{
    "info": {
        "author": "Olivier Grisel",
        "author_email": "olivier.grisel@ensta.org",
        "bugtrack_url": null,
        "classifiers": [
            "Development Status :: 3 - Alpha",
            "Intended Audience :: Developers",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved",
            "Operating System :: MacOS",
            "Operating System :: Microsoft :: Windows",
            "Operating System :: POSIX",
            "Operating System :: Unix",
            "Programming Language :: Python",
            "Topic :: Scientific/Engineering",
            "Topic :: Software Development"
        ],
        "description": "# Pyrallel - Parallel Data Analytics in Python\n\n**Overview**: experimental project to investigate distributed computation\npatterns for machine learning and other semi-interactive data analytics\ntasks.\n\n**Scope**:\n\n- focus on small to medium dataset that fits in memory on a small\n  (10+ nodes) to medium cluster (100+ nodes).\n\n- focus on small to medium data (with data locality when possible).\n\n- focus on CPU bound tasks (e.g. training Random Forests) while trying to\n  limit disk / network access to a minimum.\n\n- do not focus on HA / Fault Tolerance (yet).\n\n- do not try to invent new set of high level programming abstractions\n  (yet): use a low level programming model (IPython.parallel) to finely\n  control the cluster elements and messages transfered and help identify\n  what are the practical underlying constraints in distributed machine\n  learning setting.\n\n\n**Disclaimer**: the public API of this library will probably not be\nstable soon as the current goal of this project is to experiment.\n\n\n## Dependencies\n\nThe usual suspects: Python 2.7, NumPy, SciPy.\n\nFetch the development version (master branch) from:\n\n- https://github.com/ipython/ipython\n\n- https://github.com/scikit-learn/scikit-learn\n\nStarCluster `develop` branch and its `IPCluster` plugin is also required\nto easily startup a bunch of nodes with IPython.parallel setup.\n\n## Patterns currently under investigation\n\n- Asynchronous & randomized hyper-parameters search (a.k.a. Randomized Grid\n  Search) for machine learning models\n\n- Share numerical arrays efficiently over the nodes and make them\n  available to concurrently running Python processes without making\n  copies in memory using memory-mapped files.\n\n- Distributed Random Forests fitting.\n\n- Ensembling heterogeneous library models.\n\n- Parallel implementation of online averaged models using a MPI AllReduce, for\n  instance using MiniBatchKMeans on partitioned data.\n\n\nSee the content of the `examples/` folder for more details.\n\n\n## License\n\nSimplified BSD.\n\n\n## History\n\nThis project started at the [PyCon 2012 PyData\nsprint](http://wiki.ipython.org/PyCon12Sprint)\nas a set of proof of concept [IPython.parallel\nscripts](https://github.com/ogrisel/pycon-pydata-sprint).\n",
        "description_content_type": null,
        "docs_url": null,
        "download_url": "UNKNOWN",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "http://github.com/pydata/pyrallel",
        "keywords": null,
        "license": "MIT",
        "maintainer": null,
        "maintainer_email": null,
        "name": "pyrallel",
        "package_url": "https://pypi.org/project/pyrallel/",
        "platform": "UNKNOWN",
        "project_url": "https://pypi.org/project/pyrallel/",
        "project_urls": {
            "Download": "UNKNOWN",
            "Homepage": "http://github.com/pydata/pyrallel"
        },
        "release_url": "https://pypi.org/project/pyrallel/0.2.1/",
        "requires_dist": null,
        "requires_python": null,
        "summary": "Experimental tools for parallel machine learning",
        "version": "0.2.1"
    },
    "last_serial": 775398,
    "releases": {
        "0.2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "9cd094b4e2d7eed5511933a23fc43212",
                    "sha256": "f586c0d66b56038c32edd112e25f6d98e16c3df766abbdef0f044501e7ac8836"
                },
                "downloads": -1,
                "filename": "pyrallel-0.2.tar.gz",
                "has_sig": false,
                "md5_digest": "9cd094b4e2d7eed5511933a23fc43212",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 7841,
                "upload_time": "2013-06-13T14:46:34",
                "url": "https://files.pythonhosted.org/packages/34/7c/23761a09302544e3275cebc8d2930248e43efbe1f125fd8219807ed2691c/pyrallel-0.2.tar.gz"
            }
        ],
        "0.2.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "ffe1bed1bff178ec2f91e4155d7fe930",
                    "sha256": "b57cfdf7dfc14628d7c3b738e0e23d8750a5ffc1ac2aae83d2216cd1549b6b30"
                },
                "downloads": -1,
                "filename": "pyrallel-0.2.1.tar.gz",
                "has_sig": false,
                "md5_digest": "ffe1bed1bff178ec2f91e4155d7fe930",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 7853,
                "upload_time": "2013-06-20T14:47:34",
                "url": "https://files.pythonhosted.org/packages/ca/b8/42036676c89dcc92c90e74f8cb5ccc60bc2da860bef3115068806ef639ef/pyrallel-0.2.1.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "ffe1bed1bff178ec2f91e4155d7fe930",
                "sha256": "b57cfdf7dfc14628d7c3b738e0e23d8750a5ffc1ac2aae83d2216cd1549b6b30"
            },
            "downloads": -1,
            "filename": "pyrallel-0.2.1.tar.gz",
            "has_sig": false,
            "md5_digest": "ffe1bed1bff178ec2f91e4155d7fe930",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 7853,
            "upload_time": "2013-06-20T14:47:34",
            "url": "https://files.pythonhosted.org/packages/ca/b8/42036676c89dcc92c90e74f8cb5ccc60bc2da860bef3115068806ef639ef/pyrallel-0.2.1.tar.gz"
        }
    ]
}