Count Metric based on two physical columns

Question

I have a fact table that has Beginning Inventory On Hand (BOH) Units and Ending Inventory On Hand (EOH) Units.
It also contains sales dollars and sales units and other facts. My dimensions are: Locations, Time, Products.
Locations
     All Stores
     Region
     District
     Store
Products
     All Products
     Division
     Entity
     Subdivision
     SubClass
Time
     Week
We have fact tables at the lowest level and we have aggregrate tables at certain points.
We are trying to create some count metrics in the BMM Layer. We have created a metric that will count the number
of stores that have Sales Units at each level of the product dimension. We did this by mapping the logical column
to the physical column of Sales Units. We then defined the aggregation for this metric based on dimensions. The
first thing we needed to do is Sum the Sales Units across Products, so we set the "Other" as a sum. We then need
to apply the Count Distinct, so we set this at the Location Dimension.
Based On Dimensions
Sum(Sales Units)                              - Other Dimensions
Count (Distinct Case When Sales Units > 0 then store number else null end )     - Location Dimension
This will give a final rule as follows:
Count (Distinct Case When SUM(Sales Units) > 0 then store number else null end )
This is working great. However, we now want to create a logical metric that will count the number of stores based on the evaluation of two physical columns.
Count the number of stores that have Sales Units > 0 OR EOH Units > 0.
I do not know how to evaluate two physical columns and apply a count against them.
Thanks in advance for your help.

Answer 1

create same column for sum(eoh)
Count (Distinct Case When SUM(Sales Units) > 0 then store number
else
case when SUM(eoh) > 0 then store number else null end
else null end )