{
"info": {
"author": "Cristian Carlos dos Santos",
"author_email": "perestra.ccds@gmail.com",
"bugtrack_url": null,
"classifiers": [],
"description": "# multiverseML\n\nAplica\u00e7\u00e3o para a cria\u00e7\u00e3o de pipelines para desenvovimento de algoritmos de Machine Learning. A ideia \u00e9 facilitar o acompanhamento das altera\u00e7\u00f5es de modelos, m\u00e9tricas e par\u00e2metros sem preocupa\u00e7\u00e3o com organiza\u00e7\u00e3o.\n\nO multiverseML vem com a ideia de organizar o desenvolvimento no conceito de Multiverso. \n\n*O conceito de Multiverso tem suas ra\u00edzes em extrapola\u00e7\u00f5es, at\u00e9 o momento n\u00e3o cient\u00edficas, da moderna Cosmologia e na Teoria Qu\u00e2ntica, e engloba tamb\u00e9m v\u00e1rias ideias oriundas da Teoria da Relatividade de modo a configurar um cen\u00e1rio em que pode ser poss\u00edvel a exist\u00eancia de in\u00fameros Universos onde, em escala global, todas as probabilidades e combina\u00e7\u00f5es ocorrem em algum dos universos. Simplesmente por haver espa\u00e7o suficiente para acoplar outros universos numa estrutura dimensional maior: o chamado Multiverso.*\n\n\nDito isso, o multiverseML organizar\u00e1 o seu modelo nos conceitos de:\n\n- Multiverse(Multiverso): Diret\u00f3rio central de armazenamento de todos os universos.\n- Universe(Universo): O universo \u00e9 todo arquivo no qual exista um monitoramento ativo. Um universo pode ser um desafio a ser resolvido, como uma identifica\u00e7\u00e3o de fraude ou um reconhecimento de imagem.\n- Timeline(Linha do Tempo): Cada Universo ter\u00e1 multiplas linhas temporais. Cada linha temporal ser\u00e1 uma execu\u00e7\u00e3o com sucesso do monitoramento. Cada linha temporal pode ter um modelo diferente, m\u00e9tricas diferentes e par\u00e2metros diferentes. A timeline \u00e9 baseada no versionamento do Git.\n- Reality(Realidade): Realidade \u00e9 a linha do tempo eleita para produ\u00e7\u00e3o. Poder ser disponibilizado um servidor HTTP ou um processo Batch.\n\n\n### Instala\u00e7\u00e3o vers\u00e3o 0.1.6-Alpha\n`pip install multiverseML`\n\n### Utiliza\u00e7\u00e3o\n\nPara utiliza\u00e7\u00e3o \u00e9 necess\u00e1rio primeiramente a importa\u00e7\u00e3o do m\u00f3dulo:\n\n`import multiverseml`\n\nAp\u00f3s, ser\u00e1 necess\u00e1rio definir qual ser\u00e1 o nome do universo a ser criado. Seja criativo!\n\n`universe = 'theoretical'`\n\nEnt\u00e3o criamos uma vari\u00e1vel model com a finalidade de armazenar o nome e o modelo utilizado (\"lr\" no exemplo \u00e9 um modelo de regress\u00e3o linear).\n\n```\n model = {\n 'name': 'Linear Regression',\n 'model': lr\n }\n```\n\nAgora vamos rastrear as m\u00e9tricas:\n\n``` \n metrics = {\n 'rmse': rmse,\n 'r2': r2,\n 'mae': mae\n }\n```\n\n\u00c9 poss\u00edvel tamb\u00e9m adicionar os par\u00e2metros utilizados para o modelo.\n\n```\n param = {\n 'alpha': alpha,\n 'l1_ratio': l1_ratio\n }\n```\n\nPor fim, deve-se enviar os dados para o MultiverseML.\n\n`multiverseml.metrics(universe, model, metrics, param)`",
"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/cccadet/multiverseML/",
"keywords": "multiverseML,machine-learning,pipeline",
"license": "GNU General Public License v3.0",
"maintainer": "",
"maintainer_email": "",
"name": "multiverseML",
