DNest4 is a C++11 implementation of Diffusive Nested Sampling, a Markov Chain Monte Carlo (MCMC) algorithm for Bayesian Inference and Statistical Mechanics.

There is a manuscript describing DNest4 installation and usage in the paper/ directory of this repository. You can compile it with pdflatex. Alternatively, you can get the preprint here.

You might also want to read the paper describing the Diffusive Nested Sampling algorithm. If you find this software useful in your research, please cite one or both of these papers.

Diffusive Nested Sampling Brendon J. Brewer, Livia B. Pártay, and Gábor Csányi Statistics and Computing, 2011, 21, 4, 649-656.

The algorithm paper is freely available online at the arXiv.

• There are far fewer dependencies --- all you need is a C++ compiler that supports the C++11 standard, along with Python (and the Python packages NumPy, scipy, matplotlib, and Cython) for the post-processing scripts. Because of this, it should be much easier to compile (at least on a Unix-like operating system such as Ubuntu or Mac OS X).

• The licence is now the permissive MIT licence.

• RJObject (which allows relatively straightforward implementation of hierarchical and trans-dimensional mixture models) is now included in the same repository.

• There is now some support for implementing models in Python and Julia, making DNest4 usable to non-C++ programmers.

Some functions have slightly different names and specifications compared to DNest3. These changes are cosmetic, but mean that porting models from DNest3 to DNest4 will take a little bit of work.

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.

However, you can also compile and install the library using SCons. To do this you just need to run:

scons install

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>

where <install_location> is the base path for the install.

To install with GDB enabled during the library's compilation you can add the --debug-mode 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 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.