{
  "id": 164837,
  "title": "This leaderboard is calculated with approximately 1% of the test data.",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/164837",
  "author_name": "Andrey Lukyanenko",
  "post_date": "2020-07-07T17:49:38.614000",
  "votes": 18,
  "comment_count": 13,
  "views": 0,
  "content": "<p>1%? Is this enough?</p>",
  "messages": [
    {
      "id": 919096,
      "postDate": "2020-07-07T17:49:38.613Z",
      "content": "<p>1%? Is this enough?</p>",
      "rawMarkdown": "1%? Is this enough?",
      "votes": 18
    },
    {
      "id": 919235,
      "postDate": "2020-07-07T19:08:44.980Z",
      "content": "<p>This is caused by all the <code>Ignored</code> rows being counted as \"not public\" (we show it this way in normal competitions to not leak the actual number of ignored rows). I'll manually override this to show the actual value, which is approximately 15%. </p>",
      "rawMarkdown": "This is caused by all the `Ignored` rows being counted as \"not public\" (we show it this way in normal competitions to not leak the actual number of ignored rows). I'll manually override this to show the actual value, which is approximately 15%. ",
      "votes": 13
    },
    {
      "id": 919114,
      "postDate": "2020-07-07T18:02:56.033Z",
      "content": "<p>Quite enough for a legendary shakeup 😑 </p>",
      "rawMarkdown": "Quite enough for a legendary shakeup 😑 ",
      "votes": 8,
      "replies": [
        {
          "id": 919224,
          "postDate": "2020-07-07T19:04:57.210Z",
          "content": "<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F4051350%2Fa1a0f631ff9e44ca8c9aad7246480f29%2Fpopcorn.jpg?generation=1594148675265352&amp;alt=media\" alt=\"\"></p>",
          "rawMarkdown": "![](https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F4051350%2Fa1a0f631ff9e44ca8c9aad7246480f29%2Fpopcorn.jpg?generation=1594148675265352&amp;alt=media)\n",
          "votes": 14
        }
      ]
    },
    {
      "id": 919129,
      "postDate": "2020-07-07T18:14:25.320Z",
      "content": "<blockquote>\n  <p>The provided test set is a small representative set of files (copied from the training set)</p>\n</blockquote>\n\n<p>More like 0% :) </p>",
      "rawMarkdown": "&gt; The provided test set is a small representative set of files (copied from the training set)\n\nMore like 0% :) ",
      "votes": 2
    },
    {
      "id": 920121,
      "postDate": "2020-07-08T11:01:17.420Z",
      "content": "<p>yes, they should use at least use 10% for test.</p>",
      "rawMarkdown": "yes, they should use at least use 10% for test.",
      "votes": 1,
      "replies": [
        {
          "id": 920130,
          "postDate": "2020-07-08T11:14:36.957Z",
          "content": "<p>15% are used, please see the other comments below.</p>",
          "rawMarkdown": "15% are used, please see the other comments below."
        },
        {
          "id": 920278,
          "postDate": "2020-07-08T13:21:32.367Z",
          "content": "<p><a href=\"/ahmedhshahin\">@ahmedhshahin</a> sorry for the comment, I misinterpreted the evaluation metrics part.</p>",
          "rawMarkdown": "@ahmedhshahin sorry for the comment, I misinterpreted the evaluation metrics part."
        },
        {
          "id": 920372,
          "postDate": "2020-07-08T14:49:34.673Z",
          "content": "<p>No problem, you are welcome!</p>",
          "rawMarkdown": "No problem, you are welcome!"
        }
      ]
    },
    {
      "id": 919135,
      "postDate": "2020-07-07T18:16:38.207Z",
      "content": "<p>Woah they should really change that\nEven 20% is something we can work with\n1% is way too less</p>",
      "rawMarkdown": "Woah they should really change that\nEven 20% is something we can work with\n1% is way too less",
      "votes": 1
    },
    {
      "id": 919144,
      "postDate": "2020-07-07T18:20:50.503Z",
      "content": "<p>Without checking details, maybe the (uncertainty) metric is very prone to probing?</p>",
      "rawMarkdown": "Without checking details, maybe the (uncertainty) metric is very prone to probing?",
      "votes": 1
    },
    {
      "id": 919467,
      "postDate": "2020-07-07T22:04:15.947Z",
      "content": "<p>1% -&gt; 15%,\n99% -&gt; 85%</p>",
      "rawMarkdown": "1% -&gt; 15%,\n99% -&gt; 85%"
    },
    {
      "id": 919102,
      "postDate": "2020-07-07T17:54:49.227Z",
      "content": "<p>I felt the same. Way too low right?</p>",
      "rawMarkdown": "I felt the same. Way too low right?"
