https://github.com/djsutherland/skl-groups
scikit-learn addon to operate on set/"group"-based features
https://github.com/djsutherland/skl-groups
Last synced: 4 months ago
JSON representation
scikit-learn addon to operate on set/"group"-based features
- Host: GitHub
- URL: https://github.com/djsutherland/skl-groups
- Owner: djsutherland
- License: bsd-3-clause
- Created: 2014-06-10T22:36:35.000Z (about 12 years ago)
- Default Branch: master
- Last Pushed: 2016-08-08T22:38:51.000Z (almost 10 years ago)
- Last Synced: 2026-02-23T02:31:44.426Z (4 months ago)
- Language: Python
- Homepage:
- Size: 2.16 MB
- Stars: 41
- Watchers: 5
- Forks: 7
- Open Issues: 25
-
Metadata Files:
- Readme: README.rst
- License: LICENSE.txt
Awesome Lists containing this project
README
|Travis|_
.. |Travis| image:: https://api.travis-ci.org/dougalsutherland/skl-groups.png?branch=master
.. _Travis: https://travis-ci.org/dougalsutherland/skl-groups
skl-groups
==========
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 still in fairly early development.
The precursor package, `py-sdm `_,
is still somewhat easier to use for some tasks (though it has less functionality
and less documentation); skl-groups will hopefully match it in the next few weeks.
Feel free to get in touch (dsutherl@cs.cmu.edu) if you're interested.
Installation
------------
Full instructions are
`in the documentation `_,
but the short version is to do::
$ conda install -c dougal -c r skl-groups
if you use conda, or::
$ pip install skl-groups
if not. If you pip install and want to use the kNN divergence estimator,
you'll need to install either
`cyflann `_
or the regular pyflann bindings to FLANN,
and you'll want a version of FLANN with OpenMP support.
A much faster version of the kNN estimator is enabled by the
skl-groups-accel package, which you can get via::
$ pip install skl-groups-accel
It requires cyflann and a working C compiler with OpenMP support
(i.e. gcc, not clang).