pysmac

Simple python wrapper to [SMAC](http://www.cs.ubc.ca/labs/beta/Projects/SMAC/), a versatile tool for optimizing algorithm parameters.

SMAC is free for academic & non-commercial usage. Please contact Frank Hutter to discuss obtaining a license for commercial purposes.

This wrapper is intented to use SMAC directly from Python to minimize a objective function. It also contains some rudimentary analyzing tools that can also be applied to already existing SMAC runs.

Installation

To install pysmac, we advise using the Python package management system:

pip install git+https://github.com/automl/pysmac.git --user

If you prefer, you can clone the repository and install it manually via

python setup.py install

Documentation

The documentation for pySMAC is hosted on http://pysmac.readthedocs.org/. It contains a Quickstart guide for the impatient, but also a detailed manual.