No description, website, or topics provided.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
bcg
cpp
notebooks
.gitignore
LICENSE
README.md
makefile
setup.py

README.md

BCG

BCG provides an implementation of the Bayesian Conjugate Gradient method from the paper by Cockayne et. al. which can be found here.

Installation

The package depends on the C++ linear algebra library eigen3. On Ubuntu systems this can be installed with sudo apt install libeigen3-dev, and on OSX with brew install eigen3 (via Homebrew).

It requires Python 3 and the following libraries, all of which can be installed with pip install:

  • numpy
  • scipy
  • Cython
  • eigency

To install, ensure the above dependencies are installed. Then clone the repository and type make install. This will compile the C++ code and install the library in the active virtualenv.

Usage

Example usage can be found in the Jupyter notebook in notebooks/Demo.ipynb

Credits

This library is written and maintained by Jon Cockayne.