From: spellucci@mathematik.tu-darmstadt.de (Peter Spellucci) Subject: Re: numerical integration methods - Cube Date: 26 May 1999 11:27:48 GMT Newsgroups: sci.math.num-analysis In article <374B169C.8DA93CE0@nwu.edu>, "L. D. Marks" writes: |> On a similar theme, I need to integrate a function over a |> three-dimensional cube; note, function evaluations are VERY |> expensive in computer time. I tried the formula in Abramowitz and |> Stegun, ... snip for this special case there are good formulae in stroud: approximate calculation of multiple integrals. because of the special geometry these are more efficient than monte-carlo. but for higher dimensions monte carlo is competitive or even superior, see e.g. the recnet paper "numerical integration using sparse grids" by thomas gerstner and michael griebel, abteilung wissenschaftliches rechnen universitaet bonn (preprint) hope this helps peter