{
  "id": 89972,
  "title": "Segments of training data plotted separately",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/89972",
  "author_name": "Ammar Alyousfi",
  "post_date": "2019-04-19T07:44:01.301000",
  "votes": 2,
  "comment_count": 0,
  "views": 0,
  "content": "<p><strong>Hello everyone,</strong> 🙋🏻‍♂️</p>\n\n<p>In a <a href=\"https://www.kaggle.com/c/LANL-Earthquake-Prediction/discussion/89766#latest-518693\">previous post</a>, the whole training data was plotted in one graph. After that, I plotted each segment of training data in a separate plot where all plots have the <em>same scale</em> to enable comparison. Each segment contains a fault (where <code>time_of_failure</code> = 0).</p>\n\n<p>For example, the first segment (fault) is shown in the image below:</p>\n\n<p><img src=\"https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part1.jpg\" alt=\"\"></p>\n\n<p>The second segment:</p>\n\n<p><img src=\"https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part2.jpg\" alt=\"\"></p>\n\n<h3>You can find all plots on this Github page:</h3>\n\n<p>📈 <a href=\"https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md\"><strong>https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md</strong></a></p>\n\n<p>Hope you find them useful :)</p>",
  "messages": [
    {
      "id": 519567,
      "postDate": "2019-04-19T07:44:01.300Z",
      "content": "<p><strong>Hello everyone,</strong> 🙋🏻‍♂️</p>\n\n<p>In a <a href=\"https://www.kaggle.com/c/LANL-Earthquake-Prediction/discussion/89766#latest-518693\">previous post</a>, the whole training data was plotted in one graph. After that, I plotted each segment of training data in a separate plot where all plots have the <em>same scale</em> to enable comparison. Each segment contains a fault (where <code>time_of_failure</code> = 0).</p>\n\n<p>For example, the first segment (fault) is shown in the image below:</p>\n\n<p><img src=\"https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part1.jpg\" alt=\"\"></p>\n\n<p>The second segment:</p>\n\n<p><img src=\"https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part2.jpg\" alt=\"\"></p>\n\n<h3>You can find all plots on this Github page:</h3>\n\n<p>📈 <a href=\"https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md\"><strong>https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md</strong></a></p>\n\n<p>Hope you find them useful :)</p>",
      "rawMarkdown": "**Hello everyone,** 🙋🏻‍♂️\n\nIn a [previous post](https://www.kaggle.com/c/LANL-Earthquake-Prediction/discussion/89766#latest-518693), the whole training data was plotted in one graph. After that, I plotted each segment of training data in a separate plot where all plots have the *same scale* to enable comparison. Each segment contains a fault (where `time_of_failure` = 0).\n\nFor example, the first segment (fault) is shown in the image below:\n\n![](https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part1.jpg)\n\nThe second segment:\n\n![](https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part2.jpg)\n\n### You can find all plots on this Github page:\n\n📈 [**https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md**](https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md)\n\nHope you find them useful :)",
      "votes": 2
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "519567": "**Hello everyone,** 🙋🏻‍♂️\n\nIn a [previous post](https://www.kaggle.com/c/LANL-Earthquake-Prediction/discussion/89766#latest-518693), the whole training data was plotted in one graph. After that, I plotted each segment of training data in a separate plot where all plots have the *same scale* to enable comparison. Each segment contains a fault (where `time_of_failure` = 0).\n\nFor example, the first segment (fault) is shown in the image below:\n\n![](https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part1.jpg)\n\nThe second segment:\n\n![](https://raw.githubusercontent.com/ammar1y/LANL-Earthquake-Prediction-Competition/master/images/train_segments/part2.jpg)\n\n### You can find all plots on this Github page:\n\n📈 [**https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md**](https://github.com/ammar1y/LANL-Earthquake-Prediction-Competition/blob/master/Training-Data%20Segmented.md)\n\nHope you find them useful :)"
  }
}