Count the appearances of each value per bin within a vector of 144 bins.

I have a vector of 21600 values between [-3;2]. I want to divide this vector in 150-value bins and count how many times each value occurs within each bin.

Ideally I would get a matrix 144*6, with columns representing the bins and rows showing the number of occurrences of the values (-3, -2, -1, 0, 1, 2) within the each bin. I tried to use histcounts and bincounts, and other similar scripts of MATLAB Community, but I wasn't successful. Thanks a lot!

Answer 1

[Guillaume]

A method without a loop:

%demo data:
v = randi([-3 2], 1, 21600);
binsize = 150;

assert(mod(numel(v), binsize) == 0, 'number of elements not divisible by binsize')
destrow = repelem(1:numel(v)/binsize, binsize);
result = accumarray([destrow', v' + 4], 1)  %v' + 4 gives the destination column for values -3 to 2

[Carlos Goncalves Moreira]

Thank you very much, it works great! My next step would be adapting the script to any initial vector size, but your input makes that job already! Thanks again.

Answer 2

[Rik]

So you want to split your original vector into 144 vectors, and then get the histcounts of each vector?

data=randi([-3 2],216000,1);
results=zeros(144,6);
for n=1:144
    results(n,:)=histcounts(data(((n-1)*150+1):(n*150)),6);
end

[Carlos Goncalves Moreira]

Thanks a lot, it works great!