{
  "id": 539424,
  "title": "test accelerometer has only 2 files",
  "url": "/competitions/child-mind-institute-problematic-internet-use/discussion/539424",
  "author_name": "shelley klompus",
  "post_date": "2024-10-08T19:50:18.482000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>I worked a lot on the accelerometer data but I see that the tset has only 2 files meaning 2 pretisipents </p>\n<ul>\n<li>it looks to me far too small for any accuracy analysis</li>\n<li>it looke too small to impute from them to the test data</li>\n<li>imputing using also train seems risking in data leakage</li>\n<li>if I work only on the tabular data it seems to miss the point <br>\nany ideas? </li>\n</ul>",
  "messages": [
    {
      "id": 3012253,
      "postDate": "2024-10-08T19:50:18.483Z",
      "content": "<p>I worked a lot on the accelerometer data but I see that the tset has only 2 files meaning 2 pretisipents </p>\n<ul>\n<li>it looks to me far too small for any accuracy analysis</li>\n<li>it looke too small to impute from them to the test data</li>\n<li>imputing using also train seems risking in data leakage</li>\n<li>if I work only on the tabular data it seems to miss the point <br>\nany ideas? </li>\n</ul>",
      "rawMarkdown": "I worked a lot on the accelerometer data but I see that the tset has only 2 files meaning 2 pretisipents \n- it looks to me far too small for any accuracy analysis\n- it looke too small to impute from them to the test data\n- imputing using also train seems risking in data leakage\n- if I work only on the tabular data it seems to miss the point \nany ideas? ",
      "votes": 1
    },
    {
      "id": 3012431,
      "postDate": "2024-10-09T03:18:51.650Z",
      "content": "<p>This is a dummy test dataset used for code and syntax check <a href=\"https://www.kaggle.com/shelleyklompus\" target=\"_blank\">@shelleyklompus</a> <br>\nThe actual test set is hidden and will be available secretly upon leaderboard submission </p>",
      "rawMarkdown": "This is a dummy test dataset used for code and syntax check @shelleyklompus \nThe actual test set is hidden and will be available secretly upon leaderboard submission ",
      "replies": [
        {
          "id": 3012459,
          "postDate": "2024-10-09T04:18:15.003Z",
          "content": "<p>thanks! that makes much more sence now :)</p>",
          "rawMarkdown": "thanks! that makes much more sence now :)"
        }
      ]
    },
    {
      "id": 3012269,
      "postDate": "2024-10-08T20:20:02.837Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 3012431,
      "author_name": "Ravi Ramakrishnan",
      "author_url": "",
      "post_date": "2024-10-09T03:18:51.650000",
      "content": "<p>This is a dummy test dataset used for code and syntax check <a href=\"https://www.kaggle.com/shelleyklompus\" target=\"_blank\">@shelleyklompus</a> <br>\nThe actual test set is hidden and will be available secretly upon leaderboard submission </p>",
      "votes": 0,
      "replies": [
        {
          "id": 3012459,
          "author_name": "shelley klompus",
          "author_url": "",
          "post_date": "2024-10-09T04:18:15.003000",
          "content": "<p>thanks! that makes much more sence now :)</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 3012269,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-10-08T20:20:02.837000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3012253": "I worked a lot on the accelerometer data but I see that the tset has only 2 files meaning 2 pretisipents \n- it looks to me far too small for any accuracy analysis\n- it looke too small to impute from them to the test data\n- imputing using also train seems risking in data leakage\n- if I work only on the tabular data it seems to miss the point \nany ideas? ",
    "3012431": "This is a dummy test dataset used for code and syntax check @shelleyklompus \nThe actual test set is hidden and will be available secretly upon leaderboard submission ",
    "3012269": ""
  }
}