{
    "info": {
        "author": "Sofya Lipnitskaya",
        "author_email": "lipnitskaya.sofya@gmail.com",
        "bugtrack_url": null,
        "classifiers": [
            "Development Status :: 4 - Beta",
            "Intended Audience :: Developers",
            "Intended Audience :: Education",
            "Intended Audience :: Other Audience",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved :: MIT License",
            "Natural Language :: English",
            "Operating System :: OS Independent",
            "Programming Language :: Python",
            "Programming Language :: Python :: 3",
            "Programming Language :: Python :: 3.6",
            "Programming Language :: Python :: 3.7",
            "Topic :: Education :: Testing",
            "Topic :: Office/Business",
            "Topic :: Other/Nonlisted Topic",
            "Topic :: Scientific/Engineering",
            "Topic :: Software Development :: Libraries",
            "Typing :: Typed"
        ],
        "description": "# Experiment Runner (exp-runner)\n\n**exp-runner** is a simple and extensible framework for data analysis and machine learning experiments in Python.\n\n#### Structure\nThe framework includes following step:\n1. _Data loading_\n2. _Data transformation_\n3. _Model training and testing_\n4. _Performance evaluation_\n5. _Results saving_\n\n#### Main features\n* _Generability_: Variaty of models and methods are supported and it can be used in a number of tasks \n(such as preprocessing, dimensionality reduction, classification, \nregression, clustering, statistical tests, etc.)\n* _Flexability_: Steps can be easily skipped and/or included\n* _Dynamic loading_: Automatically imports modules during runtime - no additional lines are needed\n\n### Installation\n```bash\npip install exp-runner\n```\n\n### Usage \n\nLet's say, your project has the following structure:\n\n    MyAwesomeProject/\n            main.py\n            my_custom_module.py\n\n            data/\n                    data_00.npy\n                    data_01.npy\n                    ...\n                    data_NN.npy\n\n            protocols/\n                    experiment_config.json\n\n            results/\n\n\n#### Just give me a code!\nYou just need to describe your **framework** in the [JSON](https://json.org) configuration file:\n\n##### experiment_config.json\n```JSON\n{\n  \"Setup\": {\n    \"description\": \"You can add detailed description of the experiment\",\n    \"random_seed\": 42\n  },\n  \"Dataset\": {\n    \"class\": \"my_custom_module.MyAwesomeDataLoader\",\n    \"args\": {\"path_to_data\": \"data/*.npy\"}\n  },\n  \"Transforms\": [\n    {\n      \"class\": \"sklearn.decomposition.PCA\",\n      \"args\": {\"n_components\": 3, \"whiten\": true}\n    }\n  ],\n  \"Model\": {\n    \"class\": \"sklearn.cluster.KMeans\",\n    \"args\": {\"n_clusters\": 3, \"n_jobs\": -1, \"verbose\": 0}\n  },\n  \"Metric\": {\n    \"class\": \"my_custom_module.SklearnMetricWrapper\",\n    \"args\": {\"metric\": \"normalized_mutual_info_score\"}\n  },\n  \"Saver\": {\n    \"class\": \"my_custom_module.CSVReport\",\n    \"args\": {\"path_to_output\": \"results/evaluation_results.csv\", \"sep\": \";\"}\n  }\n}\n```\n<details><summary>Here are aforementioned classes (click):</summary>\n<p>\n\n##### my_custom_module.py\n\n```python\nimport os\nimport glob\nimport numpy as np\nimport sklearn.metrics\n\nfrom exp_runner import Dataset, Metric, Saver\n\nfrom collections import defaultdict\nfrom typing import Any, Dict, List, Union, NoReturn, Iterable, Callable\n\nfrom sklearn.model_selection import StratifiedShuffleSplit\n\n\nclass MyAwesomeDataLoader(Dataset):\n\n    def __init__(self, path_to_data: str, test_size: float = 0.1, training: bool = True):\n\n        super(MyAwesomeDataLoader, self).__init__()\n\n        self._samples = dict()\n        self._labels = dict()\n        self._splits = defaultdict(dict)\n\n        paths_to_data = glob.glob(path_to_data)\n\n        for path in paths_to_data:\n            fname = os.path.basename(path)\n\n            data = np.load(path)\n            X = data[:, :-1]   \n            y = data[:, -1]\n\n            indices_train, indices_test = next(StratifiedShuffleSplit(\n                test_size=test_size\n            ).split(X, y))\n\n            self._samples[fname] = X\n            self._labels[fname] = y\n            self._splits[fname]['train'] = indices_train\n            self._splits[fname]['test'] = indices_test\n\n        self._indices = list(self._samples.keys())\n\n        self._training = training\n\n    def __getitem__(self, index: int) -> Dict[str, Dict[str, Union[str, np.ndarray]]]:\n        if not (0 <= index < len(self._indices)):\n            raise IndexError\n\n        fname = self._indices[index]\n\n        item = {\n        'X': self._samples[fname][self._splits[fname]['train'] if self.training else self._splits[fname]['test']],\n        'y': self._labels[fname][self._splits[fname]['train'] if self.training else self._splits[fname]['test']]\n        }\n\n        item['desc'] = 'it is possible to add description for each data sample'\n\n        return {'filename': fname, 'item': item}\n\n    def __len__(self) -> int:\n        return len(self._indices)\n\n    @property\n    def training(self):\n        return self._training\n\n\nclass SklearnMetricWrapper(Metric):\n\n    def __init__(self, metric: str):\n        super(SklearnMetricWrapper, self).