{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": "                      project                                    commit  \\\n0     1adrianb/face-alignment  7bef25dfe9fa41730cb2d63b21eaa82e68dca1a1   \n1     1adrianb/face-alignment  478c38823785db76ca85c059cd455a23a586b663   \n2     1adrianb/face-alignment  b7199a49a3baf29625eee68615977a3c9f3413fc   \n3     1adrianb/face-alignment  79878d6232ef79cde9cfb8ac396d42c346e2e61e   \n4     1adrianb/face-alignment  c4eeeb6c990a9cb2dd6547763d2f68c4a107f5ca   \n...                       ...                                       ...   \n3130  ZQPei/deep_sort_pytorch  7df8404cfee854056a4d5d881be8be99d54dc55f   \n3131  ZQPei/deep_sort_pytorch  82ad8ef89135eec05d07b174ec66c53567144f3d   \n3132  ZQPei/deep_sort_pytorch  664c3065809da2bc331ed9f3a4efda19d4654958   \n3133  ZQPei/deep_sort_pytorch  3df392310def8f3ab160e73c8cdc9fbbd6234bf6   \n3134  ZQPei/deep_sort_pytorch  4c2d86229b0b69316af67d519f8476eee69c9b20   \n\n              commit_date                                            author  \\\n0     2017-09-20 08:49:02  Adrian Bulat <1adrianb@users.noreply.github.com>   \n1     2017-09-21 08:06:51  Adrian Bulat <1adrianb@users.noreply.github.com>   \n2     2017-09-25 05:49:13  Adrian Bulat <1adrianb@users.noreply.github.com>   \n3     2017-09-27 00:17:25  Adrian Bulat <1adrianb@users.noreply.github.com>   \n4     2017-10-02 10:24:45             Adrian Bulat <bulat.adrian@gmail.com>   \n...                   ...                                               ...   \n3130  2020-04-24 14:33:15                             pzq <dfzspzq@163.com>   \n3131  2020-05-08 08:28:50                             pzq <dfzspzq@163.com>   \n3132  2020-05-08 08:32:20                             pzq <dfzspzq@163.com>   \n3133  2020-05-19 13:28:13   Masoud Masoumi Moghadam <masouduut94@gmail.com>   \n3134  2020-05-23 02:54:18   Masoud Masoumi Moghadam <masouduut94@gmail.com>   \n\n                                                  files  \\\n0                                         ['README.md']   \n1                             ['face_alignment/api.py']   \n2                             ['face_alignment/api.py']   \n3                                          ['setup.py']   \n4                                        ['Dockerfile']   \n...                                                 ...   \n3130                        ['yolov3_deepsort_eval.py']   \n3131  ['ped_det_server.py', 'utils/json_logger.py', ...   \n3132                        ['deep_sort/deep/train.py']   \n3133  ['deep_sort/deep/original_model.py', 'detector...   \n3134  ['webserver/.env', 'webserver/config/config.py...   \n\n                                          changed_lines issues  is_ml phases  \\\n0                                                   [2]    [1]  False    NaN   \n1                                                   [4]    [3]  False    NaN   \n2                                                  [10]    [7]  False    NaN   \n3                                                   [3]    [8]  False    NaN   \n4                                                   [5]   [13]  False    NaN   \n...                                                 ...    ...    ...    ...   \n3130                                                [2]  [126]  False    NaN   \n3131                                 [155, 383, 39, 21]  [140]  False    NaN   \n3132                                              [378]  [130]  False    NaN   \n3133  [212, 106, 9, 44, 128, 116, 235, 13, 206, 264,...  [146]  False    NaN   \n3134                                [3, 3, 26, 11, 110]  [147]  False    NaN   \n\n      file_entropy  line_entropy            open_date           close_date  \\\n0         0.138103      0.007068  2017-09-20 07:46:57  2017-09-20 07:49:03   \n1         0.138103      0.012796  2017-09-21 03:33:09  2017-09-21 07:06:53   \n2         0.138103      0.027558  2017-09-25 04:23:20  2017-09-25 04:49:14   \n3         0.138103      0.010014  2017-09-26 09:19:48  2017-09-26 23:17:26   \n4         0.138103      0.015455  2017-10-01 20:57:56  2017-10-02 09:25:32   \n...            ...           ...                  ...                  ...   \n3130      0.072132      0.003007  2020-04-24 06:22:17  2020-04-24 06:33:16   \n3131      0.288527      0.381703  2020-05-03 16:24:06  2020-05-08 00:28:51   \n3132      0.072132      0.208893  2020-04-28 03:01:38  2020-05-08 00:32:21   \n3133      1.442634      1.409697  2020-05-19 08:54:39  2020-05-19 08:58:14   \n3134      0.504922      0.134130  2020-05-22 22:24:08  2020-05-22 22:24:19   \n\n      n_comments  words_mean  day_to_fix  \n0            1.0        22.0           0  \n1            1.0        18.0           0  \n2            1.0         1.0           0  \n3            1.0        30.0           0  \n4            1.0        34.0           0  \n...          ...         ...         ...  \n3130         1.0         6.0           0  \n3131         1.0        25.0           4  \n3132         0.0         0.0           9  \n3133         0.0         0.0           0  \n3134         0.0         0.0           0  \n\n[3135 rows x 16 columns]",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>project</th>\n      <th>commit</th>\n      <th>commit_date</th>\n      <th>author</th>\n      <th>files</th>\n      <th>changed_lines</th>\n      <th>issues</th>\n      <th>is_ml</th>\n      <th>phases</th>\n      <th>file_entropy</th>\n      <th>line_entropy</th>\n      <th>open_date</th>\n      <th>close_date</th>\n      <th>n_comments</th>\n      <th>words_mean</th>\n      <th>day_to_fix</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>1adrianb/face-alignment</td>\n      <td>7bef25dfe9fa41730cb2d63b21eaa82e68dca1a1</td>\n      <td>2017-09-20 08:49:02</td>\n      <td>Adrian Bulat &lt;1adrianb@users.noreply.github.com&gt;</td>\n      <td>['README.md']</td>\n      <td>[2]</td>\n      <td>[1]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.138103</td>\n      <td>0.007068</td>\n      <td>2017-09-20 07:46:57</td>\n      <td>2017-09-20 07:49:03</td>\n      <td>1.0</td>\n      <td>22.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1adrianb/face-alignment</td>\n      <td>478c38823785db76ca85c059cd455a23a586b663</td>\n      <td>2017-09-21 08:06:51</td>\n      <td>Adrian Bulat &lt;1adrianb@users.noreply.github.com&gt;</td>\n      <td>['face_alignment/api.py']</td>\n      <td>[4]</td>\n      <td>[3]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.138103</td>\n      <td>0.012796</td>\n      <td>2017-09-21 03:33:09</td>\n      <td>2017-09-21 07:06:53</td>\n      <td>1.0</td>\n      <td>18.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>1adrianb/face-alignment</td>\n      <td>b7199a49a3baf29625eee68615977a3c9f3413fc</td>\n      <td>2017-09-25 05:49:13</td>\n      <td>Adrian Bulat &lt;1adrianb@users.noreply.github.com&gt;</td>\n      <td>['face_alignment/api.py']</td>\n      <td>[10]</td>\n      <td>[7]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.138103</td>\n      <td>0.027558</td>\n      <td>2017-09-25 04:23:20</td>\n      <td>2017-09-25 04:49:14</td>\n      <td>1.0</td>\n      <td>1.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>1adrianb/face-alignment</td>\n      <td>79878d6232ef79cde9cfb8ac396d42c346e2e61e</td>\n      <td>2017-09-27 00:17:25</td>\n      <td>Adrian Bulat &lt;1adrianb@users.noreply.github.com&gt;</td>\n      <td>['setup.py']</td>\n      <td>[3]</td>\n      <td>[8]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.138103</td>\n      <td>0.010014</td>\n      <td>2017-09-26 09:19:48</td>\n      <td>2017-09-26 23:17:26</td>\n      <td>1.0</td>\n      <td>30.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>1adrianb/face-alignment</td>\n      <td>c4eeeb6c990a9cb2dd6547763d2f68c4a107f5ca</td>\n      <td>2017-10-02 10:24:45</td>\n      <td>Adrian Bulat &lt;bulat.adrian@gmail.com&gt;</td>\n      <td>['Dockerfile']</td>\n      <td>[5]</td>\n      <td>[13]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.138103</td>\n      <td>0.015455</td>\n      <td>2017-10-01 20:57:56</td>\n      <td>2017-10-02 09:25:32</td>\n      <td>1.0</td>\n      <td>34.