dnest4 0.2.4

Diffusive nested sampling in Python

These are experimental Python bindings to the DNest4 sampler allowing the user to implement their models in Python. See the example(s) in examples for some demos for how you might do this.

Building

DNest4 relies on recent C++11 features so building with python can be fun. On a Mac, you might need to do something like:

MACOSX_DEPLOYMENT_TARGET=10.9 python setup.py install

If you only want the postprocessing scripts and don’t want to worry about building the bindings, you can run:

python setup.py install --without-cext

and then use it as follows:

import dnest

# NOTE: the path will default to the current working directory if you
#       don't give another path.
dnest.postprocess("/path/to/where/you/ran/your/model")

Author

These bindings were written by Dan Foreman-Mackey and they are licensed as part of DNest4 with the MIT license.