选择层次结构中的所有级别作为MDX中的元组

Question

I made my first cube yesterday. I'm still new to this, so please forgive any misuse of terminology. One of my dimensions, let's call it MyDimension1 has a hierarchy, let's call it MyHierarchy , with two levels, let's call them Level1 and Level2 .

I seem to be able to do something like:

SELECT 
{ [Measures].[Whatever] } ON 0,
{ ([MyDimension1].[Level1].[Level1], [MyDimension1].[Level2].[Level2]) } ON 1
FROM MyCube

Which gives me the result I want. If I understand correctly, the set defining axis 1 contains a tuple with two components(?). I've read that each component(?) of a tuple is supposed to be from a different dimension, but I seem to be able to specify components of the same dimension, as demonstrated by the above query. I'd like to be able to just specify the hierarchy by name and then have the server create the tuples for me, though. So something like:

SELECT 
{ [Measures].[Whatever] } ON 0,
{ AllLevelsToTuple([MyDimension1].[MyHierarchy]) } ON 1
FROM MyCube

But I can't seem to figure out how to do this. Thanks in advance!

This does what I want, functionally, but isn't the syntax I'd like:

SELECT [Measures].[Original] ON 0,
([Customer].[Customer].[Customer], [Customer].[Account ID].[Account ID]) ON 1

Results (In SSMS):

               Original
---------  --  --------
CustomerA  15    306.03
CustomerA  16    754.20
...
CustomerB  17    524.43
...
CustomerC  22    760.42
...

When I expand the dimension, it looks like:

- Customer
    - Hierarchy
        + Members
        + Customer
        + Account ID

When I do something like (I'd like to do something similar syntactically):

SELECT [Measures].[Original] ON 0,
([Customer].[Hierarchy].Members) ON 1

I get the following results:

            Original
----------  --------
All         17638.15
CustomerA    2624.76
15
16
...
CustomerB    3113.67
17            524.43
...
CustomerC    3427.01
22            760.42
...

I want there to be 3 columns, not 2, basically.

Answer 1

You might want to read this MDX gentle introduction .

I've read that each component(?) of a tuple is supposed to be from a different dimension

From different hierarchy instead. In AS, I guess that for each level, you have as well a corresponding flat hierarchy; so the following looks like you're accessing levels of two different hierarchies:

{ ([MyDimension1].[Level1].[Level1], [MyDimension1].[Level2].[Level2]) } ON 1

I'm not a specialist of AS but I guess the following statement :

SELECT [Measures].[Original] ON 0,
([Customer].[Customer].[Customer], [Customer].[Account ID].[Account ID]) ON 1

is actually interpreted as a crossjoin of level members from 2 different hierarchies; it is more likely that the () notation is interpreted as the () operator instead of tuple notation and then {},{} is a crossjoin in MDX :

SELECT [Measures].[Original] ON 0,
[Customer].[Customer].[Customer].members * [Customer].[Account ID].[Account ID].members ON 1

hence the two columns in front of your measures in the result.

I want there to be 3 columns, not 2, basically.

You'll need to use calculated measures; something like :

with 
MEMBER Measures.HN as [Customer].Currentmember.Hierarchy.Name 
MEMBER Measures.LN as [Customer].Currentmember.Level.Name 
MEMBER Measures.MN as [Customer].Currentmember.Name 

SELECT { Measures.HN, Measures.LN, Measures.MN, [Measures].[Original] } ON 0,
([Customer].[Hierarchy].Members) ON 1

Answer 2

Yes, you can create a tuple from different dimensions, and you do so by placing members between parentheses and separated by commas.

But in this case, you don't need a tuple, simply a set. A set contains members from a single dimension, separated by commas and placed between curly braces: {}.

Have a look at the descendants function. You can create a set from your hierarchy using this function.