Neural network accelerated nested and MCMC sampling. The target distribution is first transformed into a diagonal, unit variance Gaussian by a series of non-linear, invertible, and non-volume preserving flows. Efficient MCMC proposals can then be made in this simpler latent space.
NNest can be installed via pip
pip install nnest
Alternatively the latest version can be obtained by
git clone https://github.com/adammoss/nnest
cd nnest
python setup.py install
Nested sampling examples can be found in the examples/nested directory, and can be run with e.g.
python run.py --x_dim 2 --likelihood rosenbrock
There is also an example notebook. Runs can be analysed by
python analyse.py
MCMC sampling (work in progress) examples can be found in the examples/mcmc directory, and can be run with e.g.
python run.py --x_dim 2 --likelihood rosenbrock
There is also an example notebook.
Please cite Moss (2019) if you find the package useful in your research.