Add particular elements of array

Let's say that k is a vector and

k = [1 1 1 1 4 1 1];

k could be anything I just assumed [1 1 1 1 4 1 1]. Let's say there is a variable

limiter = 3;

I want to add elements of k that are less than or equal to the variable 'limiter' and put it in a new vector, let's say k_new. For example, k(1),k(2),k(3),k(4) <= limiter, add them and put it in k_new resulting

k_new = [4];

Now k(5) is greater than limiter, it should be ignored and should be stored in k_new as it is, resulting

k_new = [4 4];

the final desired vector is

k_new = [4 4 2]

I worte a program using two while loops and somehow the loop keeps running. I want to do this using a single while loop if possible and I need help in doing that. Below is my code

limiter = 3;
k = [1 1 1 1 4 1 1];
run1 = true;
while (run1)
    ii = 1;
    jj = 1;
    run2 = true;
    while (run2)
        if (k(ii) < limiter)
            k_new(jj) = k(ii)+k(ii+1)
            ii = ii + 2;           % increment of 2 because we need
                                   % to check k(ii+2) as k(ii) and k(ii+1) have been checked
        else
            k_new(jj) = k(ii)
            ii = ii + 1;
        end
        jj = jj + 1;
        if (ii == length(k))
            k_new(jj) = k(ii)
            break
        end
        if (ii > length(k))
            run2 = false;
        end
    end
    k = k_new
    clear k_new
    
    if (find(k==min(k),1,'last') == length(3))   % Stop the outer while loop
        run1 = false;
        break
    end
end

Answer 1

[Rik]

Not the tidiest code, but it all happens in a single loop, with cumsum doing the addition prior to the loop.

%set your data
data=[1 1 1 1 4 1 1];
limiter=3;

%add all elements together
c1=cumsum(data);c2=NaN(size(c1));
%c1 will hold the corrected cummulative sums (masked with -inf)
%c2 will contain the end result (masked with NaNs)

while true
    ind=find(c1>limiter,1);
    if isempty(ind)
        %break loop
        if isnan(c2(end)),c2(end)=c1(end);end
        [c1;c2]
        break
    end
    c1((ind+1):end)=c1((ind+1):end)-c1(ind);
    c2(ind)=c1(ind);
    c1(1:ind)=-inf;
    [c1;c2]
end

ans = 2×7

-Inf -Inf -Inf -Inf 4 5 6 NaN NaN NaN 4 NaN NaN NaN

ans = 2×7

-Inf -Inf -Inf -Inf -Inf 1 2 NaN NaN NaN 4 4 NaN NaN

ans = 2×7

-Inf -Inf -Inf -Inf -Inf 1 2 NaN NaN NaN 4 4 NaN 2

k_new=c2(~isnan(c2))

k_new = 1×3

4 4 2

[Awais Saeed]

Hey @Rik, It worked brilliantly and that's a very good way of using cumsum. Thank you very much. Just wondering if we can add elements which are >= limiter instead of <= limiter? I might need to do this in future.

[Rik]

I don't think I completely understand what you mean, but it probably would not be too hard to modify this code. I would encourage you to try it yourself.