Arb is a C library for high-performance arbitrary-precision floating-point ball
(mid-rad interval) arithmetic. It supports complex numbers, polynomials,
matrices, and evaluation of special functions, all with rigorous error
bounding.

Arb depends on FLINT (http://flintlib.org/), MPIR (http://mpir.org)
and MPFR (http://mpfr.org). It is currently built as a standalone library,
but in the future might become an extension module for FLINT.

For documentation and additional details, see: http://fredrikj.net/arb/

Development updates are occasionally posted to: http://fredrikj.net/blog/

Author: Fredrik Johansson <fredrik.johansson@gmail.com>

The build scripts are copied from FLINT and were written by Bill Hart.