{ "info": { "author": "Ilya Razenshteyn, Ludwig Schmidt", "author_email": "falconn.lib@gmail.com", "bugtrack_url": null, "classifiers": [], "description": "|Build Status|\n\nFALCONN - FAst Lookups of Cosine and Other Nearest Neighbors\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nFALCONN is a library with algorithms for the nearest neighbor search\nproblem. The algorithms in FALCONN are based on `Locality-Sensitive\nHashing `__\n(LSH), which is a popular class of methods for nearest neighbor search\nin high-dimensional spaces. The goal of FALCONN is to provide very\nefficient and well-tested implementations of LSH-based data structures.\n\nCurrently, FALCONN supports two LSH families for the `cosine\nsimilarity `__:\nhyperplane LSH and cross polytope LSH. Both hash families are\nimplemented with multi-probe LSH in order to minimize memory usage.\nMoreover, FALCONN is optimized for both dense and sparse data. Despite\nbeing designed for the cosine similarity, FALCONN can often be used for\nnearest neighbor search under the Euclidean distance or a maximum inner\nproduct search.\n\nFALCONN is written in C++ and consists of several modular core classes\nwith a convenient wrapper around them. Many mathematical operations in\nFALCONN are vectorized through the\n`Eigen `__ and\n`FFHT `__ libraries. The core\nclasses of FALCONN rely on\n`templates `__ in\norder to avoid runtime overhead.\n\nHow to use FALCONN\n~~~~~~~~~~~~~~~~~~\n\nWe provide a C++ interface for FALCONN as well as a\n`Python `__ wrapper (that uses\n`NumPy `__). In the future, we plan to support\nmore programming languages such as `Julia `__.\nFor C++, FALCONN is a header-only library and has no dependencies\nbesides Eigen (which is also header-only), so FALCONN is easy to set up.\nFor further details, please see our\n`documentation `__.\n\nHow fast is FALCONN?\n~~~~~~~~~~~~~~~~~~~~\n\nOn data sets with about 1 million points in around 100 dimensions,\nFALCONN typically requires a few milliseconds per query (running on a\nreasonably modern desktop CPU).\n\nFor more detailed results, see\n`ann-benchmarks `__ of `Erik\nBernhardsson `__. Let us point out that FALCONN\nis especially competitive, when the RAM budget is quite restrictive,\nwhich is not the regime the above benchmarks use.\n\nQuestions\n~~~~~~~~~\n\nMaybe your question is already answered in our `Frequently Asked\nQuestions `__. If you\nhave additional questions about using FALCONN, we would be happy to\nhelp. Please send an email to falconn.lib@gmail.com.\n\nAuthors\n~~~~~~~\n\nFALCONN is mainly developed by `Ilya\nRazenshteyn `__ and `Ludwig\nSchmidt `__. FALCONN has grown out\nof a `research\nproject `__\nwith our collaborators `Alexandr\nAndoni `__, `Piotr\nIndyk `__, and `Thijs\nLaarhoven `__.\n\nMany of the ideas used in FALCONN were proposed in research papers over\nthe past 20 years (see the\n`documentation `__).\n\nIf you want to cite FALCONN in a publication, here is the bibliographic\ninformation of our research paper\n`(bibtex) `__:\n\n `Practical and Optimal LSH for Angular\n Distance `__\n Alexandr Andoni, Piotr Indyk, Thijs Laarhoven, Ilya Razenshteyn,\n Ludwig Schmidt NIPS 2015\n\nLicense\n~~~~~~~\n\nFALCONN is available under the `MIT\nLicense `__ (see LICENSE.txt). Note\nthat the third-party libraries in the ``external/`` folder are\ndistributed under other open source licenses. The Eigen library is\nlicensed under the `MPL2 `__.\nThe googletest and googlemock libraries are licensed under the `BSD\n3-Clause License `__. The\npybind11 library is licensed under a `BSD-style\nlicense `__.