in reply to Optimizing a double loop
Now I have about 1000 long arrays dumped in store files. All arrays are of the same length (~10^6), each cell contains some small positive int (0-500). No undef values.
I would like to get the average and standard deviation for each array position. For this, I use two "aggregating" arrays @sum and @sum_squares initialed with zeros. I then retrieve from the disk one array at a time and add all of its values (loop...) to the corresponding places in the "aggregating" arrays. Finally, I do the calculations for each position
Any suggestion on how to improve that? Perhaps there is a faster way to "sum up" arrays?