MATLAB looping was dramatically speeded up several years ago. It's no longer that slow, though the belief persists. If you have less than a few million iterations, it will be pretty fast.
One way to speed it up even more, when you have nested loops, is to have the inner-most loop iterate over the left most index, since MATLAB is in column-major mode, meaning it goes down rows in a single column first before moving over to the next column. So, in summary
% Slower:
for row = 1 : rows
for col = 1 : columns
m(row, col) = some formula;
end
end
% Faster:
for col = 1 : columns
for row = 1 : rows
m(row, col) = some formula;
end
end
And that might only hold for very large matrices (millions of elements). For small matrices (a few tens of thousands of elements) you might not notice a difference.
