mj-will/nessai-models

Models for use with nessai

nessai-models

Models for use with the nested sampling package nessai.

Included models

• n-dimensional unit Gaussian
• n-dimensional HalfGaussian
• n-dimensional Rosenbrock
• n-dimensional mixture of Gaussians
• n-dimensional slab plus spike model
• Gaussian mixture using data to based on this example from cpnest
• n-dimensional Egg Box based on the version in Feroz et al. 2008
• n-dimensional Pyramid-like model
• n-dimensional Brewer likelihood (Skilling's "Staistical Model") from Brewer et al.
• Linear signal plus Gaussian noise model (LinearSignal)
• Sinusoidal signal plus Gaussian noise model (SinusoidalSignal)
• Mixture of 1-dimensional distributions (MixtureOfDistributions)

Requirements

nessai_models requires:

• numpy
• scipy
• nessai>=0.6.0

Installation

We recommend following the installation instructions for nessai and then installing nessai_models since it shares all of its dependencies with nessai.

nessai_models can be install from PyPI using

pip install nessai-models

Example usage

Below is an example of using nessai_models so configure a 4-dimensional Gaussian and then sample it using nessai.

from nessai import FlowSampler
from nessai_models import Gaussian

model = Gaussian(4)
fs = FlowSampler(model, output='example/')
fs.run()

Citing

If you use nessai_models in your work please cite the Zenodo DOI