Python package for modular Bayesian optimization
A Python package for modular Bayesian optimization.
This package provides methods for performing optimization of a possibly noise-corrupted function f. In particular this package allows us to place a prior on the possible behavior of f and select points in order to gather information about the function and its maximum.
The easiest way to install this package is by running
pip install -r https://github.com/mwhoffman/pybo/raw/master/requirements.txt pip install git+https://github.com/mwhoffman/pybo.git
which will install the package and any of its dependencies. Once the package is installed the included demos can be run directly via python. For example, by running
python -m pybo.demos.animated
A full list of demos can be viewed here.
The current version of
pybohas undergone some change to its interface from previous versions. The previous stable release(s) of the package can be found here. For example, those users interested in a graphical interface to
pybocan take a look at ProjectB, however this package requires a previous version which can be installed using:
pip install -r https://github.com/mwhoffman/pybo/raw/v0.1/requirements.txt pip install git+https://github.com/mwhoffman/[email protected]
Note the v0.1 tag in the installation lines which corresponds to the relevant release.