{
    "info": {
        "author": "Austin Clyde",
        "author_email": "aclyde@uchicago.edu",
        "bugtrack_url": null,
        "classifiers": [
            "Operating System :: OS Independent",
            "Programming Language :: Python :: 2",
            "Programming Language :: Python :: 3"
        ],
        "description": "# ChebyGCN\n\n```python\npip install ChebyGCN\n```\n\nNotice, for training and testing data, permutations of the data must be done in a certain way to align with \npooling of the graph lapacian. Further, every level of graph corsening is a pool of size two, thus if you want to \npool by 2 and then 4, you need log_2(2 * 4)= 3 levels. You will also need to index your Lapancians as seen below.\n\n```python \nfrom ChebyGCN import layers, coarsening\nA = scipy.sparse.csr.csr_matrix(A) #load adjanecy matrix \ngraphs, perm = coarsening.coarsen(A, levels=3, self_connections=True) #produce graph coarsenings \nX_train = coarsening.perm_data(X_train, perm)\nX_test = coarsening.perm_data(X_test, perm)\nL = [coarsening.laplacian(A, normalized=True) for A in graphs]\n\nx_input = Input(shape=(X_train.shape[1],))\nx = Reshape((X_train.shape[1],1))(x_input)\nx = layers.GraphConvolution( 8, 2, 20, L[0])(x)\nx = layers.GraphConvolution( 8, 4, 10, L[2])(x)\nx = Flatten()(x)\nx = Dense(66, activation='softmax')(x)\n```\n\n\nThis code is 96% based on https://github.com/mdeff/cnn_graph Micha\u00ebl Defferrard, Xavier Bresson, Pierre Vandergheynst, Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering, Neural Information Processing Systems (NIPS), 2016.",
        "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/aclyde11/ChebyGCN",
        "keywords": "",
        "license": "",
        "maintainer": "",
        "maintainer_email": "",
        "name": "ChebyGCN",
        "package_url": "https://pypi.org/project/ChebyGCN/",
        "platform": "",
        "project_url": "https://pypi.org/project/ChebyGCN/",
        "project_urls": {
            "Homepage": "https://github.com/aclyde11/ChebyGCN"
        },
        "release_url": "https://pypi.org/project/ChebyGCN/0.0.3/",
        "requires_dist": null,
        "requires_python": "",
        "summary": "Implements graph convolution keras layers based on Micha\u00ebl Defferrard, Xavier Bresson, Pierre Vandergheynst, Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering, Neural Information Processing Systems (NIPS), 2016.",
        "version": "0.0.3"
    },
    "last_serial": 4744200,
    "releases": {
        "0.0.1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "10870eb4718cf92eb9230f6f9c8cabc6",
                    "sha256": "41ab5ded30091e9f312905f91288a585d2d64196d7b7f696ee65b4ba956019d0"
                },
                "downloads": -1,
                "filename": "ChebyGCN-0.0.1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "10870eb4718cf92eb9230f6f9c8cabc6",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 8473,
                "upload_time": "2019-01-25T17:15:12",
                "url": "https://files.pythonhosted.org/packages/c8/c4/50f7d0142130d99966bf3b5e4cde1b03509f20c329905d11fdc014b644a5/ChebyGCN-0.0.1-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "ef26ab87dbf0e9eec2693f3830c99c46",
                    "sha256": "062abb4779e12fbaa38a141fee13957f8466e77aee36095a0aec79793db2a34d"
                },
                "downloads": -1,
                "filename": "ChebyGCN-0.0.1.tar.gz",
                "has_sig": false,
                "md5_digest": "ef26ab87dbf0e9eec2693f3830c99c46",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 6814,
                "upload_time": "2019-01-25T17:15:14",
                "url": "https://files.pythonhosted.org/packages/fb/7b/b546e87a6a528a08d442b306f929ba45ac83b839efed2a13d3cda88740c7/ChebyGCN-0.0.1.tar.gz"
            }
        ],
        "0.0.2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "8fcec0e55fe06afc32888f0df2af0212",
                    "sha256": "53565fd0cd4a70faa126783290c740088d96c793451bec49f71c17c0c30e22af"
                },
                "downloads": -1,
                "filename": "ChebyGCN-0.0.2.tar.gz",
                "has_sig": false,
                "md5_digest": "8fcec0e55fe06afc32888f0df2af0212",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 6825,
                "upload_time": "2019-01-25T17:33:02",
                "url": "https://files.pythonhosted.org/packages/d0/65/6fa91c5dfb8f1e7d5c001332a307ed0d89d871c1903a9b72ee6e3c99df3b/ChebyGCN-0.0.2.tar.gz"
            }
        ],
        "0.0.3": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "3032c7acce08b8ec66de50eff86c6a12",
                    "sha256": "e882a07d983d4d97d38507e7519a65fa64450c9bb0f26aa4a59ab0f5843321c1"
                },
                "downloads": -1,
                "filename": "ChebyGCN-0.0.3.tar.gz",
                "has_sig": false,
                "md5_digest": "3032c7acce08b8ec66de50eff86c6a12",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 6828,
                "upload_time": "2019-01-26T18:30:26",
                "url": "https://files.pythonhosted.org/packages/ce/5c/44ee685d452af7324c6685850ac8a2cc3e5d2a2bc2dea55cdd9a2816ab90/ChebyGCN-0.0.3.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "3032c7acce08b8ec66de50eff86c6a12",
                "sha256": "e882a07d983d4d97d38507e7519a65fa64450c9bb0f26aa4a59ab0f5843321c1"
            },
            "downloads": -1,
            "filename": "ChebyGCN-0.0.3.tar.gz",
            "has_sig": false,
            "md5_digest": "3032c7acce08b8ec66de50eff86c6a12",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 6828,
            "upload_time": "2019-01-26T18:30:26",
            "url": "https://files.pythonhosted.org/packages/ce/5c/44ee685d452af7324c6685850ac8a2cc3e5d2a2bc2dea55cdd9a2816ab90/ChebyGCN-0.0.3.tar.gz"
        }
    ]
}