skip to navigation
skip to content

skl-groups 0.1.0

Addon to scikit-learn for handling set-based data.

Package Documentation

Latest Version: 0.1.7

skl-groups is a package to perform machine learning on sets (or “groups”) of features in Python. It extends the scikit-learn library with support for either transforming sets into feature vectors that can be operated on with standard scikit-learn constructs or obtaining pairwise similarity/etc matrices that can be turned into kernels for use in scikit-learn.

For an introduction to the package, why you might want to use it, and how to do so, check out the documentation.

skl-groups is currently in early stages of development; feel free to get in touch ( if you’re interested.


This code is written and tested for Python 2.7, 3.3, and 3.4. 3.2 probably works. It is also only tested on OS X and Linux, though it’s not known not to work on other platforms; let me know if you try it.

Get reasonably recent versions of numpy, scipy, and scikit-learn, then run:

python install

to install skl-groups. Once the package is a little more mature, it will also be available through pip or conda, but not just yet.

If you want to use the kNN divergence estimator, you’ll need to have FLANN and either its Python interface (pyflann) or the cyflann library. These can be installed after installing skl-groups. If you’re running on non-tiny datasets, you’ll want a version with OpenMP support. If you use Anaconda, the cyflann page has a pointer on how to install both FLANN and cyflann through conda.

You may also want the skl-groups-accel package, which contains faster implementations of some of the algorithms in skl-groups; it’s split out to make the base package of skl-groups easier to install. (Currently, it only has the kNN divergence estimator).

File Type Py Version Uploaded on Size
skl-groups-0.1.0.tar.gz (md5) Source 2014-06-30 34KB