Set analysis - data island

[#1]

Hi,

I've come across a tricky set analysis operation.

I have two tables (refer to sample app for further details):

t_fact:

LOAD * INLINE [
    Dimension A, Measure #1, Measure #2
    F, 6, 1
    F, 5, 2
    G, 4, 3
    G, 3, 4
    H, 2, 5
    H, 1, 6
];


t_dataisland:


LOAD * INLINE [
    Measure #3, Measure #4, Measure #5
    11, 3, 10
    7, 7, 20
    3, 11, 30
];

I want to create a simple straight table with Dimension A as Dim and a single expression which should be the sum of [Measure #5). Given that there is no connection between tables I want to use Set Analysis in a way similar to this:

= Sum({<[Measure #3]={"$(=Sum([Measure #1]))"}, [Measure #4]={"$(=Sum([Measure #2]))"}>}[Measure #5])

In fact, what I'm saying is that given the sum of Measures #1 and #2, retrieve the value of Measure #5:

Dimension A	Measure #5
F	10
G	20
H	30

I am missing something because I can't get it to work... maybe you'll have better clues than I do.

Thanks in advance.

[sunny_talwar · Accepted solution]

Although won't give you an amazing performance, but this expression works:

=Sum(Aggr(If([Measure #3] = Sum([Measure #1]) and [Measure #4] = Sum([Measure #2]), [Measure #5]), [Dimension A], [Measure #3], [Measure #4], [Measure #5]))

[#3]

It certainly works.

I am amazed that this can't be achieved with set analysis.

Thank you for your help.