\n\n.. |Build Status| image:: https://travis-ci.org/FALCONN-LIB/FALCONN.svg?branch=master\n :target: https://travis-ci.org/FALCONN-LIB/FALCONN\n", "description_content_type": null, "docs_url": null, "download_url": "", "downloads": { "last_day": -1, "last_month": -1, "last_week": -1 }, "home_page": "http://falconn-lib.org/", "keywords": "nearest neighbor search similarity lsh locality-sensitive hashing cosine distance euclidean", "license": "MIT", "maintainer": "", "maintainer_email": "", "name": "FALCONN", "package_url": "https://pypi.org/project/FALCONN/", "platform": "", "project_url": "https://pypi.org/project/FALCONN/", "project_urls": { "Homepage": "http://falconn-lib.org/" }, "release_url": "https://pypi.org/project/FALCONN/1.3.1/", "requires_dist": null, "requires_python": "", "summary": "A library for similarity search over high-dimensional data based on Locality-Sensitive Hashing (LSH)", "version": "1.3.1" }, "last_serial": 3174824, "releases": { "1.1": [ { "comment_text": "", "digests": { "md5": "29c69f3e86a52fdacca5f69968da648f", "sha256": "eaeb7210beccb039ac2cc2aa955d7b81ea31cdd8f3c30cbb7d2337dbfc83bdd4" }, "downloads": -1, "filename": "FALCONN-1.1.tar.gz", "has_sig": false, "md5_digest": "29c69f3e86a52fdacca5f69968da648f", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 768281, "upload_time": "2016-01-15T23:50:01", "url": "https://files.pythonhosted.org/packages/6c/ee/fe8ff1100d959fe222118c1fcc47340481b9dd9dcbc17cae31915574d86d/FALCONN-1.1.tar.gz" } ], "1.2": [ { "comment_text": "", "digests": { "md5": "6367970ce9e3669404b825a1642241a7", "sha256": "9ae9d098ee5d3a042a2d83f36dd741048da3f67e7f79881d2edb31c141946879" }, "downloads": -1, "filename": "FALCONN-1.2.tar.gz", "has_sig": false, "md5_digest": "6367970ce9e3669404b825a1642241a7", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1106652, "upload_time": "2016-03-17T19:05:51", "url": "https://files.pythonhosted.org/packages/b5/a6/0b3c5920e1d01c291311fb2fb235193801eac3f9070e1d4a0ff4ef5882ed/FALCONN-1.2.tar.gz" } ], "1.2.1": [ { "comment_text": "", "digests": { "md5": "c8355fdc6cf69bd6e87d17494fd8d71e", "sha256": "428208986a5b3566e242d20dc2987d126aa3a62d11d65812b93b5180342ac877" }, "downloads": -1, "filename": "FALCONN-1.2.1.tar.gz", "has_sig": false, "md5_digest": "c8355fdc6cf69bd6e87d17494fd8d71e", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 876862, "upload_time": "2016-10-14T01:35:27", "url": "https://files.pythonhosted.org/packages/d3/91/6a1fa5431542186b67e58ddc2d5e4e60140a79934baa0091e192095b8b2a/FALCONN-1.2.1.tar.gz" } ], "1.2.2": [ { "comment_text": "", "digests": { "md5": "590bc109481fd465d1d6333baec7045a", "sha256": "22ec3f86f342f816c0f0c71d29684feed6a7ef887fc4208c3ab43454675390b6" }, "downloads": -1, "filename": "FALCONN-1.2.2.tar.gz", "has_sig": false, "md5_digest": "590bc109481fd465d1d6333baec7045a", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 878759, "upload_time": "2016-10-22T20:46:09", "url": "https://files.pythonhosted.org/packages/82/2a/e07f1a0405de48332a58e671e697ec7f78d26b26bf966b78b2f09aed8ee5/FALCONN-1.2.2.tar.gz" } ], "1.3.0": [ { "comment_text": "", "digests": { "md5": "69fa45c5bcb83c57244d992cbcdd5232", "sha256": "c3e18531b906769914f28c634d54d4c3a1262c57afa41e5bb33eaaf9043aa610" }, "downloads": -1, "filename": "FALCONN-1.3.0.tar.gz", "has_sig": false, "md5_digest": "69fa45c5bcb83c57244d992cbcdd5232", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1381970, "upload_time": "2017-09-14T21:43:19", "url": "https://files.pythonhosted.org/packages/7d/f6/59757f289f83ccc717951b6291cba28dae2ef6fde750b371e711f432e253/FALCONN-1.3.0.tar.gz" } ], "1.3.1": [ { "comment_text": "", "digests": { "md5": "ec7c6687d2c54e56b3ec78a4ebc9cc46", "sha256": "00cf778a97e9f2d9b501e86c16ff6f39528d70142302028675b3947e4a565eef" }, "downloads": -1, "filename": "FALCONN-1.3.1.tar.gz", "has_sig": false, "md5_digest": "ec7c6687d2c54e56b3ec78a4ebc9cc46", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1382330, "upload_time": "2017-09-14T22:06:19", "url": "https://files.pythonhosted.org/packages/96/b8/0d2c629d59398a7b3ed8726ce049abf6746bbf09d1ad15878d4fcf8048a6/FALCONN-1.3.1.tar.gz" } ] }, "urls": [ { "comment_text": "", "digests": { "md5": "ec7c6687d2c54e56b3ec78a4ebc9cc46", "sha256": "00cf778a97e9f2d9b501e86c16ff6f39528d70142302028675b3947e4a565eef" }, "downloads": -1, "filename": "FALCONN-1.3.1.tar.gz", "has_sig": false, "md5_digest": "ec7c6687d2c54e56b3ec78a4ebc9cc46", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1382330, "upload_time": "2017-09-14T22:06:19", "url": "https://files.pythonhosted.org/packages/96/b8/0d2c629d59398a7b3ed8726ce049abf6746bbf09d1ad15878d4fcf8048a6/FALCONN-1.3.1.tar.gz" } ] }