Average Distinctcount of orders but disregarding blank values in field of another table

[Applicable88]

Hello,

I have a typical facttable and a dimension table. My facttable is a order table where I track all orders and from which shelf something was taken:

Shelfno		OrderID
1234		111111
1235		222222
1236		333333
1237		444444
1238		555555
1239		666666
1240		777777
1241		888888
1242		999999
1243		1111110
1244		1222221

My dimension table consist of the details about which material is in the shelf:

Shelfno		Material
9999
1240		mmm
1244		nnn
8786
1236		ggg
1239		ddd
5555
1234		dhd
3333
1243		jdj
5563

Both tables are connected via the shelf no. and bidrectional filtering is turned on. 

I need the average distinctcount of OrderID of each Shelfno. And also disregard the shelfs with blank Materials.

I tried something with Averagex, Addcolumns and Summarize but the problem is that both are in different tables, and I also don't no how to filter out the blank Materials from the counts. 

Thank you very much in advance. 

Best. 

[amitchandak · Accepted solution]

@Applicable88 , Try a measure like

AverageX(values(Table2[Shelf]), calculate(distinctcount(Table1[OrderID])))