DNest4 is a C++11 implementation of Diffusive Nested Sampling, a Markov Chain Monte Carlo (MCMC) algorithm for Bayesian Inference and Statistical Mechanics. You can use it in a few different ways:
There is a paper describing DNest4 installation and usage in the Journal of Statistical software. You might also want to read the original paper describing the Diffusive Nested Sampling algorithm itself. If you find this software useful in your research, please cite one or both of these papers. Here are the citations:
Brewer, B., & Foreman-Mackey, D. (2018).
DNest4: Diffusive Nested Sampling in C++ and Python.
Journal of Statistical Software, 86(7), 1 - 33. doi:http://dx.doi.org/10.18637/jss.v086.i07
Brewer, B. J., Pártay, L. B., & Csányi, G. (2011). Diffusive nested sampling.
Statistics and Computing, 21(4), 649-656.
You will need a C++ compiler that supports the C++11 standard, along with Python 3 and the Python packages NumPy, scipy, matplotlib, Cython, and Numba. Anaconda is a suitable Python distribution to use.
On some Macs,
g++ is an alias for
clang, which is a C compiler. If this
is the case for you, you'll need to edit the first line of the Makefile so
that it uses
clang++, which is a C++ compiler.
You can compile the DNest4 library (
libdnest4) using the Makefile in the
code directory using:
cd code make
Along with building the library this will compile all the examples. Then, install the Python package.
The easiest way to do this is with:
pip install -U dnest4
but you can install the development version using:
python setup.py install
in the root directory of this repository.
However, you can also compile and install the library using SCons. To do this you just need to run:
By default it will attempt to install the library in
/usr/local (with the library files in
/usr/local/lib and the headers in
/usr/local/include/dnest4), so the above command
must be run as a user with root access or using
sudo. To install to a different location
you can instead run:
scons install --prefix <install_location>
<install_location> is the base path for the install.
To install with GDB enabled during the library's compilation you can add the
flag to the install command.
Currently, the Scons installation does not compile the examples or the Python library. Any additions to this installation process are welcome.
(c) 2015--2018 Brendon J. Brewer and contributors. LICENCE: MIT. See the LICENSE file for details.
This work was supported by a Marsden Fast Start grant from the Royal Society of New Zealand.