{
  "id": 187420,
  "title": "Submission Error",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/187420",
  "author_name": "",
  "post_date": "2020-09-28T20:54:56.806540300Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Is it possible to train the model locally then attach the csv file in kernel for submission?<br>\nI tried this method but got submission error. Is it due to attaching the csv file externally or something wrong in submission file?</p>",
  "messages": [
    {
      "id": "1030688",
      "postDate": "09/28/2020 20:54:56",
      "content": "<p>Is it possible to train the model locally then attach the csv file in kernel for submission?<br>\nI tried this method but got submission error. Is it due to attaching the csv file externally or something wrong in submission file?</p>",
      "rawMarkdown": "Is it possible to train the model locally then attach the csv file in kernel for submission?\nI tried this method but got submission error. Is it due to attaching the csv file externally or something wrong in submission file?",
      "votes": null
    },
    {
      "id": "1030703",
      "postDate": "09/28/2020 21:40:58",
      "content": "<p>No.</p>\n<p>The submission file must be based on hidden test data. You must run your model in your committed notebook. The public test data is just a sample to show you the format (it matches the last 5 patients in the train dataset).</p>",
      "rawMarkdown": "No.\n\nThe submission file must be based on hidden test data. You must run your model in your committed notebook. The public test data is just a sample to show you the format (it matches the last 5 patients in the train dataset).",
      "votes": null
    },
    {
      "id": "1030938",
      "postDate": "09/29/2020 05:31:31",
      "content": "<p>But I submitted the sample_submission.csv file and It got accepted and showing -13.0145 score. The code is like that:</p>\n<p>import pandas as pd<br>\nsub = pd.read_csv(\"../input/osic-pulmonary-fibrosis-progression/sample_submission.csv\")<br>\nsub.to_csv(\"submission.csv\", index = False)</p>",
      "rawMarkdown": "But I submitted the sample_submission.csv file and It got accepted and showing -13.0145 score. The code is like that:\n\nimport pandas as pd\nsub = pd.read_csv(\"../input/osic-pulmonary-fibrosis-progression/sample_submission.csv\")\nsub.to_csv(\"submission.csv\", index = False)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1030703,
      "author_name": "richardepstein",
      "author_url": "",
      "post_date": "09/28/2020 21:40:58",
      "content": "<p>No.</p>\n<p>The submission file must be based on hidden test data. You must run your model in your committed notebook. The public test data is just a sample to show you the format (it matches the last 5 patients in the train dataset).</p>",
      "votes": null,
      "replies": [
        {
          "id": 1030938,
          "author_name": "dky7376",
          "author_url": "",
          "post_date": "09/29/2020 05:31:31",
          "content": "<p>But I submitted the sample_submission.csv file and It got accepted and showing -13.0145 score. The code is like that:</p>\n<p>import pandas as pd<br>\nsub = pd.read_csv(\"../input/osic-pulmonary-fibrosis-progression/sample_submission.csv\")<br>\nsub.to_csv(\"submission.csv\", index = False)</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1030688": "Is it possible to train the model locally then attach the csv file in kernel for submission?\nI tried this method but got submission error. Is it due to attaching the csv file externally or something wrong in submission file?",
    "1030703": "No.\n\nThe submission file must be based on hidden test data. You must run your model in your committed notebook. The public test data is just a sample to show you the format (it matches the last 5 patients in the train dataset).",
    "1030938": "But I submitted the sample_submission.csv file and It got accepted and showing -13.0145 score. The code is like that:\n\nimport pandas as pd\nsub = pd.read_csv(\"../input/osic-pulmonary-fibrosis-progression/sample_submission.csv\")\nsub.to_csv(\"submission.csv\", index = False)"
  },
  "source": "meta"
}