{
    "info": {
        "author": "Gregory Giecold",
        "author_email": "ggiecold@jimmy.harvard.edu",
        "bugtrack_url": null,
        "classifiers": [
            "Development Status :: 4 - Beta",
            "Environment :: Console",
            "Intended Audience :: Developers",
            "Intended Audience :: End Users/Desktop",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved :: MIT License",
            "Natural Language :: English",
            "Operating System :: MacOS :: MacOS X",
            "Operating System :: POSIX",
            "Programming Language :: Python :: 2.7",
            "Topic :: Scientific/Engineering",
            "Topic :: Scientific/Engineering :: Mathematics",
            "Topic :: Scientific/Engineering :: Visualization"
        ],
        "description": "Concurrent\\_AP\r\n==============\r\n\r\nOverview\r\n--------\r\n\r\nA scalable and concurrent programming implementation of Affinity\r\nPropagation clustering.\r\n\r\nAffinity Propagation is a clustering algorithm based on passing messages\r\nbetween data-points.\r\n\r\nStoring and updating matrices of 'affinities', 'responsibilities' and\r\n'similarities' between samples can be memory-intensive. We address this\r\nissue through the use of an HDF5 data structure, allowing Affinity\r\nPropagation clustering of arbitrary large data-sets, where other Python\r\nimplementations would return a MemoryError on most machines.\r\n\r\nWe also significantly speed up the computations by splitting them up\r\nacross subprocesses, thereby taking full advantage of the resources of\r\nmulti-core processors and bypassing the Global Interpreter Lock of the\r\nstandard Python interpreter, CPython.\r\n\r\nInstallation and Requirements\r\n-----------------------------\r\n\r\nConcurrent\\_AP requires Python 2.7 along with the following packages and\r\na few modules from the Standard Python Library:\r\n\r\n-  NumPy >= 1.9\r\n\r\n-  psutil\r\n\r\n-  PyTables\r\n\r\n-  scikit-learn\r\n\r\n-  setuptools\r\n\r\nIt is suggested that you check that the required dependencies are installed, although the ``pip`` command below should do this automatically for you. You can indeed most conveniently download Concurrent_AP from the official Python Package Index (PyPI) as follows:\r\n\r\n-  open a terminal window;\r\n\r\n-  type in the command ``pip install Concurrent_AP``.\r\n\r\nThe code herewith has been tested on Fedora, OS X and Ubuntu and should\r\nwork fine on any other member of the Unix-like family of operating systems.\r\n\r\nUsage and Command Line Options\r\n------------------------------\r\n\r\nSee the docstrings associated to each function of the Concurrent\\_AP\r\nmodule for more information and an understanding of how different tasks\r\nare organized and shared among subprocesses.\r\n\r\nUsage: ``Concurrent_AP [options] file_name``, where ``file_name``\r\ndenotes the path where the data to be processed by Affinity Propagation\r\nclustering is held. The data must consist in tab-separated rows of samples, each column corresponding to a particular feature.\r\n\r\n-  ``-c`` or ``--convergence``: specify the number of iterations without\r\n   change in the number of clusters that signals convergence (defaults\r\n   to 15);\r\n\r\n-  ``-d`` or ``--damping``: the damping parameter of Affinity\r\n   Propagation (defaults to 0.5);\r\n\r\n-  ``-f`` or ``--file``: option to specify the file name or file handle\r\n   of the hierarchical data format where the matrices involved in\r\n   Affinity Propagation clustering will be stored (defaults to a\r\n   temporary file);\r\n\r\n-  ``-i`` or ``--iterations``: maximum number of message-passing\r\n   iterations (defaults to 200);\r\n\r\n-  ``-m`` or ``--multiprocessing``: the number of processes to use;\r\n\r\n-  ``-p`` or ``--preference``: the preference parameter of Affinity\r\n   Propagation (if not specified, will be determined as the median of\r\n   the distribution of pairwise L2 Euclidean distances between samples);\r\n\r\n-  ``-s`` or ``--similarities``: determine if a similarity matrix has\r\n   been pre-computed and stored in the HDF5 data structure accessible at\r\n   the location specified through the command line option ``-f`` or\r\n   ``--file`` (see above);\r\n\r\n-  ``-v`` or ``--verbose``: whether to be verbose.\r\n\r\nDemo of Concurrent\\_AP\r\n----------------------\r\n\r\nThe following few lines illustrate the use of Concurrent\\_AP on the\r\n'Iris data-set' from the UCI Machine Learning Repository. While the\r\nnumber of samples is here way too small for the benefits of the present\r\nmulti-tasking implementation and the use of an HDF5 data structure to\r\ncome fully into play, this data-set has the advantage of being\r\nwell-known and prone to a quick comparison with scikit-learn's version\r\nof Affinity Propagation clustering.