{
  "id": 69246,
  "title": "My mistake",
  "url": "/competitions/airbus-ship-detection/discussion/69246",
  "author_name": "Emil",
  "post_date": "2018-10-21T22:42:36.436000",
  "votes": 7,
  "comment_count": 2,
  "views": 0,
  "content": "<p>UPD. It was my mistake, I used data from the first stage and found the same leak again :c\nEverything is fine, there is no leak in new data</p>\n\n<p>I examined the new test data v2 and it looks like there is a leak again. I managed to fing at least 35% of test data in train. Color map in images below is: <br>\nRed - test <br>\nGreen - train <br>\nPurple - train &amp; test</p>\n\n<p><a href=\"http://pasteboard.co/HJxuC0S.png\">example 1</a> <br>\n<a href=\"http://pasteboard.co/HJxvkED.png\">example 2</a></p>",
  "messages": [
    {
      "id": 407864,
      "postDate": "2018-10-21T22:42:36.437Z",
      "content": "<p>UPD. It was my mistake, I used data from the first stage and found the same leak again :c\nEverything is fine, there is no leak in new data</p>\n\n<p>I examined the new test data v2 and it looks like there is a leak again. I managed to fing at least 35% of test data in train. Color map in images below is: <br>\nRed - test <br>\nGreen - train <br>\nPurple - train &amp; test</p>\n\n<p><a href=\"http://pasteboard.co/HJxuC0S.png\">example 1</a> <br>\n<a href=\"http://pasteboard.co/HJxvkED.png\">example 2</a></p>",
      "rawMarkdown": "UPD. It was my mistake, I used data from the first stage and found the same leak again :c\nEverything is fine, there is no leak in new data\n\n\nI examined the new test data v2 and it looks like there is a leak again. I managed to fing at least 35% of test data in train. Color map in images below is:  \nRed - test  \nGreen - train  \nPurple - train &amp; test\n\n[example 1][1]  \n[example 2][2]\n\n\n  [1]: http://pasteboard.co/HJxuC0S.png\n  [2]: http://pasteboard.co/HJxvkED.png",
      "votes": 7
    },
    {
      "id": 407956,
      "postDate": "2018-10-22T04:06:46.257Z",
      "content": "<p>Thank you for sharing your findings. This competition may become puzzle game again :(</p>",
      "rawMarkdown": "Thank you for sharing your findings. This competition may become puzzle game again :(",
      "votes": 1
    },
    {
      "id": 407960,
      "postDate": "2018-10-22T04:16:48.867Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 407956,
      "author_name": "toshi_k",
      "author_url": "",
      "post_date": "2018-10-22T04:06:46.257000",
      "content": "<p>Thank you for sharing your findings. This competition may become puzzle game again :(</p>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 407960,
      "author_name": "",
      "author_url": "",
      "post_date": "2018-10-22T04:16:48.867000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "407864": "UPD. It was my mistake, I used data from the first stage and found the same leak again :c\nEverything is fine, there is no leak in new data\n\n\nI examined the new test data v2 and it looks like there is a leak again. I managed to fing at least 35% of test data in train. Color map in images below is:  \nRed - test  \nGreen - train  \nPurple - train &amp; test\n\n[example 1][1]  \n[example 2][2]\n\n\n  [1]: http://pasteboard.co/HJxuC0S.png\n  [2]: http://pasteboard.co/HJxvkED.png",
    "407956": "Thank you for sharing your findings. This competition may become puzzle game again :(",
    "407960": ""
  }
}