{
  "id": 337270,
  "title": "\"Submission Scoring Error\" for you submission",
  "url": "/competitions/mayo-clinic-strip-ai/discussion/337270",
  "author_name": "Jirka",
  "post_date": "2022-07-15T08:25:09.426000",
  "votes": 11,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I was playing around and found a few corner cases which are covered under this error but could be hard to decode its meaning or what needs to be fixed… So the cases I found so far are:</p>\n<ol>\n<li>duplicated in <code>patient_id</code><br>\n-&gt; which can be simply resolved with <code>sub.groupby(\"patient_id\").mean()</code></li>\n<li>number of digits in your submission (need one more sub for conformation)<br>\n-&gt; which you can simply round <code>sub[[\"CE\", \"LAA\"]].round(6)</code></li>\n</ol>\n<p>see dummy submission over here:</p>\n<blockquote>\n  <p><a href=\"https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-fake-stats-prediction\" target=\"_blank\">https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-fake-stats-prediction</a></p>\n</blockquote>",
  "messages": [
    {
      "id": 1856270,
      "postDate": "2022-07-15T08:25:09.427Z",
      "content": "<p>I was playing around and found a few corner cases which are covered under this error but could be hard to decode its meaning or what needs to be fixed… So the cases I found so far are:</p>\n<ol>\n<li>duplicated in <code>patient_id</code><br>\n-&gt; which can be simply resolved with <code>sub.groupby(\"patient_id\").mean()</code></li>\n<li>number of digits in your submission (need one more sub for conformation)<br>\n-&gt; which you can simply round <code>sub[[\"CE\", \"LAA\"]].round(6)</code></li>\n</ol>\n<p>see dummy submission over here:</p>\n<blockquote>\n  <p><a href=\"https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-fake-stats-prediction\" target=\"_blank\">https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-fake-stats-prediction</a></p>\n</blockquote>",
      "rawMarkdown": "I was playing around and found a few corner cases which are covered under this error but could be hard to decode its meaning or what needs to be fixed... So the cases I found so far are:\n\n1. duplicated in `patient_id`\n -> which can be simply resolved with `sub.groupby(\"patient_id\").mean()`\n2. number of digits in your submission (need one more sub for conformation)\n -> which you can simply round `sub[[\"CE\", \"LAA\"]].round(6)`\n\nsee dummy submission over here:\n> https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-fake-stats-prediction",
      "votes": 11
    },
    {
      "id": 1962143,
      "postDate": "2022-09-29T14:08:43.280Z",
      "content": "<p>I have also encountered an unexpected one, the evaluation tab says ours probabilities do not have to sum to 1, but I tested a submission where they do not and it resulted in an error..</p>\n<p><a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a> bugs?</p>\n<blockquote>\n  <p>number of digits in your submission (need one more sub for conformation)</p>\n</blockquote>\n<p>Did you manage to confirm this one? edit: this fixed a submission for me.. Thanks</p>",
      "rawMarkdown": "I have also encountered an unexpected one, the evaluation tab says ours probabilities do not have to sum to 1, but I tested a submission where they do not and it resulted in an error..\n\n@sohier bugs?\n\n>number of digits in your submission (need one more sub for conformation)\n\nDid you manage to confirm this one? edit: this fixed a submission for me.. Thanks"
    }
  ],
  "comments": [
    {
      "id": 1962143,
      "author_name": "JM",
      "author_url": "",
      "post_date": "2022-09-29T14:08:43.280000",
      "content": "<p>I have also encountered an unexpected one, the evaluation tab says ours probabilities do not have to sum to 1, but I tested a submission where they do not and it resulted in an error..</p>\n<p><a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a> bugs?</p>\n<blockquote>\n  <p>number of digits in your submission (need one more sub for conformation)</p>\n</blockquote>\n<p>Did you manage to confirm this one? edit: this fixed a submission for me.. Thanks</p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1856270": "I was playing around and found a few corner cases which are covered under this error but could be hard to decode its meaning or what needs to be fixed... So the cases I found so far are:\n\n1. duplicated in `patient_id`\n -> which can be simply resolved with `sub.groupby(\"patient_id\").mean()`\n2. number of digits in your submission (need one more sub for conformation)\n -> which you can simply round `sub[[\"CE\", \"LAA\"]].round(6)`\n\nsee dummy submission over here:\n> https://www.kaggle.com/code/jirkaborovec/bloodclots-classif-fake-stats-prediction",
    "1962143": "I have also encountered an unexpected one, the evaluation tab says ours probabilities do not have to sum to 1, but I tested a submission where they do not and it resulted in an error..\n\n@sohier bugs?\n\n>number of digits in your submission (need one more sub for conformation)\n\nDid you manage to confirm this one? edit: this fixed a submission for me.. Thanks"
  }
}