MCMCLib is a C++ library of Markov Chain Monte Carlo methods.
Features
- C++11 implementations of several well-known MCMC methods, including:
- Random Walk Metropolis-Hastings (RWMH);
- Metropolis-adjusted Langevin algorithm (MALA);
- Hamiltonian Monte Carlo (HMC); and
- Riemannian Manifold HMC.
- Samplers designed for multi-modal distributions:
- Equi-Energy sampling; and
- Differential Evolution
- Built on the Armadillo C++ linear algebra library for fast and efficient matrix-based computation.
- OpenMP-accelerated samplers for parallel computation.
- Straightforward linking with parallelized BLAS libraries, such as OpenBLAS.
- Released under a permissive, non-GPL license.
Author: Keith O'Hara
Download and Installation
- The source code is available on GitHub.
The library can be installed on Unix-alike systems via the standard ./configure && make method:
# clone MCMCLib into the current directory git clone https://github.com/kthohr/mcmc ./mcmc # build and install cd ./mcmc ./configure -i "/usr/local" -p make make install
The last line will install MCMCLib to /usr/local
Configuration options: (see ./configure -h)
-
Primary
-hprint help-iinstall path; default: the build directory
example:-i "/usr/local"-mspecify the BLAS and Lapack libraries to link against;
examples:-m "-lopenblas"or-m "-framework Accelerate"-ocompiler optimization options;
default:-O3 -march=native -ffp-contract=fast -flto -DARMA_NO_DEBUG-penable OpenMP parallelization features (recommended)
Secondary
-ca coverage build (used with Codecov)-da 'development' build-ga debugging build (optimization flags set to-O0 -g)