Library of arbitrary accuracy of linear algebra c / C ++ with complex numbers

Question

I am researching linear differential equations with complex coefficients in a 4-dimensional phase space. To test some hypothesis about the root of the solution, I need to be able to solve these equations numerically with arbitrary accuracy. I used the mpmathPython module , but it runs slower, so I prefer to rewrite my program in C / C ++ for maximum performance. So I have a question:

Is there a C / C ++ linear algebra library that supports both arithmetic and complex precision numbers. I need some basic features like dot products, etc. (In fact, I also need a matrix exponent, but I can implement it myself if necessary).

I tried using Eigenwith MPFR C++, but could not because it does not support complex numbers (and the type construct complex <mpreal>does not work, since it assumes that the base type is a standard float).

Answer 1

, , zkcm , . , ( ), , , .

Answer 2

uBLAS boost.

Answer 3

(, GMP http://gmplib.org/) , (, Eigen http://eigen.tuxfamily.org/)