"package_url": "https://pypi.org/project/multiverseML/",
"platform": "",
"project_url": "https://pypi.org/project/multiverseML/",
"project_urls": {
"Homepage": "https://github.com/cccadet/multiverseML/"
},
"release_url": "https://pypi.org/project/multiverseML/0.1.7a0/",
"requires_dist": null,
"requires_python": "",
"summary": "Aplica\u00e7\u00e3o para facilitar a cria\u00e7\u00e3o de pipelines para desenvovimento de algoritmos de Machine Learning",
"version": "0.1.7a0"
},
"last_serial": 5463127,
"releases": {
"0.1.4a0": [
{
"comment_text": "",
"digests": {
"md5": "bb0a298c31dd59bc479f5e945d8df6b1",
"sha256": "34680e5d4b4990b0c588dc767619ca0140fee714d417cd70ee99ce5afd301fc1"
},
"downloads": -1,
"filename": "multiverseML-0.1.4a0.tar.gz",
"has_sig": false,
"md5_digest": "bb0a298c31dd59bc479f5e945d8df6b1",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 14183,
"upload_time": "2019-06-26T01:16:24",
"url": "https://files.pythonhosted.org/packages/74/6c/3d7d7a5963d509228b589418b9506ab28e9561f8986f4b3a14fcc73b35f9/multiverseML-0.1.4a0.tar.gz"
}
],
"0.1.5a0": [
{
"comment_text": "",
"digests": {
"md5": "2eeceac4d3ebbd4411dd3c44808532c9",
"sha256": "97b0de2fb7ccc699680a2c836d5c349abc3bdfcfa3e51e1352aa5f6e472053e0"
},
"downloads": -1,
"filename": "multiverseML-0.1.5a0.tar.gz",
"has_sig": false,
"md5_digest": "2eeceac4d3ebbd4411dd3c44808532c9",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 15353,
"upload_time": "2019-06-28T00:09:58",
"url": "https://files.pythonhosted.org/packages/e1/df/a1729fd48cb12189ecc817961e69ab8efbf2d670451c261bd79032608ae0/multiverseML-0.1.5a0.tar.gz"
}
],
"0.1.6a0": [
{
"comment_text": "",
"digests": {
"md5": "ec1775dc6eab2e774d762b8beb3e0bf1",
"sha256": "c11bff9b02d530f27af83d308b445306153b8f7ec29c3590ae98b3cfb159ac3e"
},
"downloads": -1,
"filename": "multiverseML-0.1.6a0.tar.gz",
"has_sig": false,
"md5_digest": "ec1775dc6eab2e774d762b8beb3e0bf1",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 4172,
"upload_time": "2019-06-28T19:37:16",
"url": "https://files.pythonhosted.org/packages/ce/2b/0db015d0455f6cb19e5c1f8da4c4bfc6e5f59dbb7c9a839fd9b6cdc5991d/multiverseML-0.1.6a0.tar.gz"
}
],
"0.1.7a0": [
{
"comment_text": "",
"digests": {
"md5": "ce9e024d4a1df6d361dab39a80a31596",
"sha256": "5d23567f0ca424545bf92d9bfe51e21005bfb9ee1662aa80ca46c41036aa7245"
},
"downloads": -1,
"filename": "multiverseML-0.1.7a0.tar.gz",
"has_sig": false,
"md5_digest": "ce9e024d4a1df6d361dab39a80a31596",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 16560,
"upload_time": "2019-06-28T19:44:35",
"url": "https://files.pythonhosted.org/packages/6b/1a/1e809cbabd19f9716de14efc938f4bdcd94a0ebf55e7a11129c18187ee38/multiverseML-0.1.7a0.tar.gz"
}
]
},
"urls": [
{
"comment_text": "",
"digests": {
"md5": "ce9e024d4a1df6d361dab39a80a31596",
"sha256": "5d23567f0ca424545bf92d9bfe51e21005bfb9ee1662aa80ca46c41036aa7245"
},
"downloads": -1,
"filename": "multiverseML-0.1.7a0.tar.gz",
"has_sig": false,
"md5_digest": "ce9e024d4a1df6d361dab39a80a31596",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 16560,
"upload_time": "2019-06-28T19:44:35",
"url": "https://files.pythonhosted.org/packages/6b/1a/1e809cbabd19f9716de14efc938f4bdcd94a0ebf55e7a11129c18187ee38/multiverseML-0.1.7a0.tar.gz"
}
]
}