Concatenated Fact Table Issue

[#1]

Hello. I am having trouble calculating the % part churn by a dimension. I have an example data set included.

The part churn is calculated by  =Count(Part)/Sum(Total_Part_Count).  I am able to graph by manager using =Count(Part)/Aggr(Total_Part_Count, Key) but as soon as I select a Manager the denominator is obviously null.

Is there a way that when I filter to a manager I can still get the Total_Part_Count for the key?  I have tried several ways with Set Analysis but nothing has worked.  Any help would be great. Thank you.

Part	Manager	Key	Total_Part_Count
Part 1	Joe	1A
Part 2	Joe	1A
Part 3	Bill	2B
Part 4	Bill	2B
Part 5	Ted	3C
		1A	5
		2B	10
		3C	1

[Anonymous]

you may use set Analysis to desekect Manager

like this

Count({<Manager=>}Part)

the part {<Manager=>} must be placed in every calculation - Count, sum

[sunny_talwar]

Are you sure you want to do a concatenate here? To me you will be better off doing a Left Join (or Join)

Table:

LOAD Part,

          Manager,

          Key

FROM xyz;

Join(Table)

LOAD Key,

          Total_Part_Count

FROM abc;

Best,

Sunny

[#4]

I tried both suggestions to either join the data or use set analysis but could not get either to work. Any time I filter to a manager I lose the total_parts in the denominator.  Any other suggestions?

I have included the qvw for assistance.

[sunny_talwar · Accepted solution]

Have a look with join. (PFA)

Best,

Sunny

[#6]

It looks like it is working. The only difference that I had done was an inner join. Thank you!

[sunny_talwar]

Inner join should have also worked, not sure why that did not work for you.

Best,

Sunny