How to split an array of integers into arrays of varying lengths where all the elements are close in value to each other?

A function I wrote (ThermoOptimzation) outputs two 15x999 matricies (G_exp, and G_model), representing thermodyanmic data over 15 temperatures and 999 compositions. I am trying to find where these two matricies are equal, but have to use a wide tolerance range or rounding in order to obtain at least a single value for each temperature point (row).

[G_exp, G_model] = ThermoOptimization(nVars, nIter);
G_exp = round(G_exp,3,'significant');
G_model = round(G_model,3,'significant');

for i = 1:length(G_exp(:,1))
    ind = find(eq(G_exp(i,:),G_model(i,:)))
end  

In doing this, I get multiple points for some rows. For example, at row 2:

ind =

     2     3     4   998   999

This is clearly two distinct "groupings", an equality at composition (column) 2-4 and another at 998-999. I'd like to seperate ind into two sets, that I can use later on. I've tried a for loop and a while loop checking if the next element was less than 5 away from the current one, but I'm unsure what to do to split the array when the difference is greater than 5.

    for j = 2:length(ind)
        if (ind(1,j) - ind(1,j-1)) < 5
            continue
        else 
        
        end
    end

Answer 1

[Bruno Luong]

ind = [2     3     4   998   999]

lgt = diff(find(diff([-Inf, ind, Inf])>=5));
gind = mat2cell(ind,1,lgt);

results:

>> gind{:}

ans =

     2     3     4


ans =

   998   999