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>3130</th>\n      <td>ZQPei/deep_sort_pytorch</td>\n      <td>7df8404cfee854056a4d5d881be8be99d54dc55f</td>\n      <td>2020-04-24 14:33:15</td>\n      <td>pzq &lt;dfzspzq@163.com&gt;</td>\n      <td>['yolov3_deepsort_eval.py']</td>\n      <td>[2]</td>\n      <td>[126]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.072132</td>\n      <td>0.003007</td>\n      <td>2020-04-24 06:22:17</td>\n      <td>2020-04-24 06:33:16</td>\n      <td>1.0</td>\n      <td>6.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3131</th>\n      <td>ZQPei/deep_sort_pytorch</td>\n      <td>82ad8ef89135eec05d07b174ec66c53567144f3d</td>\n      <td>2020-05-08 08:28:50</td>\n      <td>pzq &lt;dfzspzq@163.com&gt;</td>\n      <td>['ped_det_server.py', 'utils/json_logger.py', ...</td>\n      <td>[155, 383, 39, 21]</td>\n      <td>[140]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.288527</td>\n      <td>0.381703</td>\n      <td>2020-05-03 16:24:06</td>\n      <td>2020-05-08 00:28:51</td>\n      <td>1.0</td>\n      <td>25.0</td>\n      <td>4</td>\n    </tr>\n    <tr>\n      <th>3132</th>\n      <td>ZQPei/deep_sort_pytorch</td>\n      <td>664c3065809da2bc331ed9f3a4efda19d4654958</td>\n      <td>2020-05-08 08:32:20</td>\n      <td>pzq &lt;dfzspzq@163.com&gt;</td>\n      <td>['deep_sort/deep/train.py']</td>\n      <td>[378]</td>\n      <td>[130]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.072132</td>\n      <td>0.208893</td>\n      <td>2020-04-28 03:01:38</td>\n      <td>2020-05-08 00:32:21</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>9</td>\n    </tr>\n    <tr>\n      <th>3133</th>\n      <td>ZQPei/deep_sort_pytorch</td>\n      <td>3df392310def8f3ab160e73c8cdc9fbbd6234bf6</td>\n      <td>2020-05-19 13:28:13</td>\n      <td>Masoud Masoumi Moghadam &lt;masouduut94@gmail.com&gt;</td>\n      <td>['deep_sort/deep/original_model.py', 'detector...</td>\n      <td>[212, 106, 9, 44, 128, 116, 235, 13, 206, 264,...</td>\n      <td>[146]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>1.442634</td>\n      <td>1.409697</td>\n      <td>2020-05-19 08:54:39</td>\n      <td>2020-05-19 08:58:14</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>3134</th>\n      <td>ZQPei/deep_sort_pytorch</td>\n      <td>4c2d86229b0b69316af67d519f8476eee69c9b20</td>\n      <td>2020-05-23 02:54:18</td>\n      <td>Masoud Masoumi Moghadam &lt;masouduut94@gmail.com&gt;</td>\n      <td>['webserver/.env', 'webserver/config/config.py...</td>\n      <td>[3, 3, 26, 11, 110]</td>\n      <td>[147]</td>\n      <td>False</td>\n      <td>NaN</td>\n      <td>0.504922</td>\n      <td>0.134130</td>\n      <td>2020-05-22 22:24:08</td>\n      <td>2020-05-22 22:24:19</td>\n      <td>0.0</td>\n      <td>0.0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n<p>3135 rows × 16 columns</p>\n</div>"
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "data = pd.read_csv('commit_analysis.csv')\n",
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "outputs": [
    {
     "data": {
      "text/plain": "                         project                                    commit  \\\n353   CamDavidsonPilon/lifelines  d2be007e7aea254670e244942d07df547beee603   \n482   CamDavidsonPilon/lifelines  55a81f179b87e6370bd0fa5ff573c9ab8582d91f   \n501   CamDavidsonPilon/lifelines  f7ca1239b150ee5928d7e2937a2551f40272b4f7   \n516   CamDavidsonPilon/lifelines  af56f7a7925feeb36df59ac4bc30e0b8e78b2c12   \n853           deepfakes/faceswap  27a685383ec2fc1df2f60a5648d95b722b2f4e94   \n1246               gboeing/osmnx  2d1756bc161061d0d3f55a16734dcd185af5b8dc   \n2327      ljvmiranda921/pyswarms  50f6e159f0cc50f906a1c2202ddbfe63a212c1ac   \n\n              commit_date                                             author  \\\n353   2015-08-01 13:57:16  Cameron Davidson-Pilon <cam.davidson.pilon@gma...   \n482   2019-01-31 20:07:46  Cameron Davidson-Pilon <cam.davidson.pilon@gma...   \n501   2019-03-21 12:34:34  Cameron Davidson-Pilon <cam.davidson.pilon@gma...   \n516   2019-05-16 08:35:35  Cameron Davidson-Pilon <cam.davidson.pilon@gma...   \n853   2019-04-21 19:19:06  torzdf <36920800+torzdf@users.noreply.github.com>   \n1246  2020-08-11 08:56:17  Nick Bristow <36126654+AtelierLibre@users.nore...   \n2327  2019-02-09 04:47:50    Aaron <39431903+whzup@users.noreply.github.com>   \n\n                                                  files  \\\n353   ['.travis.yml', 'CHANGELOG.md', 'docs/Examples...   \n482   ['.prospector.yaml', 'CHANGELOG.md', 'docs/Exa...   \n501   ['CHANGELOG.md', 'docs/Changelog.rst', 'docs/E...   \n516   ['CHANGELOG.md', 'Makefile', 'docs/Changelog.r...   \n853   ['.gitignore', 'lib/cli.py', 'lib/config.py', ...   \n1246  ['environments/docker/Dockerfile', 'environmen...   \n2327  ['docs/api/_pyswarms.backend.rst', 'docs/api/p...   \n\n                                          changed_lines           issues  \\\n353   [2, 10, 21, 21, 6, 4, 2, 2, 2, 2, 4, 2, 16, 2,...            [170]   \n482   [5, 14, 38, 14, 12, 74, 146, 2, 31, 11, 18, 66...            [618]   \n501   [18, 191, 41, 87, 2, 8, 54, 49, 2, 167, 103, 1...  [672, 671, 673]   \n516   [13, 11, 224, 2, 16, 2, 2, 1, 122, 307, 551, 2...            [715]   \n853   [1, 342, 30, 202, 120, 2, 3, 6, 216, 31, 20, 5...            [703]   \n1246            [17, 6, 183, 5, 25, 4, 164, 4, 140, 91]             [19]   \n2327  [1, 15, 17, 3, 536, 64, 31, 36, 36, 32, 12, 20...       [238, 237]   \n\n      is_ml phases  file_entropy  line_entropy            open_date  \\\n353    True    NaN      0.919340      0.045754  2015-06-24 02:43:00   \n482    True    NaN      0.952173      0.208786  2019-01-28 01:57:48   \n501    True    NaN      0.952173      0.160658  2019-03-19 11:22:23   \n516    True    NaN      0.952173      0.278338  2019-05-02 18:20:00   \n853    True    NaN      0.926474      0.322994  2019-04-09 15:42:16   \n1246   True    NaN      0.937500      0.317626  2016-12-15 12:54:24   \n2327   True    NaN      0.910964      0.362021  2018-08-26 13:11:51   \n\n               close_date  n_comments  words_mean  day_to_fix  \n353   2015-08-01 17:57:16    0.000000    0.000000          38  \n482   2019-02-01 01:07:48    0.000000    0.000000           3  \n501   2019-03-21 16:34:35    1.333333    3.083333           2  \n516   2019-05-16 12:35:36    0.000000    0.000000          13  \n853   2019-04-21 19:19:07    0.000000    0.000000          12  \n1246  2016-12-29 13:05:10    3.000000  119.000000          14  \n2327  2019-02-09 03:47:51   18.500000   58.731183         166  ",
      "text/html": "<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>project</th>\n      <th>commit</th>\n      <th>commit_date</th>\n      <th>author</th>\n      <th>files</th>\n      <th>changed_lines</th>\n      <th>issues</th>\n      <th>is_ml</th>\n      <th>phases</th>\n      <th>file_entropy</th>\n      <th>line_entropy</th>\n      <th>open_date</th>\n      <th>close_date</th>\n      <th>n_comments</th>\n      <th>words_mean</th>\n      <th>day_to_fix</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>353</th>\n      <td>CamDavidsonPilon/lifelines</td>\n      <td>d2be007e7aea254670e244942d07df547beee603</td>\n      <td>2015-08-01 13:57:16</td>\n      <td>Cameron Davidson-Pilon &lt;cam.davidson.pilon@gma...</td>\n      <td>['.travis.yml', 'CHANGELOG.md', 'docs/Examples...</td>\n      <td>[2, 10, 21, 21, 6, 4, 2, 2, 2, 2, 4, 2, 16, 2,...</td>\n      <td>[170]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.919340</td>\n      <td>0.045754</td>\n      <td>2015-06-24 02:43:00</td>\n      <td>2015-08-01 17:57:16</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>38</td>\n    </tr>\n    <tr>\n      <th>482</th>\n      <td>CamDavidsonPilon/lifelines</td>\n      <td>55a81f179b87e6370bd0fa5ff573c9ab8582d91f</td>\n      <td>2019-01-31 20:07:46</td>\n      <td>Cameron Davidson-Pilon &lt;cam.davidson.pilon@gma...