    },
    {
      "id": 919115,
      "postDate": "2020-07-07T18:03:27.187Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 919235,
      "author_name": "Will Cukierski",
      "author_url": "",
      "post_date": "2020-07-07T19:08:44.980000",
      "content": "<p>This is caused by all the <code>Ignored</code> rows being counted as \"not public\" (we show it this way in normal competitions to not leak the actual number of ignored rows). I'll manually override this to show the actual value, which is approximately 15%. </p>",
      "votes": 13,
      "replies": []
    },
    {
      "id": 919114,
      "author_name": "Ruslan Vdovychenko",
      "author_url": "",
      "post_date": "2020-07-07T18:02:56.033000",
      "content": "<p>Quite enough for a legendary shakeup 😑 </p>",
      "votes": 8,
      "replies": [
        {
          "id": 919224,
          "author_name": "Carl McBride Ellis",
          "author_url": "",
          "post_date": "2020-07-07T19:04:57.210000",
          "content": "<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-user-content/o/inbox%2F4051350%2Fa1a0f631ff9e44ca8c9aad7246480f29%2Fpopcorn.jpg?generation=1594148675265352&amp;alt=media\" alt=\"\"></p>",
          "votes": 14,
          "replies": []
        }
      ]
    },
    {
      "id": 919129,
      "author_name": "Theo Viel",
      "author_url": "",
      "post_date": "2020-07-07T18:14:25.320000",
      "content": "<blockquote>\n  <p>The provided test set is a small representative set of files (copied from the training set)</p>\n</blockquote>\n\n<p>More like 0% :) </p>",
      "votes": 2,
      "replies": []
    },
    {
      "id": 920121,
      "author_name": "MithileshThakkar",
      "author_url": "",
      "post_date": "2020-07-08T11:01:17.420000",
      "content": "<p>yes, they should use at least use 10% for test.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 920130,
          "author_name": "Ahmed Shahin",
          "author_url": "",
          "post_date": "2020-07-08T11:14:36.957000",
          "content": "<p>15% are used, please see the other comments below.</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 920278,
          "author_name": "MithileshThakkar",
          "author_url": "",
          "post_date": "2020-07-08T13:21:32.367000",
          "content": "<p><a href=\"/ahmedhshahin\">@ahmedhshahin</a> sorry for the comment, I misinterpreted the evaluation metrics part.</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 920372,
          "author_name": "Ahmed Shahin",
          "author_url": "",
          "post_date": "2020-07-08T14:49:34.673000",
          "content": "<p>No problem, you are welcome!</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 919135,
      "author_name": "Aryan Pandey",
      "author_url": "",
      "post_date": "2020-07-07T18:16:38.207000",
      "content": "<p>Woah they should really change that\nEven 20% is something we can work with\n1% is way too less</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 919144,
      "author_name": "Psi",
      "author_url": "",
      "post_date": "2020-07-07T18:20:50.503000",
      "content": "<p>Without checking details, maybe the (uncertainty) metric is very prone to probing?</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 919467,
      "author_name": "Heroseo",
      "author_url": "",
      "post_date": "2020-07-07T22:04:15.947000",
      "content": "<p>1% -&gt; 15%,\n99% -&gt; 85%</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 919102,
      "author_name": "Ameer Ul Islam",
      "author_url": "",
      "post_date": "2020-07-07T17:54:49.227000",
      "content": "<p>I felt the same. Way too low right?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 919115,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-07-07T18:03:27.187000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "919096": "1%? Is this enough?",
    "919235": "This is caused by all the `Ignored` rows being counted as \"not public\" (we show it this way in normal competitions to not leak the actual number of ignored rows). I'll manually override this to show the actual value, which is approximately 15%. ",
    "919114": "Quite enough for a legendary shakeup 😑 ",
    "919129": "&gt; The provided test set is a small representative set of files (copied from the training set)\n\nMore like 0% :) ",
    "920121": "yes, they should use at least use 10% for test.",
    "919135": "Woah they should really change that\nEven 20% is something we can work with\n1% is way too less",
    "919144": "Without checking details, maybe the (uncertainty) metric is very prone to probing?",
    "919467": "1% -&gt; 15%,\n99% -&gt; 85%",
    "919102": "I felt the same. Way too low right?",
    "919115": ""
  }
}