__init__()\n\n        metric = getattr(sklearn.metrics, metric)\n        self._metric: Callable[[Iterable[Union[float, int]], Iterable[Union[float, int]]], float] = metric\n\n    def __call__(self, y_true: Iterable[Union[float, int]], y_pred: Iterable[Union[float, int]]) -> float:\n        return self._metric(y_true, y_pred)\n\n\nclass CSVReport(Saver):\n\n    def __init__(self, path_to_output: str, sep: str = ';', append: bool = True):\n        super(CSVReport, self).__init__()\n\n        self.path_to_output = path_to_output\n        self.sep = sep\n        self.mode = 'a+' if append else 'w+'\n\n    def save(self, report: List[Dict[str, Any]]) -> NoReturn:\n        with open(self.path_to_output, self.mode) as csv:\n            for entry in report:\n                line = self.sep.join([\n                    entry['filename'],\n                    entry['desc'],\n                    entry['perf']\n                ]) + '\\n'\n                csv.write(line)\n```\n\n</p>\n</details>\n\nFinally, to run your experiment type in your terminal:\n```bash\ncd /path/to/MyAwesomeProject\npython main.py --config protocols/experiment_config.json\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/slipnitskaya/exp-runner.git",
        "keywords": "pipeline framework modelling model training testing classification regression clustering,pipeline framework data analysis",
        "license": "",
        "maintainer": "",
        "maintainer_email": "",
        "name": "exp-runner",
        "package_url": "https://pypi.org/project/exp-runner/",
        "platform": "",
        "project_url": "https://pypi.org/project/exp-runner/",
        "project_urls": {
            "Homepage": "https://github.com/slipnitskaya/exp-runner.git"
        },
        "release_url": "https://pypi.org/project/exp-runner/0.1.0b2/",
        "requires_dist": [
            "tqdm (>=4.28.1)",
            "numpy (>=1.15.4)",
            "scikit-learn (>=0.20.1)"
        ],
        "requires_python": ">=3.6",
        "summary": "Framework for data analysis and machine learning experiments",
        "version": "0.1.0b2"
    },
    "last_serial": 5229623,
    "releases": {
        "0.1.0b1": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "2ca3883f6edcb11650a9f850c0815bb6",
                    "sha256": "1f5ed8db3b1f07d341ab0c776634473b4fd66755c868b2bda101f1e5d8025c52"
                },
                "downloads": -1,
                "filename": "exp_runner-0.1.0b1-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "2ca3883f6edcb11650a9f850c0815bb6",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": ">=3.6",
                "size": 6603,
                "upload_time": "2019-03-25T19:51:35",
                "url": "https://files.pythonhosted.org/packages/1a/bf/f14372c4d8c2bd97c74d1ee9f6d0cb7134218e017a1130bb78bc69e1ddf0/exp_runner-0.1.0b1-py3-none-any.whl"
            },
            {
                "comment_text": "",
                "digests": {
                    "md5": "6b4a80ee0d161cc5b1be778e22ed5cc5",
                    "sha256": "6ab14d9a791879b9ab200aee5fbe1c850a8d10850a79dd5f6a0240255a74d269"
                },
                "downloads": -1,
                "filename": "exp-runner-0.1.0b1.tar.gz",
                "has_sig": false,
                "md5_digest": "6b4a80ee0d161cc5b1be778e22ed5cc5",
                "packagetype": "sdist",
                "python_version": "source",
                "requires_python": ">=3.6",
                "size": 5498,
                "upload_time": "2019-03-25T19:51:42",
                "url": "https://files.pythonhosted.org/packages/1f/25/72585f764dc7da4451903160fd4dcf78158fda90492d023c2e784b485cf8/exp-runner-0.1.0b1.tar.gz"
            }
        ],
        "0.1.0b2": [
            {
                "comment_text": "",
                "digests": {
                    "md5": "5cbca2ebb141bb49036031448c596be6",
                    "sha256": "ed7118cada6a5f2dcc74ac84485bdddc0fc988514a18f7fac905da42e822b202"
                },
                "downloads": -1,
                "filename": "exp_runner-0.1.0b2-py3-none-any.whl",
                "has_sig": false,
                "md5_digest": "5cbca2ebb141bb49036031448c596be6",
                "packagetype": "bdist_wheel",
                "python_version": "py3",
                "requires_python": ">=3.6",
                "size": 6637,
                "upload_time": "2019-05-05T20:49:45",
                "url": "https://files.pythonhosted.org/packages/3a/5c/a9af5166e6c2dbf84e945b248409494a0f33e59bb4ac0d500203ccb38b1d/exp_runner-0.1.0b2-py3-none-any.whl"
            }
        ]
    },
    "urls": [
        {
            "comment_text": "",
            "digests": {
                "md5": "5cbca2ebb141bb49036031448c596be6",
                "sha256": "ed7118cada6a5f2dcc74ac84485bdddc0fc988514a18f7fac905da42e822b202"
            },
            "downloads": -1,
            "filename": "exp_runner-0.1.0b2-py3-none-any.whl",
            "has_sig": false,
            "md5_digest": "5cbca2ebb141bb49036031448c596be6",
            "packagetype": "bdist_wheel",
            "python_version": "py3",
            "requires_python": ">=3.6",
            "size": 6637,
            "upload_time": "2019-05-05T20:49:45",
            "url": "https://files.pythonhosted.org/packages/3a/5c/a9af5166e6c2dbf84e945b248409494a0f33e59bb4ac0d500203ccb38b1d/exp_runner-0.1.0b2-py3-none-any.whl"
        }
    ]
}