3D Matrix distance

I have a 15x3x3 matrix and I want to calculate the euclidean distance between one point and all other points in the matrix and store those values. I want to do this for every single element in my matrix. Logic:

element 1 of 135: calculate distance between element 1 and 2, 1 and 3, 1 and 4...., 1 and 135 -> then story distances in array

element 2 of 135: calculate distance between element 2 and 1, 2 and 2, 2 and 3, 2 and 4, ..... 2 and 135 -> then store distances in arry

element 3 of 135 .... " "

element 4of 135..... " "

you get the picture

This is what I have so far:

EtoIratio = 0.8;
gridspace = zeros(15,3,3);
nSelect = round(numel(gridspace)*EtoIratio);
idx = randperm(numel(gridspace),nSelect);
gridspace(idx) = 1;
neurontype = [];
[row, col, page] = size(gridspace);
for i = 1:row
    for j = 1:col
        for k = 1:page
        if gridspace(i,j,k) == 1
            neurontype(i,j,k) = 1;
        else neurontype(i,j,k) = -1;
      end
    end
    end
end
 ind = find(neurontype);
 [r c p] = ind2sub(size(neurontype),ind);

That is what I have so far... anyone have any idea on how to accomplish this?

Thanks!!!

[Rik]

Your code doesn't seem related to the text you posted. Do you want an array with Euclidean distances? Why would you want to store all of them? You will get 18225 values just for this small array. What is your end goal?

[Alex Henderson]

Hi Rik! I'll tell you what I'm trying to do here:

I have a 3D grid of excitatory and inhibitory neurons denoted as 1 and -1, respectively

Now I am trying to implement this probality function to get a probability matrix that determines the chance that a neuron will connect with another neuron (keep in mind that one neuron can have connections with many other neurons):

Once I get those probabilities I can use that as a constraint in an overall matrix that represents the connections (for exmaple, 1 if they form a connection 0 they dont form a connection)....

lambda is a control parameter already defined

D(a,b) is the Euclidean distance between neurons a and b

C is a constant whose value depends on the type of neurons that are being connected: 0.3 (EE), 0.2 (EI), 0.4 (IE), 0.1 (II) (still trying to figure out who to implement this as well)

As of now I have the coordinates of all 135 neurons. I believe my next step is getting those 18,225 values representing those distances. I think that's where I'm stuck!!

[Alan Stevens]

If lambda is of the order of, or smaller than, your grid spacing then the probability will fall off extremely quickly with distance. I suggest you first do the calculation using nearest neighbour distances, then extend that to next nearest neighbours if that isn't sufficient.

First, perhaps, find what distance (i.e. what value of D) makes the probability insignificant for your problem.