{
    "info": {
        "author": "Rachit Jas, Jiaqi Liu, Qiuyi Han, Jie Ding",
        "author_email": "dingj@umn.edu",
        "bugtrack_url": null,
        "classifiers": [
            "Intended Audience :: Developers",
            "Intended Audience :: Education",
            "Intended Audience :: Information Technology",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved :: GNU General Public License v3 (GPLv3)",
            "Programming Language :: Python",
            "Programming Language :: Python :: 3.6",
            "Programming Language :: Python :: 3.7",
            "Programming Language :: Python :: 3.8",
            "Topic :: Education",
            "Topic :: Scientific/Engineering",
            "Topic :: Scientific/Engineering :: Information Analysis",
            "Topic :: Scientific/Engineering :: Mathematics"
        ],
        "description": "# The 'slants' Python API\n\nThis package provides a Python implementation as well as an user-friendly API for the **Sequential Adaptive Nonlinear Modeling of Time Series** (SLANTS) algorithm, as proposed by **Q. Han, J. Ding, E. M. Airoldi and V. Tarokh** in the _IEEE Transactions on Signal Processing Journal, Vol. 65, NO. 19, October 2017, pages [4994, 5005]_. SLANTS provides a new method for online modeling and prediction of nonlinear and nonparametric autoregressive time series. \n\n- It uses **splines** to approximate a wide range of **nonlinear functions and adaptive filtering** to accommodate time varying data generating processes. \n- It is built on a **new online group LASSO algorithm** proposed in the reference paper. \n- It can be **applied to high dimensional time series** where the dimension is larger than the sample size.\n\nFurther details and theory about the algorithm can be found here [[PDF](http://jding.org/jie-uploads/2018/11/slant.pdf)]\n\n## A Quick Setup Guide\n\n### Getting Started \n\n#### 1. Install the 'slants' package using pip\n\n```bash\n# Installing test package\npython -m pip install slants\n```\n\n#### 2. Import the Model and Experiment API classes\n\nTwo fundamental API classes are available. The **Model Class API** provides online model fitting capabilities of the SLANTS algorithm for your data as well as diagnostic plot graphing functions to view learned optimal model parameters. The **Experiment Class API** is a utility data-loader class that will quickly set you up with the environment to test the SLANTS algorithm API by loading sample synthetic experiments. You may import both classes as follows:\n\n```python\nfrom slants import Model, Experiment\n```\n\n### Using the package\n\n#### Experiment Class API\n\n##### 1. Load a synthetic experiment dataset using the Experiment.load() method \nThis method returns a dictionary containing the design data matrix + tunable parameters for the model.\n\n```python\nexp = Experiment.load(1) # load synthetic experiment no. (1)\n```\n##### 2. View number of available synthetic datasets\n\n```python\nExperiment.load('?') \n```\n\nAll the tunable parameters for model fitting, inclusive of the data design matrix, can be accessed from the 'exp' dictionary.\n\n#### Model Class API\n\n##### 1. Instantiate Model objects and viewing available methods:\n\n```python\nmodel = Model()\nmodel.help(ls=True) #ls ~ inspiration from linux bash-style command\n```\n##### 2. Fit SLANTS algorithm on data design matrix:\n\n```python\n# Consolidating spline spline configuration + optional tunable parameters \nspline_info = {\n    'order': spline_info_order, \n    'num_Bsplines': spline_info_num_Bsplines\n}\n\noptional_params = {\n    'test_size': test_size, \n    'safe_shrink_lambda': safe_shrink_lambda, \n    'random_beta': random_beta\n}\n## Note: Many more optional tunable parameters are available; see documentation\n\n# Fitting the SLANTS model on the data \nmodel.fit(data, predict, maximum_lag, spline_info, forget_factor,  \n          lambda_init, tau2_init, sparse_lambda_tolerance,\t\t \t\t\t\t  \n          delta_multiplier, shrink_step_size, **optional_params) # unpacking optional args dictionary \n```\n\nEither check the **Complete_User_Guide.IPYNB** notebook or type the following to get additional documentation regarding the Model.fit() method:\n\n```python\nmodel.help('fit')\n```\n\n##### 3. View diagnostic plots\n\n```python\nmodel.plot('coef') # show trend of beta coefficients with time\nmodel.plot('lambda') # show trend of LASSO penalty term with time\nmodel.plot('tau') # show trend of EM decomposition parameter with time\nmodel.plot('trace') # show trace of nonlinear functions\n``` \n\n##### 4. Return learned model parameters\n\n```python\nlearned_model_params = model.model_params() # dictionary consisting of all learned parameters\n```\n\nEither check **Complete_User_Guide.IPYNB** notebook or type the following to get additional documentation regarding the Model.model_params() method:\n\n```python\nmodel.help('model_params')\n```\n\n##### 5. Reset the