</td>\n      <td>['.prospector.yaml', 'CHANGELOG.md', 'docs/Exa...</td>\n      <td>[5, 14, 38, 14, 12, 74, 146, 2, 31, 11, 18, 66...</td>\n      <td>[618]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.952173</td>\n      <td>0.208786</td>\n      <td>2019-01-28 01:57:48</td>\n      <td>2019-02-01 01:07:48</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>3</td>\n    </tr>\n    <tr>\n      <th>501</th>\n      <td>CamDavidsonPilon/lifelines</td>\n      <td>f7ca1239b150ee5928d7e2937a2551f40272b4f7</td>\n      <td>2019-03-21 12:34:34</td>\n      <td>Cameron Davidson-Pilon &lt;cam.davidson.pilon@gma...</td>\n      <td>['CHANGELOG.md', 'docs/Changelog.rst', 'docs/E...</td>\n      <td>[18, 191, 41, 87, 2, 8, 54, 49, 2, 167, 103, 1...</td>\n      <td>[672, 671, 673]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.952173</td>\n      <td>0.160658</td>\n      <td>2019-03-19 11:22:23</td>\n      <td>2019-03-21 16:34:35</td>\n      <td>1.333333</td>\n      <td>3.083333</td>\n      <td>2</td>\n    </tr>\n    <tr>\n      <th>516</th>\n      <td>CamDavidsonPilon/lifelines</td>\n      <td>af56f7a7925feeb36df59ac4bc30e0b8e78b2c12</td>\n      <td>2019-05-16 08:35:35</td>\n      <td>Cameron Davidson-Pilon &lt;cam.davidson.pilon@gma...</td>\n      <td>['CHANGELOG.md', 'Makefile', 'docs/Changelog.r...</td>\n      <td>[13, 11, 224, 2, 16, 2, 2, 1, 122, 307, 551, 2...</td>\n      <td>[715]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.952173</td>\n      <td>0.278338</td>\n      <td>2019-05-02 18:20:00</td>\n      <td>2019-05-16 12:35:36</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>13</td>\n    </tr>\n    <tr>\n      <th>853</th>\n      <td>deepfakes/faceswap</td>\n      <td>27a685383ec2fc1df2f60a5648d95b722b2f4e94</td>\n      <td>2019-04-21 19:19:06</td>\n      <td>torzdf &lt;36920800+torzdf@users.noreply.github.com&gt;</td>\n      <td>['.gitignore', 'lib/cli.py', 'lib/config.py', ...</td>\n      <td>[1, 342, 30, 202, 120, 2, 3, 6, 216, 31, 20, 5...</td>\n      <td>[703]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.926474</td>\n      <td>0.322994</td>\n      <td>2019-04-09 15:42:16</td>\n      <td>2019-04-21 19:19:07</td>\n      <td>0.000000</td>\n      <td>0.000000</td>\n      <td>12</td>\n    </tr>\n    <tr>\n      <th>1246</th>\n      <td>gboeing/osmnx</td>\n      <td>2d1756bc161061d0d3f55a16734dcd185af5b8dc</td>\n      <td>2020-08-11 08:56:17</td>\n      <td>Nick Bristow &lt;36126654+AtelierLibre@users.nore...</td>\n      <td>['environments/docker/Dockerfile', 'environmen...</td>\n      <td>[17, 6, 183, 5, 25, 4, 164, 4, 140, 91]</td>\n      <td>[19]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.937500</td>\n      <td>0.317626</td>\n      <td>2016-12-15 12:54:24</td>\n      <td>2016-12-29 13:05:10</td>\n      <td>3.000000</td>\n      <td>119.000000</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>2327</th>\n      <td>ljvmiranda921/pyswarms</td>\n      <td>50f6e159f0cc50f906a1c2202ddbfe63a212c1ac</td>\n      <td>2019-02-09 04:47:50</td>\n      <td>Aaron &lt;39431903+whzup@users.noreply.github.com&gt;</td>\n      <td>['docs/api/_pyswarms.backend.rst', 'docs/api/p...</td>\n      <td>[1, 15, 17, 3, 536, 64, 31, 36, 36, 32, 12, 20...</td>\n      <td>[238, 237]</td>\n      <td>True</td>\n      <td>NaN</td>\n      <td>0.910964</td>\n      <td>0.362021</td>\n      <td>2018-08-26 13:11:51</td>\n      <td>2019-02-09 03:47:51</td>\n      <td>18.500000</td>\n      <td>58.731183</td>\n      <td>166</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data[(data['file_entropy'] > 0.9) & (data['file_entropy'] < 0.96) & (data['is_ml'])]\n",
    "\n"
   ],
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    }
   }
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}