skrebate
A scikit-learn-compatible Python implementation of ReBATE, a suite of Relief-based feature selection algorithms for Machine Learning.
Science Score: 20.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
○codemeta.json file
-
○.zenodo.json file
-
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
✓Committers with academic emails
5 of 14 committers (35.7%) from academic institutions -
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (13.5%) to scientific vocabulary
Keywords
Repository
A scikit-learn-compatible Python implementation of ReBATE, a suite of Relief-based feature selection algorithms for Machine Learning.
Basic Info
- Host: GitHub
- Owner: EpistasisLab
- License: mit
- Language: Python
- Default Branch: master
- Homepage: https://EpistasisLab.github.io/scikit-rebate/
- Size: 12.9 MB
Statistics
- Stars: 420
- Watchers: 24
- Forks: 73
- Open Issues: 18
- Releases: 0
Topics
Metadata Files
README.md
Master status: 
Development status: 
Package information: 
scikit-rebate
This package includes a scikit-learn-compatible Python implementation of ReBATE, a suite of Relief-based feature selection algorithms for Machine Learning. These Relief-Based algorithms (RBAs) are designed for feature weighting/selection as part of a machine learning pipeline (supervised learning). Presently this includes the following core RBAs: ReliefF, SURF, SURF*, MultiSURF*, and MultiSURF. Additionally, an implementation of the iterative TuRF mechanism and VLSRelief is included. It is still under active development and we encourage you to check back on this repository regularly for updates.
These algorithms offer a computationally efficient way to perform feature selection that is sensitive to feature interactions as well as simple univariate associations, unlike most currently available filter-based feature selection methods. The main benefit of Relief algorithms is that they identify feature interactions without having to exhaustively check every pairwise interaction, thus taking significantly less time than exhaustive pairwise search.
Certain algorithms require user specified run parameters (e.g. ReliefF requires the user to specify some 'k' number of nearest neighbors).
Relief algorithms are commonly applied to genetic analyses, where epistasis (i.e., feature interactions) is common. However, the algorithms implemented in this package can be applied to almost any supervised classification data set and supports:
Feature sets that are discrete/categorical, continuous-valued or a mix of both
Data with missing values
Binary endpoints (i.e., classification)
Multi-class endpoints (i.e., classification)
Continuous endpoints (i.e., regression)
Built into this code, is a strategy to 'automatically' detect from the loaded data, these relevant characteristics.
Of our two initial ReBATE software releases, this scikit-learn compatible version primarily focuses on ease of incorporation into a scikit learn analysis pipeline. This code is most appropriate for scikit-learn users, Windows operating system users, beginners, or those looking for the most recent ReBATE developments.
An alternative 'stand-alone' version of ReBATE is also available that focuses on improving run-time with the use of Cython for optimization. This implementation also outputs feature names and associated feature scores as a text file by default.
License
Please see the repository license for the licensing and usage information for scikit-rebate.
Generally, we have licensed scikit-rebate to make it as widely usable as possible.
Installation
scikit-rebate is built on top of the following existing Python packages:
NumPy
SciPy
scikit-learn
All of the necessary Python packages can be installed via the Anaconda Python distribution, which we strongly recommend that you use. We also strongly recommend that you use Python 3 over Python 2 if you're given the choice.
NumPy, SciPy, and scikit-learn can be installed in Anaconda via the command:
conda install numpy scipy scikit-learn
Once the prerequisites are installed, you should be able to install scikit-rebate with a pip command:
pip install skrebate
Please file a new issue if you run into installation problems.
Usage
We have designed the Relief algorithms to be integrated directly into scikit-learn machine learning workflows. For example, the ReliefF algorithm can be used as a feature selection step in a scikit-learn pipeline as follows.
```python import pandas as pd import numpy as np from sklearn.pipeline import makepipeline from skrebate import ReliefF from sklearn.ensemble import RandomForestClassifier from sklearn.modelselection import crossvalscore
geneticdata = pd.readcsv('https://github.com/EpistasisLab/scikit-rebate/raw/master/data/' 'GAMETESEpistasis2-Way20atts0.4HEDM-11.tsv.gz', sep='\t', compression='gzip')
features, labels = geneticdata.drop('class', axis=1).values, geneticdata['class'].values
clf = makepipeline(ReliefF(nfeaturestoselect=2, nneighbors=100), RandomForestClassifier(nestimators=100))
print(np.mean(crossvalscore(clf, features, labels)))
0.795 ```
For more information on the Relief algorithms available in this package and how to use them, please refer to our usage documentation.