\r\n\r\n-  In a Python interpreter console, enter the following few lines, whose\r\n   purpose is to create a file containing the Iris data-set that will be\r\n   later subjected to Affinity Propagation clustering via\r\n   Concurrent\\_AP:\r\n\r\n::\r\n\r\n    >>> import numpy as np\r\n    >>> from sklearn import datasets\r\n\r\n    >>> iris = datasets.load_iris()\r\n    >>> data = iris.data\r\n    >>> with open('./iris_data.txt', 'w') as f:\r\n           np.savetxt(f, data, fmt = '%.4f', delimiter = '\\t')\r\n\r\n-  Open a terminal window.\r\n-  Type in ``Concurrent_AP --preference 5.47 --v iris_data.txt`` or\r\n   simply ``Concurrent_AP iris_data.txt``.\r\n\r\nThe latter will automatically compute a preference parameter from the\r\ndata-set.\r\n\r\nWhen the rounds of message-passing among data-points have completed, a\r\nfolder containing a file of cluster labels and a file of cluster centers\r\nindices both in tab-separated format is created in your current working\r\ndirectory.\r\n\r\nReference\r\n---------\r\n\r\nBrendan J. Frey and Delbert Dueck. \"Clustering by Passing Messages\r\nbetween Data Points\", Science Feb. 2007",
        "description_content_type": null,
        "docs_url": null,
        "download_url": "https://github.com/GGiecold/Concurrent_AP",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/GGiecold/Concurrent_AP",
        "keywords": "parallel multiprocessing machine-learning concurrent clustering",
        "license": "MIT License",
        "maintainer": "",
        "maintainer_email": "",
        "name": "Concurrent_AP",
        "package_url": "https://pypi.org/project/Concurrent_AP/",
        "platform": "Any",
        "project_url": "https://pypi.org/project/Concurrent_AP/",
        "project_urls": {
            "Download": "https://github.com/GGiecold/Concurrent_AP",
            "Homepage": "https://github.com/GGiecold/Concurrent_AP"
        },
        "release_url": "https://pypi.org/project/Concurrent_AP/1.4/",
        "requires_dist": null,
        "requires_python": null,
        "summary": "Scalable and parallel programming implementation of Affinity Propagation clustering",
        "version": "1.4"
    },
    "last_serial": 2001344,
    "releases": {
        "1.2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "9e7abc7693e2bb5899d7b96760733aec",
                    "sha256": "52ee3b36c7cd8fd38d74ce08a48e3d1426bbb1a4d50671721bc37b0f8364ee90"
                },
                "downloads": -1,
                "filename": "Concurrent_AP-1.2.tar.gz",
                "has_sig": false,
                "md5_digest": "9e7abc7693e2bb5899d7b96760733aec",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 14033,
                "upload_time": "2015-11-23T04:27:04",
                "url": "https://files.pythonhosted.org/packages/de/f8/89932f731710eda22a55168ad409514c539776a33ce4faef75d25b9170d5/Concurrent_AP-1.2.tar.gz"
            }
        ],
        "1.3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "6b8f424bfcc093f02e5d5fe8415de237",
                    "sha256": "aee3a6e5f35a6d18492e4c33543b83ceeafb4d3ad913df70803dcae46a3e9468"
                },
                "downloads": -1,
                "filename": "Concurrent_AP-1.3.tar.gz",
                "has_sig": false,
                "md5_digest": "6b8f424bfcc093f02e5d5fe8415de237",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 14098,
                "upload_time": "2015-12-07T16:32:37",
                "url": "https://files.pythonhosted.org/packages/57/e6/14a30f14dda029cbc8fac6c590fca56179d450ef86d6429666a8ce7c82c8/Concurrent_AP-1.3.tar.gz"
            }
        ],
        "1.4": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "2bb16a3e49dca04f2613977fcefe2104",
                    "sha256": "d6f4b5e82be60d7d4fa9f677e4f823731c5baf5b2487b583f7e329132bb70a2a"
                },
                "downloads": -1,
                "filename": "Concurrent_AP-1.4.tar.gz",
                "has_sig": false,
                "md5_digest": "2bb16a3e49dca04f2613977fcefe2104",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 14269,
                "upload_time": "2016-01-15T18:44:36",
                "url": "https://files.pythonhosted.org/packages/46/ae/4120492cfa5569ce13bc9fa2457e444fc877c54ed6178c7f7d8499cea252/Concurrent_AP-1.4.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "2bb16a3e49dca04f2613977fcefe2104",
                "sha256": "d6f4b5e82be60d7d4fa9f677e4f823731c5baf5b2487b583f7e329132bb70a2a"
            },
            "downloads": -1,
            "filename": "Concurrent_AP-1.4.tar.gz",
            "has_sig": false,
            "md5_digest": "2bb16a3e49dca04f2613977fcefe2104",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 14269,
            "upload_time": "2016-01-15T18:44:36",
            "url": "https://files.pythonhosted.org/packages/46/ae/4120492cfa5569ce13bc9fa2457e444fc877c54ed6178c7f7d8499cea252/Concurrent_AP-1.4.tar.gz"
        }
    ]
}