[tanderson92]

I've always found DCTs and particularly Clenshaw-Curtis quadrature (better than Gauss quadrature arguably!) attractive: you can do an incredible amount of heavy lifting with this corner of approximation theory.