{
  "id": 444673,
  "title": "Why submission.csv is empty?",
  "url": "/competitions/child-mind-institute-detect-sleep-states/discussion/444673",
  "author_name": "",
  "post_date": "2023-10-03T05:05:56.089742700Z",
  "votes": 2,
  "comment_count": 2,
  "views": 0,
  "content": "<p>I am new to kaggle and this is my first competition. I wonder why the output of the public code (submission.csv) is empty but still get good score. </p>",
  "messages": [
    {
      "id": "2465559",
      "postDate": "10/03/2023 05:05:56",
      "content": "<p>I am new to kaggle and this is my first competition. I wonder why the output of the public code (submission.csv) is empty but still get good score. </p>",
      "rawMarkdown": "I am new to kaggle and this is my first competition. I wonder why the output of the public code (submission.csv) is empty but still get good score.",
      "votes": null
    },
    {
      "id": "2465731",
      "postDate": "10/03/2023 08:31:06",
      "content": "<p>Hi,</p>\n<p>This is because when they are running their code without submission, they are using the dummy test dataset. The test dataset may not even have a onset or awake event since it is too short, which result in a empty submission.csv. when you submit your code, the code will be re-run on the public LB test data, which will most likely give a non-empty submission.csv. If you are unsure with whether your model is working, do a train test split on your train data, and test if the model gives any result.</p>\n<p>Hope it helps!</p>",
      "rawMarkdown": "Hi,\n\nThis is because when they are running their code without submission, they are using the dummy test dataset. The test dataset may not even have a onset or awake event since it is too short, which result in a empty submission.csv. when you submit your code, the code will be re-run on the public LB test data, which will most likely give a non-empty submission.csv. If you are unsure with whether your model is working, do a train test split on your train data, and test if the model gives any result.\n\nHope it helps!",
      "votes": null
    },
    {
      "id": "2465760",
      "postDate": "10/03/2023 09:06:04",
      "content": "<p>thanks a lot</p>",
      "rawMarkdown": "thanks a lot",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2465731,
      "author_name": "yuchen2066",
      "author_url": "",
      "post_date": "10/03/2023 08:31:06",
      "content": "<p>Hi,</p>\n<p>This is because when they are running their code without submission, they are using the dummy test dataset. The test dataset may not even have a onset or awake event since it is too short, which result in a empty submission.csv. when you submit your code, the code will be re-run on the public LB test data, which will most likely give a non-empty submission.csv. If you are unsure with whether your model is working, do a train test split on your train data, and test if the model gives any result.</p>\n<p>Hope it helps!</p>",
      "votes": null,
      "replies": [
        {
          "id": 2465760,
          "author_name": "met4physics",
          "author_url": "",
          "post_date": "10/03/2023 09:06:04",
          "content": "<p>thanks a lot</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2465559": "I am new to kaggle and this is my first competition. I wonder why the output of the public code (submission.csv) is empty but still get good score.",
    "2465731": "Hi,\n\nThis is because when they are running their code without submission, they are using the dummy test dataset. The test dataset may not even have a onset or awake event since it is too short, which result in a empty submission.csv. when you submit your code, the code will be re-run on the public LB test data, which will most likely give a non-empty submission.csv. If you are unsure with whether your model is working, do a train test split on your train data, and test if the model gives any result.\n\nHope it helps!",
    "2465760": "thanks a lot"
  },
  "source": "meta"
}