Using the dimensional Member() function

[snowy]

I'd be interested to see if anyone can tackle this thorny dimensional issue.

I have a prompt which consists of a set of members AND an "invented" member which I want to define. The code for the prompt_selection data item is as follows:

Code Select Expand

union(
[Sales].[Retailer type].[Retailer (by type)].[Retailer type]
,
set( member( [Golf Shop] + [Sports Store], 'SPORTS AND GOLF', 'SPORTS AND GOLF', [Sales].[Retailer type].[Retailer (by type)] ) )
)

and this shows exactly the list I want in a value prompt: a list of all the retailer types, plus 'SPORTS AND GOLF' (my "invented" member which represents the total of Sports and Golf).

BUT when I select SPORTS AND GOLF and use the prompt output in a data item:

Code Select Expand

set( [Sales].[Retailer type].[Retailer (by type)] -> ?pSelectedRetailer? )

...I get a stinking error OP-ERR-0217, complaining that "An unexpected type of parameter value was encountered. The specified parameter type is 'String', but the expected type is 'memberUniqueName'"

So what gives? Can the member() function ever REALLY create a genuine MUN? Have I invented a member() or not?!

[Francis aka khayman]

let me take a shot...

I can see that your question is about the new member rather than making your report run, so i'll confine my reply to that.

i think a MUN is used to identify a member from a dimension/hierachy/level. since your new member created on the fly, it do not need a MUN.

and this command:
set( [Sales].[Retailer type].[Retailer (by type)] -> ?pSelectedRetailer? )

doesn't make sense if you choose the invented member from your prompt as you know in the first place the invented member is not in found in that level.