Using find in a 4D matrix

I have multiple 9 x 96 x 14 x 1356 matrixes. I want to find when one of them meets a certain criteria and then select the same indexes on the other matrixes.

I did this and looked liked it worked...

idx = find(neutral<24.5);
[r, c, v, l] = ind2sub(size(neutral),idx);

With this, I get a r, c, v, l with 4411652 x 1 each, instead of a 4D logical matrix, which was what I was expecting (like what happens with a 2D matrix when using find).

I tried using those indexes to cut my other variables, but I could't do it.

temp_AT = temp(r,c,v,l);

Thanks!

[Rik]

Do you really need it to be 4D logical? If not, you can directly use the indices.

[Thaís Lobato Sarmento]

Not necessarily... I just need a way to use the information (neutral<24.5) to select the other 4D variables... And this is the way I know how to do it (in 2D matrixes). But I am open to suggestions...

[Adam Danz]

d = other4Darray(neutral < 24.5)  %assuming both arrays are same size
% d will be a vector

idx = find(neutral<24.5);
% d(i) comes from the index idx(i)

[Thaís Lobato Sarmento]

this kind of works... but I get the same result as I put in my question (4411652 x 1 matrix). I need it to keep the 4D format...

[Adam Danz]

When you remove or isolate elements from an array, they no longer have the same shape unless you're indexing all of the data. Here's a simple example.

d = [2, 3, 2;
    5, 2, 2]; 
idx = d == 2; %same shape as d

d(idx) % a vector with 4 elements, not 6!

If you're trying to get rid of data and maintain the shape of the data, you can replace unwanted data with NaNs (or any other value).

d = [2, 3, 2;
    5, 2, 2]; 
idx = d == 2; %same shape as d

d(~idx) = NaN;  % same shape as d!

[Thaís Lobato Sarmento]

Thanks, this worked!

Answer 1

[Walter Roberson]

neutral<24.5

is the 4d logical matrix. No need for find()

[Adam Danz]

+1 (moving my inferior answer here)

idx = find(neutral < 24.5);
logIdx = false(size(neutral)); %Logical index (default: all false)
logIdx(idx) = true;            %Set 'idx' indices as true

[Thaís Lobato Sarmento]

Thank you! This also worked!