{
    "info": {
        "author": "Nikolay Novik",
        "author_email": "nickolainovik@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "Development Status :: 2 - Pre-Alpha",
            "Intended Audience :: Developers",
            "License :: OSI Approved :: Apache Software License",
            "Operating System :: POSIX",
            "Programming Language :: Python :: 3",
            "Programming Language :: Python :: 3.6",
            "Programming Language :: Python :: 3.7"
        ],
        "description": "ibreakdown\n==========\n.. image:: https://travis-ci.com/jettify/ibreakdown.svg?branch=master\n    :target: https://travis-ci.com/jettify/ibreakdown\n.. image:: https://codecov.io/gh/jettify/ibreakdown/branch/master/graph/badge.svg\n    :target: https://codecov.io/gh/jettify/ibreakdown\n.. image:: https://img.shields.io/pypi/pyversions/ibreakdown.svg\n    :target: https://pypi.org/project/ibreakdown\n.. image:: https://img.shields.io/pypi/v/ibreakdown.svg\n    :target: https://pypi.python.org/pypi/ibreakdown\n\n\n**ibreakdown** is model agnostic predictions explainer with interactions support,\nlibrary can show contribution of each feature in your prediction value.\n\n**SHAP** or **LIME** consider only local additive feature attributions, when\n**ibreakdown** also evaluates local feature interactions.\n\nAlgorithm\n=========\n\nAlgorithm is based on ideas describe in paper *\"iBreakDown: Uncertainty of Model\nExplanations for Non-additive Predictive Models\"* https://arxiv.org/abs/1903.11420 and\nreference implementation in **R** (iBreakDown_)\n\nIntuition behind algorithm is following:\n\n  ::\n\n   The algorithm works in a similar spirit as SHAP or Break Down but is not\n   restricted to additive effects. The intuition is the following:\n\n   1. Calculate a single-step additive contribution for each feature.\n   2. Calculate a single-step contribution for every pair of features. Subtract additive contribution to assess the interaction specific contribution.\n   3. Order interaction effects and additive effects in a list that is used to determine sequential contributions.\n\n   This simple intuition may be generalized into higher order interactions.\n\nIn depth explanation can be found in algorithm authors free book:\n*Predictive Models: Explore, Explain, and Debug* https://pbiecek.github.io/PM_VEE/iBreakDown.html\n\n\nSimple example\n--------------\n\n.. code:: python\n\n    # model = RandomForestClassifier(...)\n    explainer = ClassificationExplainer(model)\n    classes = ['Deceased', 'Survived']\n    explainer.fit(X_train, columns, classes)\n    exp = explainer.explain(observation)\n    exp.print()\n\nPlease check full Titanic example here: https://github.com/jettify/ibreakdown/blob/master/examples/titanic.py\n\n.. code::\n\n   +------------------------------------+-----------------+--------------------+--------------------+\n   | Feature Name                       | Feature Value   |   Contrib:Deceased |   Contrib:Survived |\n   +------------------------------------+-----------------+--------------------+--------------------|\n   | intercept                          |                 |          0.613286  |          0.386714  |\n   | Sex                                | female          |         -0.305838  |          0.305838  |\n   | Pclass                             | 3               |          0.242448  |         -0.242448  |\n   | Fare                               | 7.7375          |         -0.119392  |          0.119392  |\n   | Siblings/Spouses Aboard            | 0               |         -0.0372811 |          0.0372811 |\n   | ('Age', 'Parents/Children Aboard') | [28.0 0]        |          0.0122196 |         -0.0122196 |\n   | PREDICTION                         |                 |          0.405443  |          0.594557  |\n   +------------------------------------+-----------------+--------------------+--------------------+\n\n\n\nFeatures\n========\n* Supports predictions explanations for classification and regression\n* Easy to use API.\n* Works with ``pandas`` and ``numpy``\n* Support interactions between features\n\n\nInstallation\n------------\nInstallation process is simple, just::\n\n    $ pip install ibreakdown\n\n\nRequirements\n------------\n\n* Python_ 3.6+\n* numpy_\n\n.. _Python: https://www.python.org\n.. _numpy: http://www.numpy.org/\n.. _iBreakDown: https://github.com/ModelOriented/iBreakDown\n.. _Shapley: https://en.wikipedia.org/wiki/Shapley_value\n\nCHANGES\n=======",
