Numerical integration from -Inf to Inf of an integrand in the form f(x)*exp(-x^2)
f(x) must be a slowly varying function
fn_val = GaussHermite(func, npt, varargin)
func is a handle for f(x)
npt is the number of points at which f(x) is to be evaluated. A look-up table of abscissas and weights is used, so npt must be 2, 4, 6, 8, 10, 12, 16 or 20.
varargin contains any extra parameters for f
TestGaussHermite(k,b) tests the integration for f(x) = 1, x^2, cos(bx),
selected with k = 0, 1, 2 respectively.
Inspired by: benbarrowes/f2matlab