So, I think I've found a way to push the down into the dimensions on the results cube based on multiple attributes on the source cube without an intermediary cube. I created an alias attribute (called Attribute Alias) on the dimension in the source cube and concatenated all the attributes I was interested in, PLUS the principle name...
Code: Select all
['Attribute Alias']=S:
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'CPU Vendor')
|' '|
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'CPU Type')
|' ('|
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'CPU Max Sockets')
|' CPU) '|
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'Form Factor')
|'||'|
!MMAP Servers Generic Models
;
In the destination (results) cube that has more dimensions, I then used those additional dimensions (that are represented by attributes in the other cube) plus the principle name to get data into the correct position (effectively pushing the data down into more dimensions)...
Code: Select all
['Actual','Vendor Revenue']=N:db(
'MMAP MS Servers Input',
!MMAP Versions,
!MMAP Quarters,
!MMAP Countries,
!MMAP Vendors,
!MMAP CPU Vendors|' '|!MMAP CPU Types|' ('|!MMAP CPU Max Sockets|' CPU) '|!MMAP Form Factors|'||'|!MMAP Servers Generic Models,
'Actual',
'Vendor Revenue'
);
And finally, I had to have feeders to push into the results cube...
Code: Select all
FEEDERS;
['Actual','Vendor Revenue']=>DB('MMAP MS Servers Results',
!MMAP Versions,
!MMAP Quarters,
!MMAP Countries,
!MMAP Vendors,
!MMAP Servers Generic Models,
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'CPU Vendor'),
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'CPU Type'),
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'CPU Max Sockets'),
attrs('MMAP Servers Generic Models',!MMAP Servers Generic Models,'Form Factor'),
'Actual',
'Vendor Revenue'
);
I think this makes sense, and maybe can help others trying to do a similar thing.