        "description_content_type": "",
        "docs_url": null,
        "download_url": "https://pypi.python.org/pypi/ibreakdown",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/jettify/ibreakdown",
        "keywords": "ibreakdown",
        "license": "Apache 2",
        "maintainer": "",
        "maintainer_email": "",
        "name": "ibreakdown",
        "package_url": "https://pypi.org/project/ibreakdown/",
        "platform": "POSIX",
        "project_url": "https://pypi.org/project/ibreakdown/",
        "project_urls": {
            "Download": "https://pypi.python.org/pypi/ibreakdown",
            "Homepage": "https://github.com/jettify/ibreakdown"
        },
        "release_url": "https://pypi.org/project/ibreakdown/0.0.1a4/",
        "requires_dist": null,
        "requires_python": "",
        "summary": "ibreakdown - model agnostic explanations with interactions",
        "version": "0.0.1a4",
        "yanked": false,
        "yanked_reason": null
    },
    "last_serial": 6038822,
    "releases": {
        "0.0.1a2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "8cbea48855fe1e837254fa3d3ca90801",
                    "sha256": "11a8fc74c4c479dc4891dc82b546eaef4199e8c957f858c3a6f9114d6f4ddd79"
                },
                "downloads": -1,
                "filename": "ibreakdown-0.0.1a2.tar.gz",
                "has_sig": false,
                "md5_digest": "8cbea48855fe1e837254fa3d3ca90801",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 7540,
                "upload_time": "2019-10-26T20:59:33",
                "upload_time_iso_8601": "2019-10-26T20:59:33.281728Z",
                "url": "https://files.pythonhosted.org/packages/d6/bb/ea7ff3ff604e72c16dc53c061a2933b943dd6b487689c343563fb7588d36/ibreakdown-0.0.1a2.tar.gz",
                "yanked": false,
                "yanked_reason": null
            }
        ],
        "0.0.1a3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "88fa70197b480a54f8e4bbccc3e926c2",
                    "sha256": "d28e292c3e14a96a27288b2de4a60411d5e5e6a09468d36549009846477f88f3"
                },
                "downloads": -1,
                "filename": "ibreakdown-0.0.1a3.tar.gz",
                "has_sig": false,
                "md5_digest": "88fa70197b480a54f8e4bbccc3e926c2",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 8666,
                "upload_time": "2019-10-27T22:12:35",
                "upload_time_iso_8601": "2019-10-27T22:12:35.463424Z",
                "url": "https://files.pythonhosted.org/packages/c2/33/1ebe296746a37d9d4072573c40cb7e1589e31d129a2614456421dab89063/ibreakdown-0.0.1a3.tar.gz",
                "yanked": false,
                "yanked_reason": null
            }
        ],
        "0.0.1a4": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "4c1c6af8f21bc9a484939edc2c97f7b2",
                    "sha256": "fd3cccd6ade5be548a6f6830c41a08208302461491df05941b444ad2b71b1c88"
                },
                "downloads": -1,
                "filename": "ibreakdown-0.0.1a4.tar.gz",
                "has_sig": false,
                "md5_digest": "4c1c6af8f21bc9a484939edc2c97f7b2",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 9231,
                "upload_time": "2019-10-28T00:35:06",
                "upload_time_iso_8601": "2019-10-28T00:35:06.637150Z",
                "url": "https://files.pythonhosted.org/packages/08/39/6a6cf178790750dc2dd099969de0fed5119e86344b8f04f72109303b7215/ibreakdown-0.0.1a4.tar.gz",
                "yanked": false,
                "yanked_reason": null
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "4c1c6af8f21bc9a484939edc2c97f7b2",
                "sha256": "fd3cccd6ade5be548a6f6830c41a08208302461491df05941b444ad2b71b1c88"
            },
            "downloads": -1,
            "filename": "ibreakdown-0.0.1a4.tar.gz",
            "has_sig": false,
            "md5_digest": "4c1c6af8f21bc9a484939edc2c97f7b2",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 9231,
            "upload_time": "2019-10-28T00:35:06",
            "upload_time_iso_8601": "2019-10-28T00:35:06.637150Z",
            "url": "https://files.pythonhosted.org/packages/08/39/6a6cf178790750dc2dd099969de0fed5119e86344b8f04f72109303b7215/ibreakdown-0.0.1a4.tar.gz",
            "yanked": false,
            "yanked_reason": null
        }
    ],
    "vulnerabilities": []
}