model in rare cases of unexpected behaviour\n\n```python\nmodel.reset_model()\n``` \n\n##### 6. View man-style documentation of the Model class/methods\n\n```python\nmodel.help(ls=True) # show list of methods in Model class \nmodel.help('fit') # show info about function name, for eg. fit()\nmodel.help('all') # show info about entire class\n```\n\n## Software Guide\n\nInteractive hands-on guides pertaining to the package may be accessed [[online](https://colab.research.google.com/drive/131JrKNapbWYMLUk67N2UUSBcFiiY_1kd)] on Google Colab or is available for download in the ```docs/``` folder of this project. Developers can avail all software-design charts of this project in the ```docs/``` folder. \n\n## Relevant Papers\n\nQ .Han, J. Ding, E. Airoldi, V. Tarokh, \"SLANTS: Sequential adaptive nonlinear modeling of time series,\" IEEE Transactions on Signal Processing 65 (19), 4994-5005. [[PDF](http://jding.org/jie-uploads/2018/11/slant.pdf)]\n\nX. Xian, J. Ding, \"Physics-assisted online learning,\" preprint.\n\n## Acknowledgement\n\nThis research is funded by the Defense Advanced Research Projects Agency (DARPA) under grant number HR00111890040.\n\n## License\n\nThe software is subjected to the GNU GPLv3 licensing terms and agreements. \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/JieGroup/slants",
        "keywords": "sequential adaptive online time series machine learning expectation maximization algorithm group lasso statistics",
        "license": "GNU GPLv3",
        "maintainer": "",
        "maintainer_email": "",
        "name": "slants",
        "package_url": "https://pypi.org/project/slants/",
        "platform": "",
        "project_url": "https://pypi.org/project/slants/",
        "project_urls": {
            "Homepage": "https://github.com/JieGroup/slants"
        },
        "release_url": "https://pypi.org/project/slants/1.0.0/",
        "requires_dist": [
            "numpy",
            "pandas",
            "matplotlib"
        ],
        "requires_python": "",
        "summary": "Sequential Learning Algorithm for Nonlinear Time Series (SLANTS)",
        "version": "1.0.0"
    },
    "last_serial": 5575321,
    "releases": {
        "1.0.0": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "8e5e0ee48a383ac28daa9b4d0ca7be0c",
                    "sha256": "d94562d25fdabee8e38a22d9474fa4fac9ea235e4dde97d395d1a8e1b6e666a3"
                },
                "downloads": -1,
                "filename": "slants-1.0.0-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "8e5e0ee48a383ac28daa9b4d0ca7be0c",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": null,
                "size": 359266,
                "upload_time": "2019-07-24T02:10:37",
                "url": "https://files.pythonhosted.org/packages/8d/9f/a1d81c36d72b4f050ef1740ee580cff4f9da68857b0304dab04ddef52a5b/slants-1.0.0-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "e258bd5ffeed899361753f1501e15f4c",
                    "sha256": "c194a62d57a328d599549f44723e0a7a3cee9e2085f861ad20e0098e33bf5175"
                },
                "downloads": -1,
                "filename": "slants-1.0.0.tar.gz",
                "has_sig": false,
                "md5_digest": "e258bd5ffeed899361753f1501e15f4c",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": null,
                "size": 339967,
                "upload_time": "2019-07-24T02:10:40",
                "url": "https://files.pythonhosted.org/packages/61/78/0f7f1a1b1abf46b1935b38c18c85645ef707c105660ae326389888a679a7/slants-1.0.0.tar.gz"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "8e5e0ee48a383ac28daa9b4d0ca7be0c",
                "sha256": "d94562d25fdabee8e38a22d9474fa4fac9ea235e4dde97d395d1a8e1b6e666a3"
            },
            "downloads": -1,
            "filename": "slants-1.0.0-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "8e5e0ee48a383ac28daa9b4d0ca7be0c",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": null,
            "size": 359266,
            "upload_time": "2019-07-24T02:10:37",
            "url": "https://files.pythonhosted.org/packages/8d/9f/a1d81c36d72b4f050ef1740ee580cff4f9da68857b0304dab04ddef52a5b/slants-1.0.0-py3-none-any.whl"
        },
        {
            "comment_text": "",
            "digests": {
                "md5": "e258bd5ffeed899361753f1501e15f4c",
                "sha256": "c194a62d57a328d599549f44723e0a7a3cee9e2085f861ad20e0098e33bf5175"
            },
            "downloads": -1,
            "filename": "slants-1.0.0.tar.gz",
            "has_sig": false,
            "md5_digest": "e258bd5ffeed899361753f1501e15f4c",
            "packagetype": "sdist",
            "python_version": "source",
            "requires_python": null,
            "size": 339967,
            "upload_time": "2019-07-24T02:10:40",
            "url": "https://files.pythonhosted.org/packages/61/78/0f7f1a1b1abf46b1935b38c18c85645ef707c105660ae326389888a679a7/slants-1.0.0.tar.gz"
        }
    ]
}