Since you don't specify the kind of integrals or the actual dimensionality, I can only suggest that you take into account that
where the function F(x) is defined as
and use this fact to compute your integrals with the usual quadrature techniques. For example, you could use trapz or quad in MATLAB. However, if the dimensionality is truly high, then you are better off using Monte Carlo algorithms.
发布评论
评论（4）
由于您没有指定积分的类型或实际维度，我只能建议您考虑
其中函数 F(x) 为定义为
并利用这一事实通过常用的求积技术来计算积分。例如，您可以在 MATLAB 中使用
trapz或quad。但是，如果维度确实很高，那么最好使用蒙特卡罗算法。Since you don't specify the kind of integrals or the actual dimensionality, I can only suggest that you take into account that
where the function F(x) is defined as
and use this fact to compute your integrals with the usual quadrature techniques. For example, you could use
trapzorquadin MATLAB. However, if the dimensionality is truly high, then you are better off using Monte Carlo algorithms.Google 上的第一个链接。
看起来相当强大。
First link off google.
Seems pretty roboust.
“Numerical Recipes In C”中有一个关于数值积分的非常好的章节。
也许高斯求积可以帮助你。
"Numerical Recipes In C" has a very nice chapter on numerical integration.
Maybe Gaussian quadrature can help you out.
是的，有TESTPACK，它是
C++ 程序，演示了多维集成例程的测试。Yes there is TESTPACK which is
C++program which demonstrates the testing of a routine for multidimensional integration.