R MPFR - Multiple Precision Floating-Point Reliable

Arithmetic (via S4 classes and methods) for arbitrary precision floating point numbers, including transcendental ("special") functions. To this end, the package interfaces to the 'LGPL' licensed 'MPFR' (Multiple Precision Floating-Point Reliable) Library which itself is based on the 'GMP' (GNU Multiple Precision) Library.

Installation is non-trivial if you install from _source because of the SystemRequirements (listed in ./DESCRIPTION):

The package Rmpfr interfaces R to the C Library MPFR:

MPFR, the "Multiple Precision Floating-Point Reliably" library

which is Free/Libre Software, available under the LGPL license. MPFR Website

MPFR itself is built on and requires the GMP library

GNU Multiple Precision arithmetic library (GMP)

Obtain that from GMP Website or from your operating system vendor / package system:

+ Under _Debian_, _Ubuntu_ (and other Debian derivative) Linux distributions,
  it is sufficient (for *both* libraries) to simply do

  sudo apt-get install libmpfr-dev

+ In Fedora, Redhat, CentOS, opensuse, etc, you get these via

  sudo dnf install mpfr-devel
 

The standard reference to MPFR is

@article{FouLHLPZ-2007,
 author = {Laurent Fousse and Guillaume Hanrot and Vincent Lef\`{e}vre and
 	   Patrick P\'{e}lissier and Paul Zimmermann},
 title = {MPFR: A multiple-precision binary floating-point library with
          correct rounding},
 year = {2007},
 journal = {ACM Trans. Math. Softw.},
 volume = {33},
 number = {2},
 issn = {0098-3500},
 pages = {13},
 doi = {http://doi.acm.org/10.1145/1236463.1236468},
 publisher = {ACM},
 address = {New York, NY, USA},
}