[net.math] Non-random Monte-carlo?

mwm@ea.UUCP (04/14/84)

#N:ea:6900002:000:539
ea!mwm    Apr 13 17:51:00 1984

I am currently working on a problem that involves integration via
monte-carlo techniques. However, instead of choosing n*n random points, I'm
using a regularly spaced grid. The code I'm writing looks like:

	for i: int in int$from_to(1, n) do
		for j: int in int$from_to(1, n) do
			x := i * xdelta + xmin
			y := j * ydelta + ymin
			end
		end

As far as I can tell, this won't make things worse than using a true
monte-carlo, and should make things better. Could somebody with more
statistics than I have please comment?

	Thanx,
	<mike