what popular advance mathematics libraries for c++ are present out there, so that they can be used as a 1 stop solution and avoiding reinventing the wheel ?
+1
A:
Intel's MKL (Math Kernel Library) is to be looked at especially if doing large scale matrix operations; it's C based, but should not really be an issue IMO.
Other than that, maybe the boost math library could be interesting as it is free. (but I have no experience with it, so YMMV).
Max.
Max
2010-05-10 14:27:52
Most of the MKL is based on the BLAS and LAPACK interfaces, so it is really Fortran-based, not C-based (it's obvious when you use it since matrices are assumed to be column-major and everything gets passed by pointer). It has C language bindings, though, which make it easier to use with C++. Boost uBLAS is only a BLAS implementation, but there are LAPACK bindings in the Boost incubation (I've never used them, so I can't make a recommendation on their use; I've generally just done the binding myself when I've needed to use LAPACK with uBLAS's matrix classes).
James McNellis
2010-05-10 15:09:38
+4
A:
Check out GNU Scientific Library -- it's in C, but I use it all the time to avoid re-writing the Numerical Recipes code.
jeff7
2010-05-10 14:35:22
A:
For commercial libraries, both NAG (Numerical Algorithms Group, http://www.nag.co.uk/) and IMSL ( http://www.vni.com/products/imsl/ ) are standards and provide industrial-strength numerical analysis algorithms.
c-urchin
2010-05-10 14:44:03
A:
look through the list and mix-and-match. You want very many things, unlikely any single package is going to do them all.
octave is the only one that is going to be more or less comprehensive (functionality comparable/clone to Matlab)
aaa
2010-05-10 14:59:29
A:
Like others have said, you will probably not find a single library to handle all of the areas you listed. For matrix algebra, I've heard good things about the Eigen C++ library from coworkers who are using it.
Jason
2010-05-10 15:32:03