# choix
[](https://github.com/lucasmaystre/choix/actions/workflows/build.yml)
[](https://codecov.io/gh/lucasmaystre/choix)
[](http://choix.lum.li/en/latest/?badge=latest)
`choix` is a Python library that provides inference algorithms for models based on Luce\'s choice axiom. These probabilistic models can be used to explain and predict outcomes of comparisons between items.
- **Pairwise comparisons**: when the data consists of comparisons between two items, the model variant is usually referred to as the *Bradley-Terry* model. It is closely related to the Elo rating system used to rank chess players.
- **Partial rankings**: when the data consists of rankings over (a subset of) the items, the model variant is usually referred to as the *Plackett-Luce* model.
- **Top-1 lists**: another variation of the model arises when the data consists of discrete choices, i.e., we observe the selection of one item out of a subset of items.
- **Choices in a network**: when the data consists of counts of the number of visits to each node in a network, the model is known as the *Network Choice Model*.
`choix` makes it easy to infer model parameters from these different types of data, using a variety of algorithms:
- Luce Spectral Ranking
- Minorization-Maximization
- Rank Centrality
- Approximate Bayesian inference with expectation propagation
## Getting started
To install the latest release directly from PyPI, simply type:
pip install choix
To get started, you might want to explore one of these notebooks:
- [Introduction using pairwise-comparison data](https://github.com/lucasmaystre/choix/blob/master/notebooks/intro-pairwise.ipynb)
- [Case study: analyzing the GIFGIF dataset](https://github.com/lucasmaystre/choix/blob/master/notebooks/gifgif-dataset.ipynb)
- [Using ChoiceRank to understand traffic on a network](https://github.com/lucasmaystre/choix/blob/master/notebooks/choicerank-tutorial.ipynb)
- [Approximate Bayesian inference using EP](https://github.com/lucasmaystre/choix/blob/master/notebooks/ep-example.ipynb)
You can also find more information on the [official documentation](http://choix.lum.li/en/latest/). In particular, the [API reference](http://choix.lum.li/en/latest/api.html) contains a good summary of the library's features.
## References
- Hossein Azari Soufiani, William Z. Chen, David C. Parkes, and Lirong Xia, [Generalized Method-of-Moments for Rank Aggregation](https://papers.nips.cc/paper/4997-generalized-method-of-moments-for-rank-aggregation.pdf), NIPS 2013
- François Caron and Arnaud Doucet. [Efficient Bayesian Inference for Generalized Bradley-Terry models](https://hal.inria.fr/inria-00533638/document). Journal of Computational and Graphical Statistics, 21(1):174-196, 2012.
- Wei Chu and Zoubin Ghahramani, [Extensions of Gaussian processes for ranking: semi-supervised and active learning](http://www.gatsby.ucl.ac.uk/~chuwei/paper/gprl.pdf), NIPS 2005 Workshop on Learning to Rank.
- David R. Hunter. [MM algorithms for generalized Bradley-Terry models](http://sites.stat.psu.edu/~dhunter/papers/bt.pdf), The Annals of Statistics 32(1):384-406, 2004.
- Ravi Kumar, Andrew Tomkins, Sergei Vassilvitskii and Erik Vee,
[Inverting a Steady-State](http://theory.stanford.edu/~sergei/papers/wsdm15-cset.pdf), WSDM 2015.
- Lucas Maystre and Matthias Grossglauser, [Fast and Accurate Inference of Plackett-Luce Models](https://infoscience.epfl.ch/record/213486/files/fastinference.pdf), NIPS, 2015.
- Lucas Maystre and M. Grossglauser, [ChoiceRank: Identifying Preferences from Node Traffic in Networks](https://infoscience.epfl.ch/record/229164/files/choicerank.pdf), ICML 2017.
- Sahand Negahban, Sewoong Oh, and Devavrat Shah, [Iterative Ranking from Pair-wise Comparison](https://papers.nips.cc/paper/4701-iterative-ranking-from-pair-wise-comparisons.pdf), NIPS 2012.