Contributing to scikit-rebate
We welcome you to check the existing issues for bugs or enhancements to work on. If you have an idea for an extension to scikit-rebate, please file a new issue so we can discuss it.
Please refer to our contribution guidelines prior to working on a new feature or bug fix.
Citing scikit-rebate
If you use scikit-rebate in a scientific publication, please consider citing the following paper:
Ryan J. Urbanowicz, Randal S. Olson, Peter Schmitt, Melissa Meeker, Jason H. Moore (2017). Benchmarking Relief-Based Feature Selection Methods. arXiv preprint, under review.
BibTeX entry:
bibtex
@misc{Urbanowicz2017Benchmarking,
author = {Urbanowicz, Ryan J. and Olson, Randal S. and Schmitt, Peter and Meeker, Melissa and Moore, Jason H.},
title = {Benchmarking Relief-Based Feature Selection Methods},
year = {2017},
howpublished = {arXiv e-print. https://arxiv.org/abs/1711.08477},
}
Owner
- Name: Epistasis Lab at Cedars Sinai
- Login: EpistasisLab
- Kind: organization
- Email: jason.moore@csmc.edu
- Location: United States of America
- Website: http://epistasis.org
- Repositories: 63
- Profile: https://github.com/EpistasisLab
Prof. Jason H. Moore's research lab at Cedars Sinai
GitHub Events
Total
- Watch event: 12
- Fork event: 1
Last Year
- Watch event: 12
- Fork event: 1
Committers
Last synced: 9 months ago
Top Committers
| Name | Commits | |
|---|---|---|
| Randy Olson | r****o@r****m | 103 |
| Ryan Urbanowicz | r****b@u****u | 35 |
| Alex | a****9@g****m | 24 |
| sauravbose | s****3@g****m | 15 |
| Alex Xu | a****u@A****l | 12 |
| weixuanfu2016 | w****f@u****u | 10 |
| UPENN | U****N@U****u | 6 |
| robertfrankzhang | r****0@g****m | 5 |
| Pete Schmitt | p****t@i****m | 5 |
| Weixuan Fu | w****f@m****u | 5 |
| Alex Xu | a****u@a****t | 2 |
| Weixuan Fu | w****f@p****u | 1 |
| Nicholas Bollweg | n****g@g****m | 1 |
| = | = | 1 |
Committer Domains (Top 20 + Academic)
Issues and Pull Requests
Last synced: 7 months ago
All Time
- Total issues: 34
- Total pull requests: 48
- Average time to close issues: about 1 month
- Average time to close pull requests: 15 days
- Total issue authors: 24
- Total pull request authors: 12
- Average comments per issue: 1.71
- Average comments per pull request: 0.69
- Merged pull requests: 35
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 0
- Pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Issue authors: 0
- Pull request authors: 0
- Average comments per issue: 0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
- rhiever (9)
- brunofacca (2)
- ryanurbs (2)
- jgoecks (1)
- garciaev (1)
- swzCuroverse (1)
- greenspray9 (1)
- gjj07 (1)
- moumitam28 (1)
- weixuanfu (1)
- ghost (1)
- l0o0 (1)
- arnaud-fossop (1)
- k8iechen (1)
- raomidi (1)
Pull Request Authors
- rhiever (17)
- alexmxu (9)
- ryanurbs (7)
- weixuanfu (4)
- sauravbose (3)
- garciaev (2)
- bollwyvl (1)
- bukson (1)
- robertfrankzhang (1)
- jckkvs (1)
- aadu (1)
- CaptainKanuk (1)
Top Labels
Issue Labels
Pull Request Labels
Packages
- Total packages: 2
-
Total downloads:
- pypi 8,996 last-month
- Total docker downloads: 220
-
Total dependent packages: 10
(may contain duplicates) -
Total dependent repositories: 52
(may contain duplicates) - Total versions: 18
- Total maintainers: 4
pypi.org: skrebate
Relief-based feature selection algorithms
- Homepage: https://github.com/EpistasisLab/scikit-rebate
- Documentation: https://skrebate.readthedocs.io/
- License: License :: OSI Approved :: MIT License
-
Latest release: 0.3.4
published almost 9 years ago
Rankings
conda-forge.org: skrebate
These algorithms excel at identifying features that are predictive of the outcome in supervised learning problems, and are especially good at identifying feature interactions that are normally overlooked by standard feature selection methods.
- Homepage: https://epistasislab.github.io/scikit-rebate
- License: MIT
-
Latest release: 0.5
published over 3 years ago
Rankings
Dependencies
- numpy *
- scikit-learn *
- scipy *