirement more clearly.

Thanks,
Xi Jin.

[DaveyP]

Sorry let me try and be more clear, in the example below 4 stores accross 4 weeks. The average store sales for the group is not simply total sales / number of stores / number of weeks. This wont work because 2 stores were not open in week 4 so to work it out i need to work out store by store then average it eg - 

Store 1 total sales / 4

store 2 total sales / 4 

store 3 total sales / 3

store 4 total sales / 3 

Then average the 4 results

Question is how can i do this in DAX in a way that both the stores and weeks are still filterable on the page.

Store	Week	Sales
Store 1	week 1	400
Store 2	week 1	500
Store 3	week 1	600
Store 4	week 1	700
Store 1	week 2	500
Store 2	week 2	600
Store 3	week 2	700
Store 4	week 2	800
Store 1	week 3	700
Store 2	week 3	800
Store 3	week 3	900
Store 1	week 4	600
Store 2	week 4	700
Store 3	week 4	800

[ChrisMendoza]

DaveyP, 

I have something that may work for you as it is something similar to a project of mine.

With the data you supplied:

Create two measures:

TotalSales = SUM(Table1[Sales])

CountWeeks = DISTINCTCOUNT(Table1[Week]) 

The third measure for Avg/Week:

AvgSales/Week = DIVIDE([TotalSales],[CountWeeks],BLANK())

To get something like the following:

This should work with your dataset.

In my project, while I have a dCalendar table, I had to take a DISTINCTCOUNT of the 'Dates' from the working table rather than the dCalendar table so the denominator was in the 1000's range rather than10,000's.

[DaveyP]

Hi Dozer

Thanks for your reply but this still doesnt give the right result, it is still total sales/total distinct weeks and doesnt work when 1 store wasnt open for that week

Dave

[ChrisMendoza]

Hi @DaveyP, may I ask what you expect the answer to be in your dataset that you provided?

[DaveyP]

Hi Dozer

Sorry to confuse, below is the example in more detail, the results i keep getting are essentially all stores sales added together (9300) divided by count of stores (2325) divided by count of weeks (581.25)

Week	Store1	Store2	Store3	Store4	AWUS
1	400	500	700	600	550	(All stores sales divide by count of store (4))
2	500	600	800	700	650	(All stores sales divide by count of store (4))
3	600	700	900	800	750	(All stores sales divide by count of store (4))
4	700	800			750	(All stores sales divide by count of store (2))
					675	Average sales = all weeks AWUS / Count of weeks

[ChrisMendoza]

@DaveyP,

Well this is interesting...adding an Average Line, You get the $675 but I do not know how it's calculated.

[jszlm]

How about this solution?

1) create a measure

AWUS = AVERAGEX(VALUES(Table1[Week]),CALCULATE(AVERAGE(Table1[Sales])))

 2) use matrix, set AWUS as Values.

[DaveyP · Accepted solution]

GOT IT!!!

AWUS Actual =
VAR __CATEGORY_VALUES = VALUES('Keys'[MonDate])
RETURN
 DIVIDE(
  SUMX(
   KEEPFILTERS(__CATEGORY_VALUES),
   CALCULATE(SUM('Keys'[Royalties]) / DISTINCTCOUNT('Keys'[Store Name]))
  ),
  SUMX(
   KEEPFILTERS(__CATEGORY_VALUES),
   CALCULATE(DISTINCTCOUNT('Keys'[MonDate]))
  )
 )