Raw data
{
"_id": null,
"home_page": null,
"name": "choix",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.10",
"maintainer_email": null,
"keywords": "statistics, ml, bradley-terry, plackett-luce, choice, comparison, ranking",
"author": "Lucas Maystre",
"author_email": "Lucas Maystre <lucas@maystre.ch>",
"download_url": "https://files.pythonhosted.org/packages/cc/83/9c9b9aa651e16813ca893b70f4384f2e9c369eafd5922f4ddaf32cde2e9e/choix-0.4.1.tar.gz",
"platform": null,
"description": "# choix\n\n[](https://github.com/lucasmaystre/choix/actions/workflows/build.yml)\n[](https://codecov.io/gh/lucasmaystre/choix)\n[](http://choix.lum.li/en/latest/?badge=latest)\n\n`choix` is a Python library that provides inference algorithms for models based on Luce\\'s choice axiom. These probabilistic models can be used to explain and predict outcomes of comparisons between items.\n\n- **Pairwise comparisons**: when the data consists of comparisons between two items, the model variant is usually referred to as the *Bradley-Terry* model. It is closely related to the Elo rating system used to rank chess players.\n- **Partial rankings**: when the data consists of rankings over (a subset of) the items, the model variant is usually referred to as the *Plackett-Luce* model.\n- **Top-1 lists**: another variation of the model arises when the data consists of discrete choices, i.e., we observe the selection of one item out of a subset of items.\n- **Choices in a network**: when the data consists of counts of the number of visits to each node in a network, the model is known as the *Network Choice Model*.\n\n`choix` makes it easy to infer model parameters from these different types of data, using a variety of algorithms:\n\n- Luce Spectral Ranking\n- Minorization-Maximization\n- Rank Centrality\n- Approximate Bayesian inference with expectation propagation\n\n## Getting started\n\nTo install the latest release directly from PyPI, simply type:\n\n pip install choix\n\nTo get started, you might want to explore one of these notebooks:\n\n- [Introduction using pairwise-comparison data](https://github.com/lucasmaystre/choix/blob/master/notebooks/intro-pairwise.ipynb)\n- [Case study: analyzing the GIFGIF dataset](https://github.com/lucasmaystre/choix/blob/master/notebooks/gifgif-dataset.ipynb)\n- [Using ChoiceRank to understand traffic on a network](https://github.com/lucasmaystre/choix/blob/master/notebooks/choicerank-tutorial.ipynb)\n- [Approximate Bayesian inference using EP](https://github.com/lucasmaystre/choix/blob/master/notebooks/ep-example.ipynb)\n\nYou can also find more information on the [official documentation](http://choix.lum.li/en/latest/). In particular, the [API reference](http://choix.lum.li/en/latest/api.html) contains a good summary of the library's features.\n\n## References\n\n- Hossein Azari Soufiani, William Z. Chen, David C. Parkes, and Lirong Xia, [Generalized Method-of-Moments for Rank Aggregation](https://papers.nips.cc/paper/4997-generalized-method-of-moments-for-rank-aggregation.pdf), NIPS 2013\n- Fran\u00e7ois Caron and Arnaud Doucet. [Efficient Bayesian Inference for Generalized Bradley-Terry models](https://hal.inria.fr/inria-00533638/document). Journal of Computational and Graphical Statistics, 21(1):174-196, 2012.\n- Wei Chu and Zoubin Ghahramani, [Extensions of Gaussian processes for ranking: semi-supervised and active learning](http://www.gatsby.ucl.ac.uk/~chuwei/paper/gprl.pdf), NIPS 2005 Workshop on Learning to Rank.\n- David R. Hunter. [MM algorithms for generalized Bradley-Terry models](http://sites.stat.psu.edu/~dhunter/papers/bt.pdf), The Annals of Statistics 32(1):384-406, 2004.\n- Ravi Kumar, Andrew Tomkins, Sergei Vassilvitskii and Erik Vee,\n [Inverting a Steady-State](http://theory.stanford.edu/~sergei/papers/wsdm15-cset.pdf), WSDM 2015.\n- Lucas Maystre and Matthias Grossglauser, [Fast and Accurate Inference of Plackett-Luce Models](https://infoscience.epfl.ch/record/213486/files/fastinference.pdf), NIPS, 2015.\n- Lucas Maystre and M. Grossglauser, [ChoiceRank: Identifying Preferences from Node Traffic in Networks](https://infoscience.epfl.ch/record/229164/files/choicerank.pdf), ICML 2017.\n- Sahand Negahban, Sewoong Oh, and Devavrat Shah, [Iterative Ranking from Pair-wise Comparison](https://papers.nips.cc/paper/4701-iterative-ranking-from-pair-wise-comparisons.pdf), NIPS 2012.\n",
"bugtrack_url": null,
"license": null,
"summary": "Inference algorithms for models based on Luce's choice axiom.",
"version": "0.4.1",
"project_urls": {
"Documentation": "https://choix.lum.li/en/latest/",
"Homepage": "https://github.com/lucasmaystre/choix",
"Issues": "https://github.com/lucasmaystre/choix/issues",
"Repository": "https://github.com/lucasmaystre/choix.git"
},
"split_keywords": [
"statistics",
" ml",
" bradley-terry",
" plackett-luce",
" choice",
" comparison",
" ranking"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "81f18ce3fa2fe1649c93ab7e177c3b11b9fd98c66180af3a365cfe7bd0118806",
"md5": "bbf1529b1e525fd5a27dbf0f001926e1",
"sha256": "15fe145ffc194dc6b4ba5a44ef24667fa0107578fb5b36723ab0b5e87eeae29f"
},
"downloads": -1,
"filename": "choix-0.4.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "bbf1529b1e525fd5a27dbf0f001926e1",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.10",
"size": 20575,
"upload_time": "2025-08-10T20:19:44",
"upload_time_iso_8601": "2025-08-10T20:19:44.255036Z",
"url": "https://files.pythonhosted.org/packages/81/f1/8ce3fa2fe1649c93ab7e177c3b11b9fd98c66180af3a365cfe7bd0118806/choix-0.4.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "cc839c9b9aa651e16813ca893b70f4384f2e9c369eafd5922f4ddaf32cde2e9e",
"md5": "1962dfa59d64337b9857cff053d9c7d3",
"sha256": "7c85d13845400932508c45a6a6a3e7a44f637969ffd506e1c7f5db07de1560b4"
},
"downloads": -1,
"filename": "choix-0.4.1.tar.gz",
"has_sig": false,
"md5_digest": "1962dfa59d64337b9857cff053d9c7d3",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.10",
"size": 112425,
"upload_time": "2025-08-10T20:19:45",
"upload_time_iso_8601": "2025-08-10T20:19:45.346610Z",
"url": "https://files.pythonhosted.org/packages/cc/83/9c9b9aa651e16813ca893b70f4384f2e9c369eafd5922f4ddaf32cde2e9e/choix-0.4.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-10 20:19:45",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "lucasmaystre",
"github_project": "choix",
"travis_ci": false,
"coveralls": false,
"github_actions": true,
"lcname": "choix"
}
JSON
