Statistics Using Field Theory
Project description
SUFTware is a lightweight Python package that provides provides fast and robust implementations of Bayesian Field Theory (BFT) methods for low-dimensional statistical inference. BFT is a grid-based approach to Bayesian nonparametric inference. By using a grid in lieu of specific stochastic processes (such as Dirichlet processes or Gaussian processes), BFT allows certain types of problems to be solved in a fully Bayesian manner without requiring any large-data approximations.
Currently, SUFTware supports a one-dimensional density estimation called DEFT. DEFT has substantial advantages over standard density estimation methods, including, including kernel density estimation and Dirichlet process mixture modeling. See [Chen et al., 2018; Kinney 2015; Kinney 2014].
Installation¶ pip install suftware Requirements
Python >= 3.6.3 or Python = 2.7.10 numpy >= 1.13.3 scipy >= 1.0.0 matplotlib >= 2.1.0 Quick Start import numpy as np import suftware as sw
# Generate random data data = np.random.randn(100)
# Perform one-dimensional density estimation using SUFTware density = sw.Density(data)
# Visualize results density.plot()
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for suftware-0.6a0-py2.py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 73dee079aaaccd4b8ba8eb0c1fa62b9ddb29fdbcac2e6c112e366accebc599c8 |
|
MD5 | 3b33b9472ea4e2f163909ecf3b7ed885 |
|
BLAKE2b-256 | e45b57ce65dea3f13c0fb5914ba9f48db8c44f5921cc936d0c4b72bb1756aae1 |