Efficiency help

Hi, i need some more help with effeciency. This following code is running very slowly (took over 800 seconds for one of my data sets)

 for l= 1:listsize;
     for m = 1:length(peak_loc2{l});
        counter2 =1;
        while peak_loc2{l}(m)~= mymatrix2(1, counter2)
            counter2=counter2+1;

        end
        mymatrix2(2,counter2) = mymatrix2(2,counter2) +1;
    end
end

here listsize is 271 peak_loc2 is a list of cell, around 100 cells, each containing anywhere from 2 to a few hundred variables

mymatrix2 is a 2X3000s that list all the peaks in peak_loc2, and non-recurring

Answer 1

[Sean de Wolski]

you can replace the while loop with

counter2 = find(mymatrix(1,:)==peak_loc2{1}(m),1,'first')

[Andy]

for l= 1:listsize;

for m = 1:length(peak_loc2{l});

counter2 =1;

counter2 = find(mymatrix(1,:)==peak_loc2{1}(m),1,'first')

counter2=counter2+1;

mymatrix2(2,counter2) = mymatrix2(2,counter2) +1;

end

end

like that?

[Andy]

thanks it worked well

[Andy]

hey, i ran both my version and your version, turns out my version gives over 3000 variables, while the one you gave me is faster by a lot, it only gave me a bit over 100 variables, why is this?

[Sean de Wolski]

If you have over 3000 variables there's something we're not seeing. Perhaps you could post a sample matrix (small) or a *.mat file to somewhere.

[Andy]

it turns out it still graphs it properly the way i wanted, so its fine thanks