{ "info": { "author": "Ray Yang", "author_email": "yangruipis@163.com", "bugtrack_url": null, "classifiers": [ "Environment :: Console", "Intended Audience :: End Users/Desktop", "License :: OSI Approved :: MIT License", "Operating System :: Microsoft", "Operating System :: POSIX", "Programming Language :: Python :: 3.5", "Programming Language :: Python :: 3.6" ], "description": "\n
\n\n
\n\n\nSimple Machine Learning\n\n\u4e00\u4e2a\u7b80\u5355\u7684\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u5b9e\u73b0\n\n\n![](https://img.shields.io/pypi/v/simple_male.svg) [![Build Status](https://travis-ci.org/Yangruipis/simple_ml.svg?branch=master)](https://travis-ci.org/Yangruipis/simple_ml) [![Coverage Status](https://coveralls.io/repos/github/Yangruipis/simple_ml/badge.svg?branch=master)](https://coveralls.io/github/Yangruipis/simple_ml?branch=master)\n[![Codacy Badge](https://api.codacy.com/project/badge/Grade/00c639db60364d12b0102456552fe806)](https://www.codacy.com/app/Yangruipis/simpleML?utm_source=github.com&utm_medium=referral&utm_content=Yangruipis/simpleML&utm_campaign=Badge_Grade)\n\n![](https://img.shields.io/npm/l/express.svg) [![Join the chat at https://gitter.im/simple_ml/Lobby](https://badges.gitter.im/simple_ml/Lobby.svg)](https://gitter.im/simple_ml/Lobby?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)\n\n---\n\n\n# \u5feb\u901f\u5f00\u59cb\n\n## \u5b89\u88c5\n\n**\u73af\u5883\u548c\u4f9d\u8d56\u5e93**\n- python3.5\u53ca\u4ee5\u4e0a\n- windows or Linux\n- numpy (\u6570\u7ec4)\n- matplotlib (\u4f5c\u56fe)\n- scipy (\u6c42\u89e3\u4f18\u5316\u95ee\u9898)\n- requests (\u7528\u4e8e\u5728\u7ebf\u6570\u636e\u96c6\u83b7\u53d6)\n- cvxopt (\u652f\u6301\u5411\u91cf\u673a\u4e2d\u7684\u4e8c\u6b21\u89c4\u5212\u95ee\u9898)\n\n`\u5f3a\u70c8\u63a8\u8350Anaconda\u73af\u5883`\n\n**pip\u5b89\u88c5**\n\n```bash\npip install simple_male\n```\n\u5bf9\uff0c\u662f`simple_ma(chine)le(arning)`\uff0c\u7b80\u5355\u7537\u4eba\uff0c\u56e0\u4e3a`simple_ml`\u5df2\u7ecf\u5728pypi\u4e0a\u88ab\u4eba\u6ce8\u518c\u4e86\u55b5\n\n**git\u5b89\u88c5**\n\n```bash\ngit clone https://github.com/Yangruipis/simple_ml.git\ncd ./simple_ml\npython setup.py install\n```\n\n## \u4f7f\u7528\n\n```python\n# \u4e00\u4e2a\u7b80\u5355\u7684\u4f8b\u5b50\uff0c\u7528CART\u6811\u8fdb\u884c\u4e8c\u5206\u7c7b\nfrom simple_ml.tree import CART\nimport numpy as np\n\nX = np.array([[1,1.1],\n [1,2.0],\n [0,3.0],\n [0,2.2]])\ny = np.array([1,1,0,0])\ncart = CART(min_samples_leaf=1)\ncart.fit(X, y)\nx_test = np.array([[1,2],[3,4]])\nprint(cart.predict(x_test))\n```\n```python\nOut[1]: np.array([1,1])\n```\n\n- `./simple_ml/examples`\u6587\u4ef6\u5939\u4e2d\u63d0\u4f9b\u4e86\u5927\u591a\u6570\u65b9\u6cd5\u7684\u4f7f\u7528\u8303\u4f8b\n- \u66f4\u8be6\u7ec6\u7684\u7528\u6cd5\u89c1\u5e2e\u52a9\u6587\u6863\uff1a [https://yangruipis.github.io/simple_ml/](https://yangruipis.github.io/simple_ml/)\n\n# \u5b83\u80fd\u505a\u4ec0\u4e48\n\n## \u6700\u6700\u6700\u6700\u4e3b\u8981\u7684\u4efb\u52a1\n\n\u5982\u679c\u4f60\u540c\u65f6\u6ee1\u8db3\uff1a\n1. **\u673a\u5668\u5b66\u4e60\u5165\u95e8\u9636\u6bb5**\n2. **python \u8fdb\u9636\u9636\u6bb5**\n\n\u90a3\u4e48\u606d\u559c\u4f60\uff0c\u8fd9\u4e2a\u9879\u76ee\u53ef\u4ee5\u7ed9\u4f60\u63d0\u4f9b\u5982\u4e0b\u5e2e\u52a9\uff1a\n\n- **\u9605\u8bfb\u6e90\u7801**\uff0c \u4e0d\u50cfsklearn\u8fc7\u4e8e\u590d\u6742\u96be\u8bfb\u7684\u6e90\u7801\uff0c\u8fd9\u4e2a\u8f7b\u91cf\u7ea7\u7684\u9879\u76ee\u975e\u5e38\u6613\u8bfb\uff0c\u5e76\u4e14\u6211\u5c3d\u53ef\u80fd\u7684\u589e\u52a0\u4e86\u6ce8\u91ca\uff0c\u63d0\u9ad8\u4ee3\u7801\u7684\u53ef\u8bfb\u6027\n- **\u5b66\u4e60\u77e5\u8bc6**\uff0c\u8be5\u9879\u76ee\u68b3\u7406\u57fa\u672c\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u7684\u79cd\u7c7b\u548c\u6d41\u7a0b\uff0c\u5de5\u7a0b\u5b9e\u73b0\u4e0a\u7684\u5927\u81f4\u6b65\u9aa4\uff0c\u4e2d\u95f4\u51fa\u73b0\u7684\u4e00\u4e9b\u7ec6\u8282\u95ee\u9898\u4ee5\u53ca\u5982\u4f55\u89e3\u51b3\n- **\u5b9e\u65f6\u4ea4\u6d41**\uff0c\u6211\u5728 gitter \u4e0a\u5efa\u7acb\u4e86 [gitchat \u804a\u5929\u5ba4](https://gitter.im/simple_ml/Lobby?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)\uff0c\u6b22\u8fce\u5927\u5bb6\u5c31\u9879\u76ee\u672c\u8eab\u7684\u5177\u4f53\u95ee\u9898\uff0c\u6216\u8005\u5176\u4ed6\u4efb\u4f55\u76f8\u5173\u4e8b\u9879\u8fdb\u884c\u8ba8\u8bba\uff0c\u6b22\u8fce\u5927\u5bb6\u79ef\u6781\u63d0 issues\uff0c\u6211\u4f1a\u7b2c\u4e00\u65f6\u95f4\u56de\u590d\n\n\n## \u4f5c\u4e3a\u4e00\u4e2a\u673a\u5668\u5b66\u4e60\u9879\u76ee\u7684\u4efb\u52a1\n\n### 1. \u6570\u636e\u548c\u7279\u5f81\n\n#### 1.1 \u6570\u636e\u96c6\u83b7\u53d6\n\n`simple_ml`\u63d0\u4f9b\u4e86\u5927\u91cf\u7ecf\u5178\u7684\u673a\u5668\u5b66\u4e60\u6570\u636e\u96c6\u7684\u83b7\u53d6\u63a5\u53e3`DataCollector`\uff0c\u6570\u636e\u96c6\u6765\u81ea[UCI](http://archive.ics.uci.edu/ml/index.php)\u3002\n\n#### 1.2 \u6570\u636e\u9884\u5904\u7406\n\n`simple_ml` \u63d0\u4f9b\u4e86\u5e38\u7528\u7684\u6570\u636e\u9884\u5904\u7406\u65b9\u6cd5\uff0c\u5305\u62ec\u4e86\u7f16\u7801\u3001\u72ec\u70ed\u7f16\u7801\u3001\u7f3a\u5931\u503c\u5904\u7406\u3001\u5f02\u5e38\u503c\u5904\u7406\u4ee5\u53ca\u968f\u673a\u6570\u636e\u96c6\u5212\u5206\u7b49\u3002\n\n\u540c\u65f6\uff0c`simple_ml` \u63d0\u4f9b\u4e86`PCA`\u964d\u7ef4\u65b9\u6cd5\u4ee5\u53ca\u9488\u5bf9\u9ad8\u7ef4\u6570\u636e\u7684`SuperPCA`\u964d\u7ef4\u65b9\u6cd5\u3002\n\n#### 1.3 \u7279\u5f81\u9009\u62e9\n`simple_ml`\u63d0\u4f9b\u4e86Filter\u548cEmbedded\u4e24\u79cd\u7279\u5f81\u9009\u62e9\u65b9\u6cd5\uff0c\u5305\u62ec\u4e86\uff1a\n1. \u65b9\u5dee\u6cd5\n2. \u76f8\u5173\u7cfb\u6570\u6cd5\n3. \u5361\u65b9\u68c0\u9a8c\u6cd5\n4. L1\u6b63\u5219\n5. GBDT\u7279\u5f81\u9009\u62e9\n\n### 2. \u6a21\u578b\n\n#### 2.1 \u4e8c\u5206\u7c7b\n`simple_ml`\u63d0\u4f9b\u4e86\u975e\u5e38\u591a\u7684\u4e8c\u5206\u7c7b\u65b9\u6cd5\uff0c\u4ee5[wine\u6570\u636e\u96c6](http://archive.ics.uci.edu/ml/datasets/Wine)\u4e3a\u4f8b\uff08\u89c1`./simple_ml/examples`\uff09\uff0c\u5206\u7c7b\u6548\u679c\u548c\u65b9\u6cd5\u540d\u79f0\u89c1\u56fe1\u3002\n\n
\n\n\n\u56fe 1. \u4e8c\u5206\u7c7b\u6548\u679c\u56fe\n
\n\n#### 2.2 \u591a\u5206\u7c7b\n\n`simple_ml`\u6682\u65f6\u53ea\u63d0\u4f9b\u4e86\u4e00\u4e9b\u591a\u5206\u7c7b\u7b97\u6cd5\uff0c\u89c1\u4e0b\u56fe\uff0c\u540c\u6837\u662f[wine\u6570\u636e\u96c6](http://archive.ics.uci.edu/ml/datasets/Wine)\uff0c\u540e\u9762\u4f5c\u8005\u5c06\u4f1a\u8fdb\u884c\u8865\u5145\u3002\n\n
\n\n\n\u56fe 2. \u591a\u5206\u7c7b\u6548\u679c\u56fe\n
\n\n\n#### 2.3 \u56de\u5f52\n\n`simple_ml`\u63d0\u4f9b\u4e86\u56de\u5f52\u65b9\u6cd5\u5982\u4e0b\n- `MultiRegression`\n- `CART`\n- `GBDT`\n- `SVR`\n\n#### 2.4 \u805a\u7c7b\n\n`simple_ml`\u63d0\u4f9b\u4e86`K-means\u805a\u7c7b`\u548c`\u5c42\u6b21\u805a\u7c7b`\u4e24\u79cd\u805a\u7c7b\u65b9\u6cd5\n\n`\u6ce8:`\u4ee5\u4e0a\u6240\u6709\u56fe\u5747\u4e3asimple_ml\u76f4\u51fa\uff08\u9700\u8981matplotlib\uff09\n\n### 3. \u6548\u679c\u8bc4\u4ef7\n\n\u5305\u62ec\u4e86\u5206\u7c7b\u548c\u56de\u5f52\u4f5c\u56fe\u3001\u4ee5\u53ca\u9488\u5bf9\u4e8c\u5206\u7c7b\u3001\u591a\u5206\u7c7b\u3001\u56de\u5f52\u95ee\u9898\u7684\u8bc4\u4ef7\u6307\u6807\u8ba1\u7b97\uff0c\u5305\u62ecPrecision, Recall\u7b49\u7b49\n\n# \u4e3a\u4ec0\u4e48\u4f1a\u6709\u8fd9\u4e2a\u9879\u76ee & \u81f4\u8c22\n\n\u4f5c\u8005\u5c31\u8bfb\u4e8e\u4e0a\u6d77\u67d0\u5546\u79d1\u9662\u6821\u7ecf\u6d4e\u5b66\uff0c\u4ece\u5927\u4e8c\u5f00\u59cb\u63a5\u89e6\u6570\u636e\u6316\u6398\uff0c\u4ee5\u53ca\u7f16\u7a0b\u76f8\u5173\u77e5\u8bc6\uff08stata->R->C#->python)\uff0c\u5bf9\u6570\u636e\u548c\u7f16\u7a0b\u975e\u5e38\u611f\u5174\u8da3\uff0c\u57fa\u672c\u4e0a\u4e00\u8def\u8d70\u8fc7\u6765\u5168\u9760\u81ea\u5b66\u3002\u4f5c\u8005\u5e0c\u671b\u53ef\u4ee5\u7528\u5fc3\u505a\u597d\u4e00\u4e2a\u9879\u76ee\uff0c\u8bb0\u5f55\u81ea\u5df1\u5b66\u4e60\u7684\u8f68\u8ff9\uff0c\u5c24\u5176\u662f\u5373\u5c06\u6bd5\u4e1a\u4e4b\u9645\u3002\n\n\u5728\u63a5\u4e0b\u6765\u7684\u4e00\u5e74\u627e\u5de5\u4f5c\u7684\u540c\u65f6\uff0c\u4f5c\u8005\u5c06\u5c3d\u5168\u529b\u7ef4\u62a4\u8be5\u9879\u76ee\uff0c\u4e0d\u65ad\u66f4\u65b0\u548c\u4fee\u6539\uff0c\u70ed\u70c8\u6b22\u8fce\u4efb\u4f55\u8d21\u732e\u548c\u8ba8\u8bba\u3002\n\n**\u81f4\u8c22\uff1a**\n- \u9996\u5148\u611f\u8c22\u6211\u81ea\u5df1\uff0c\u4e00\u8def\u8d70\u6765\u7684\u4e0d\u6613\u5982\u4eba\u996e\u6c34\n- \u5176\u6b21\u611f\u8c22\u6211\u7684\u597d\u53cb[\u4f55\u71d5\u6770](https://github.com/YanjieHe)\u548c[\u7a0b\u521a](https://github.com/chenggang0815)\u5bf9\u6211\u5728\u5b66\u4e60\u548c\u5de5\u4f5c\u4e0a\u7684\u5e2e\u52a9\n- \u6700\u540e\u611f\u8c22\u6240\u6709\u76f8\u5173\u4e66\u7c4d\u3001\u535a\u5ba2\u7684\u4f5c\u8005\uff0c\u5c24\u5176\u611f\u8c22[\u5218\u5efa\u5e73Pinard](https://www.cnblogs.com/pinard/)\u4e00\u4e1d\u4e0d\u82df\u7684\u673a\u5668\u5b66\u4e60\u535a\u5ba2\uff0c\u65e0\u8bba\u662f\u77e5\u8bc6\u8fd8\u662f\u6001\u5ea6\uff0c\u90fd\u4ee4\u4eba\u8083\u7136\u8d77\u656c\n\n\n# \u66f4\u65b0\u65e5\u5fd7\n\n- 2018-04-20\n - \u52a0\u5165BP\u795e\u7ecf\u7f51\u7edc\u7b97\u6cd5`simple.neural_network`\u548c\u76f8\u5173\u7684example\n - \u66f4\u65b0github pages\n- 2018-04-23\n - \u52a0\u5165`Stacking model`\n - \u66f4\u65b0\u6bcf\u4e2a\u6a21\u578b\u7684new()\u51fd\u6570\n - \u91cd\u5199`BaseModel`\u7684`predict`\u548c`score`\u62bd\u8c61\u65b9\u6cd5\uff0c\u4ee5\u68c0\u67e5\u6d4b\u8bd5\u96c6\u662f\u5426\u6ee1\u8db3\u8981\u6c42\n - fix SuperPCA bugs\n- 2018-04-24\n - \u52a0\u5165\u7c7b\uff1a\n `Multi2binary`\uff0c\u7ee7\u627f\u8be5\u7c7b\u7684`BaseClassifier`\u53ef\u4ee5\u5c06\u591a\u5206\u7c7b\u95ee\u9898\u8f6c\u4e3a\u4e8c\u5206\u7c7b\u95ee\u9898\n - \u6dfb\u52a0SVM\uff0c Logistic\uff0cNeuralNetwork, AdaBoost \u7684\u7ee7\u627f\u5173\u7cfb\u548c\u591a\u5206\u7c7b\u65b9\u6cd5\n - \u589e\u52a0\u76f8\u5173\u7684\u591a\u5206\u7c7b\u4f8b\u5b50\uff0c\u4ee5\u53ca\u5e2e\u52a9\u6587\u6863\n - \u91cd\u5199\u7279\u5f81\u7c7b\u578b\u63a8\u65ad\u51fd\u6570\uff0c\u6839\u636e\u591a\u79cd\u7ebf\u7d22\u8fdb\u884c\u63a8\u65ad\n- 2018-04-26\n - \u91cd\u5199\u81ea\u52a8\u5316\u6a21\u5757`auto`\uff0c\u5b9e\u73b0`BaseAuto`\u62bd\u8c61\u7c7b\u4ee5\u53ca\u6570\u636e\u81ea\u52a8\u9884\u5904\u7406\u7684`AutoDataHandle`\u7c7b\n - \u52a0\u5165\u7f51\u683c\u641c\u7d22\u65b9\u6cd5\n - \u52a0\u5165[\u5bab\u9888\u764c\u6570\u636e](https://archive.ics.uci.edu/ml/datasets/Cervical+cancer+%28Risk+Factors%29)\u7684\u5b8c\u6574\u5904\u7406Example\n - \u52a0\u5165`helper`\u6a21\u5757\uff0c\u7528\u4e8e\u683c\u5f0f\u5316\u8f93\u51fa\n - \u52a0\u5165`data_handle`\u6a21\u5757\u7684\u7f3a\u5931\u503c\u7edf\u8ba1\u65b9\u6cd5`nan_summary`\n- 2018-04-27\n - \u52a0\u5165\u81ea\u52a8\u7279\u5f81\u9009\u62e9\u7c7b `AutoFeatureHandle` \uff0c\u4ee5\u53ca\u5bf9\u5e94\u7684\u5bab\u9888\u764cExample\n- 2018-06-11\n - \u52a0\u5165\u56de\u5f52\u4f5c\u56fe\u65b9\u6cd5\n- 2018-06-12\n - \u52a0\u5165\u591a\u5143\u56de\u5f52\u65b9\u6cd5`MultiRegression`\u5e76\u6d4b\u8bd5\n - `MultiRegression`\u4e2d\u52a0\u5165\u52a0\u6743\u56de\u5f52\u65b9\u6cd5\n- 2018-06-13\n - \u91cd\u5199\u652f\u6301\u5411\u91cf\u673a`SVM`\uff0c\u8c03\u7528\u4f18\u5316\u5e93\u8fdb\u884c\u6c42\u89e3\uff0c\u800c\u4e0d\u662f\u624b\u5199SMO\n - \u52a0\u5165\u652f\u6301\u5411\u91cf\u56de\u5f52`SVR`\n - \u6574\u4e2a\u652f\u6301\u5411\u91cf\u76f8\u5173\u7b97\u6cd5,\u5305\u62ec\u4e86`Kernel`\u7c7b, `BaseSupportVector`\u7c7b\u4ee5\u53ca`SVM`,`SVR`\n- 2018-06-20\n - \u6dfb\u52a0\u76f8\u5173\u6d4b\u8bd5\u7528\u4f8b\n- 2018-06-23\n - \u6dfb\u52a0`optimal`\u6a21\u5757\uff0c\u5305\u62ec\u4e86\u722c\u5c71\u6cd5\u548c\u6a21\u62df\u9000\u706b\u6cd5\u8fdb\u884c\u6700\u5c0f\u503c\u6c42\u89e3\n - pypi\u53d1\u5e03\uff0c\u7248\u672c `0.1.2`\n\n# TODO list:\n\n- [ ] test cases\n- [x] an efficient bp network\n- [ ] more optimal methods\n- [x] train test split func in helper\n- [x] other feature select method to add\n- [x] lasso and Ridge\n- [x] add GBDT feature select\n- [x] update Readme\n- [x] setup.py\n- [x] examples\n- [x] get more datasets\n- [x] regression plot\n- [x] more regression method\n- [ ] kd_tree\n- [x] Support Machine Regression\n- [x] more metrics\n- [x] github pages, especially the class map\n- [x] stacking\n- [x] \u4e8c\u5206\u7c7b\u8f6c\u591a\u5206\u7c7b\u5668\n- [x] recognize nan and inf\n- [x] check x before predict, check x and y before score\n- [x] \"self.new()\" function in each model\n- [x] \u652f\u6301\u5411\u91cf\u76f8\u5173\u7b97\u6cd5\u6d4b\u8bd5\u548c\u6587\u6863\u64b0\u5199\n- [ ] pypi\u53d1\u5e03\n- [ ] \u79fb\u9664logistic.py \u4e2d\u5bf9scipy\u7684\u4f9d\u8d56,\u81ea\u5df1\u5199fmin(),\u4ee5\u53ca\u5b9e\u73b0\u5e95\u5c42\u4f18\u5316\u7b97\u6cd5\n- [ ] LSTM\n\n# TODO List AUTO MODEL\n\n- [x] auto data handle\n- [x] auto feature select\n- [ ] auto param select\n- [ ] auto model select\n", "description_content_type": "", "docs_url": null, "download_url": "", "downloads": { "last_day": -1, "last_month": -1, "last_week": -1 }, "home_page": "https://yangruipis.github.io/simple_ml/", "keywords": "", "license": "MIT", "maintainer": "", "maintainer_email": "", "name": "simple_male", "package_url": "https://pypi.org/project/simple_male/", "platform": "", "project_url": "https://pypi.org/project/simple_male/", "project_urls": { "Homepage": "https://yangruipis.github.io/simple_ml/" }, "release_url": "https://pypi.org/project/simple_male/0.1.2/", "requires_dist": null, "requires_python": "", "summary": "A machine learning algorithm implementation", "version": "0.1.2" }, "last_serial": 3992455, "releases": { "0.1.1": [ { "comment_text": "", "digests": { "md5": "aab7400a42e3e6359f543e01470e29d2", "sha256": "3ee016c84a427f241fc1f87ba0d505730e56e2237592ad1a64250c9bc8cc61e0" }, "downloads": -1, "filename": "simple_male-0.1.1.tar.gz", "has_sig": false, "md5_digest": "aab7400a42e3e6359f543e01470e29d2", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1592996, "upload_time": "2018-06-13T11:10:42", "url": "https://files.pythonhosted.org/packages/be/38/b2a9530b50ed66cbedc80dc88f9cd52f72fa4e788ed3683acbce1699814b/simple_male-0.1.1.tar.gz" } ], "0.1.2": [ { "comment_text": "", "digests": { "md5": "d58d0fd6080c8e09ced2ee174c0c2464", "sha256": "06dea56debe6289d9daa5535d5726f1fc036e48176fe4f9fd4cfb3f96cbc7f6d" }, "downloads": -1, "filename": "simple_male-0.1.2.tar.gz", "has_sig": false, "md5_digest": "d58d0fd6080c8e09ced2ee174c0c2464", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1595825, "upload_time": "2018-06-23T07:59:51", "url": "https://files.pythonhosted.org/packages/ca/ed/af7b7310b56161f2fc8d4e79caa31488c6530bf5066ace973452234a59a5/simple_male-0.1.2.tar.gz" } ] }, "urls": [ { "comment_text": "", "digests": { "md5": "d58d0fd6080c8e09ced2ee174c0c2464", "sha256": "06dea56debe6289d9daa5535d5726f1fc036e48176fe4f9fd4cfb3f96cbc7f6d" }, "downloads": -1, "filename": "simple_male-0.1.2.tar.gz", "has_sig": false, "md5_digest": "d58d0fd6080c8e09ced2ee174c0c2464", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 1595825, "upload_time": "2018-06-23T07:59:51", "url": "https://files.pythonhosted.org/packages/ca/ed/af7b7310b56161f2fc8d4e79caa31488c6530bf5066ace973452234a59a5/simple_male-0.1.2.tar.gz" } ] }