{
    "info": {
        "author": "Lucio Anderlini",
        "author_email": "l.anderlini@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "License :: OSI Approved :: MIT License",
            "Operating System :: OS Independent",
            "Programming Language :: Python"
        ],
        "description": "# FastQuantileLayer\nFastQuantileLayer is a Layer for Keras implementing the QuantileTransform \nsimilarly to scikit-learn QuantileTransformer. \nA similar implementation, more precise but not bound to Keras, can be found here: \n  https://github.com/yandexdataschool/QuantileTransformerTF/blob/master/README.md\n\nThe purpose of this package is:\n - remove all dependencies on scikit-learn\n - obtain an evaluation of the direct and inverse transform as fast as possible \n   (trading some precision for performance)\n - obtain a TensorFlow graph runnable in a Sequential model in Keras \n\n\nThe package is composed of two classes:\n - *FixedBinInterpolator*: intended to interpolate a point-defined function\n   y = f(x) with equidistant x samples (x-grid)\n - *FastQuantileLayer*: intended to compute the transform to preprocess\n   the input data into a uniform- or normal-distributed variable. \n\n\n### Example outside Keras \n\n```\n  ## Creates the training dataset \n  dataset = np.random.uniform ( 0., 1., 1000 ) \n\n  ## Train the QuantileTransformer \n  transformer = FastQuantileLayer (output_distribution='normal')\n  transformer . fit ( dataset ) \n\n  ## Gets a new dataset with the same distribution as the training dataset\n  test_dataset = tf.constant(np.random.uniform ( 0., 1., 100000 ))\n\n  ## Transform the variable into a Gaussian-distributed variable t \n  t = transformer . transform ( test_dataset ) \n\n  [...] \n\n  ## Appiles the inverted transform to the Gaussian distributed variable t  \n  bkwd = transformer . transform ( t, inverse=True ) \n\n  ## bkwd differs from test_dataset only for computational errors \n  ## (order 1e-5) that can be reduced tuning the arguments of QuantileTransformer\n```\n\n\n### Example within Keras \n\n```\n  ## Creates the training dataset \n  dataset = np.random.uniform ( 0., 1., 1000 ) \n\n  model = tf.keras.models.Sequential()\n  model.add ( FastQuantileLayer ( output_distribution = 'normal' ).fit ( dataset ) )\n  model.add ( Dense ( 10, activation = 'tanh' ) )\n  model.add ( Dense ( 1, activation = 'sigmoid' ) ) \n\n\n```\n\n\n",
        "description_content_type": "text/markdown",
        "docs_url": null,
        "download_url": "",
        "downloads": {
            "last_day": -1,
            "last_month": -1,
            "last_week": -1
        },
        "home_page": "https://github.com/landerli/FastQuantileLayer",
        "keywords": "",
        "license": "",
        "maintainer": "",
        "maintainer_email": "",
        "name": "FastQuantileLayer",
        "package_url": "https://pypi.org/project/FastQuantileLayer/",
        "platform": "",
        "project_url": "https://pypi.org/project/FastQuantileLayer/",
        "project_urls": {
            "Homepage": "https://github.com/landerli/FastQuantileLayer"
        },
        "release_url": "https://pypi.org/project/FastQuantileLayer/0.2.1/",
        "requires_dist": null,
        "requires_python": "",
        "summary": "Keras Layer to apply Quantile transform and its inverse",
        "version": "0.2.1"
    },
    "last_serial": 5349413,
    "releases": {
        "0.2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "cfb2c36b712fc2e9be036ca2af1c804a",
                    "sha256": "df20435f173f27ae16936a2e582e664bb4d3e29940e509b5cb7c7d73cda4f15a"
                },
                "downloads": -1,
                "filename": "FastQuantileLayer-0.2-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "cfb2c36b712fc2e9be036ca2af1c804a",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 7481,
                "upload_time": "2019-05-31T17:20:49",
                "url": "https://files.pythonhosted.org/packages/8c/92/30f60f7f3a7a605ddebf1656392410495ff18935a4a27cbf5ef9a7029bd6/FastQuantileLayer-0.2-py3-none-any.whl"
            }
        ],
        "0.2.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "723a9d98f3a17bbcceb9eb3e21b46930",
                    "sha256": "51278f9f4b6f15ef4a87649a45437a4e7533c36b83f7d98d0516868b4719e333"
                },
                "downloads": -1,
                "filename": "FastQuantileLayer-0.2.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "723a9d98f3a17bbcceb9eb3e21b46930",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 10588,
                "upload_time": "2019-06-02T17:27:50",
                "url": "https://files.pythonhosted.org/packages/0f/0e/e4030e40c4571d68f85bef054d379911f2a8e6491325be1a02a2ec0964c4/FastQuantileLayer-0.2.1-py3-none-any.whl"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "723a9d98f3a17bbcceb9eb3e21b46930",
                "sha256": "51278f9f4b6f15ef4a87649a45437a4e7533c36b83f7d98d0516868b4719e333"
            },
            "downloads": -1,
            "filename": "FastQuantileLayer-0.2.1-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "723a9d98f3a17bbcceb9eb3e21b46930",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": null,
            "size": 10588,
            "upload_time": "2019-06-02T17:27:50",
            "url": "https://files.pythonhosted.org/packages/0f/0e/e4030e40c4571d68f85bef054d379911f2a8e6491325be1a02a2ec0964c4/FastQuantileLayer-0.2.1-py3-none-any.whl"
        }
    ]
}