How to slice an array across multiple dimensions with two limit arrays?

I'm interested in taking an arbitrary N-dimensional array, and two 1xN vectors, and using the two vectors as start/stop indices to slice the array, but I'm not sure how. Here's what I mean:

Let X be an N-dimensional array.

Let A and B be 1xN vectors like so:

A=[a1, a2, ..., aN]

B=[b1, b2, ..., bN]

How do I slice X such that I get this:

Y = X(a1:b1, a2:b2, ..., aN:bN)

I tried X(A:B), but that doesn't work - not even sure what that operation is doing, but it isn't doing what I want. I tried to think of some way to do it with logical indexing, but nothing simple came to mind. I could slice the array one dimension at a time in a loop, but it seems like there must be a one-liner. Thanks!

Answer 1

[Bruno Luong]

This should work

X=rand(6,6,6);
A=[1 2 3]; B=[6 5 4];

cidx = arrayfun(@(i1,i2) i1:i2, A, B, 'unif', 0);
Y = X(cidx{:})

Y = 
Y(:,:,1) =

    0.7739    0.3475    0.5936    0.5165
    0.5514    0.0120    0.8295    0.7020
    0.7219    0.0188    0.0139    0.2095
    0.6081    0.2888    0.2122    0.9462
    0.7840    0.4617    0.7677    0.1717
    0.3399    0.1233    0.8264    0.1021


Y(:,:,2) =

    0.8648    0.4441    0.7968    0.1729
    0.2677    0.8310    0.6914    0.3000
    0.7044    0.0935    0.5411    0.2597
    0.1844    0.8915    0.9166    0.0287
    0.8937    0.8579    0.5319    0.6198
    0.1937    0.8000    0.0968    0.2708

[Brian Kardon]

Aha, very clever indeed! Thank you!