{
  "id": 93025,
  "title": "Interesting feature for time series data",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/93025",
  "author_name": "",
  "post_date": "2019-05-22T12:40:03.273165700Z",
  "votes": 4,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Even though I subtract mean during features calculation, I am able to derive such feature, which is alarming because test data can look much different in terms of some features.</p>\n\n<p><img src=\"https://storage.googleapis.com/kaggle-forum-message-attachments/535185/13266/feat.png\" alt=\"upward trend\"></p>",
  "messages": [
    {
      "id": "535185",
      "postDate": "05/22/2019 12:40:03",
      "content": "<p>Even though I subtract mean during features calculation, I am able to derive such feature, which is alarming because test data can look much different in terms of some features.</p>\n\n<p><img src=\"https://storage.googleapis.com/kaggle-forum-message-attachments/535185/13266/feat.png\" alt=\"upward trend\"></p>",
      "rawMarkdown": "Even though I subtract mean during features calculation, I am able to derive such feature, which is alarming because test data can look much different in terms of some features.\n\n![upward trend](https://storage.googleapis.com/kaggle-forum-message-attachments/535185/13266/feat.png)",
      "votes": null
    },
    {
      "id": "535194",
      "postDate": "05/22/2019 13:16:18",
      "content": "<p>Your feature hace a clear upward trend.  Are your test data values even higher?  </p>",
      "rawMarkdown": "Your feature hace a clear upward trend.  Are your test data values even higher?",
      "votes": null
    },
    {
      "id": "535224",
      "postDate": "05/22/2019 14:20:23",
      "content": "<p>I have many features that don't seem to have such a problem but a few have.\n<img src=\"https://storage.googleapis.com/kaggle-forum-message-attachments/535224/13269/feat.png\" alt=\"train and test\"> </p>",
      "rawMarkdown": "I have many features that don't seem to have such a problem but a few have.\n![train and test](https://storage.googleapis.com/kaggle-forum-message-attachments/535224/13269/feat.png)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 535194,
      "author_name": "cpmpml",
      "author_url": "",
      "post_date": "05/22/2019 13:16:18",
      "content": "<p>Your feature hace a clear upward trend.  Are your test data values even higher?  </p>",
      "votes": null,
      "replies": [
        {
          "id": 535224,
          "author_name": "davids1992",
          "author_url": "",
          "post_date": "05/22/2019 14:20:23",
          "content": "<p>I have many features that don't seem to have such a problem but a few have.\n<img src=\"https://storage.googleapis.com/kaggle-forum-message-attachments/535224/13269/feat.png\" alt=\"train and test\"> </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "535185": "Even though I subtract mean during features calculation, I am able to derive such feature, which is alarming because test data can look much different in terms of some features.\n\n![upward trend](https://storage.googleapis.com/kaggle-forum-message-attachments/535185/13266/feat.png)",
    "535194": "Your feature hace a clear upward trend.  Are your test data values even higher?",
    "535224": "I have many features that don't seem to have such a problem but a few have.\n![train and test](https://storage.googleapis.com/kaggle-forum-message-attachments/535224/13269/feat.png)"
  },
  "source": "meta"
}