{
  "id": 215962,
  "title": "How to solve submission.csv not found error",
  "url": "/competitions/cassava-leaf-disease-classification/discussion/215962",
  "author_name": "Trushant Kalyanpur",
  "post_date": "2021-02-01T04:14:07.860000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I noticed numerous posts on people having a difficulty with scoring which can definitely lead to a lot of frustration. I am outlining the main things to try when you see this error and why you dont see this locally.</p>\n<ul>\n<li>The main thing to remember is that the Public LB is calculated on an unseen test dataset that is different from the submission.csv/test set that you try to score locally on.  At submission time, the submission.csv and test images are replaced with a much larger dataset with more than just one image. With that being said,  these are the main things to try.</li>\n<li>Try with more than just one image in your submission.csv file. This can be either a duplication of the same entry or it's better to use something like the train dataset for scoring/prediction. This will expose a lot of issues in how you are preparing the submission.csv file and help you root cause the issue.</li>\n<li>Make sure your filename is 'submission.csv' and you have the right column names and store the csv like this:</li>\n</ul>\n<pre><code> submission_df.to_csv('submission.csv', index=False)\n</code></pre>\n<p>Hope this helps you figure out what is wrong with your code and resolve the errors seen during submission. </p>",
  "messages": [
    {
      "id": 1180052,
      "postDate": "2021-02-01T04:14:07.860Z",
      "content": "<p>I noticed numerous posts on people having a difficulty with scoring which can definitely lead to a lot of frustration. I am outlining the main things to try when you see this error and why you dont see this locally.</p>\n<ul>\n<li>The main thing to remember is that the Public LB is calculated on an unseen test dataset that is different from the submission.csv/test set that you try to score locally on.  At submission time, the submission.csv and test images are replaced with a much larger dataset with more than just one image. With that being said,  these are the main things to try.</li>\n<li>Try with more than just one image in your submission.csv file. This can be either a duplication of the same entry or it's better to use something like the train dataset for scoring/prediction. This will expose a lot of issues in how you are preparing the submission.csv file and help you root cause the issue.</li>\n<li>Make sure your filename is 'submission.csv' and you have the right column names and store the csv like this:</li>\n</ul>\n<pre><code> submission_df.to_csv('submission.csv', index=False)\n</code></pre>\n<p>Hope this helps you figure out what is wrong with your code and resolve the errors seen during submission. </p>",
      "rawMarkdown": "I noticed numerous posts on people having a difficulty with scoring which can definitely lead to a lot of frustration. I am outlining the main things to try when you see this error and why you dont see this locally.\n\n- The main thing to remember is that the Public LB is calculated on an unseen test dataset that is different from the submission.csv/test set that you try to score locally on.  At submission time, the submission.csv and test images are replaced with a much larger dataset with more than just one image. With that being said,  these are the main things to try.\n-  Try with more than just one image in your submission.csv file. This can be either a duplication of the same entry or it's better to use something like the train dataset for scoring/prediction. This will expose a lot of issues in how you are preparing the submission.csv file and help you root cause the issue.\n- Make sure your filename is 'submission.csv' and you have the right column names and store the csv like this:\n```\n submission_df.to_csv('submission.csv', index=False)\n\n```\n\nHope this helps you figure out what is wrong with your code and resolve the errors seen during submission. ",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1180052": "I noticed numerous posts on people having a difficulty with scoring which can definitely lead to a lot of frustration. I am outlining the main things to try when you see this error and why you dont see this locally.\n\n- The main thing to remember is that the Public LB is calculated on an unseen test dataset that is different from the submission.csv/test set that you try to score locally on.  At submission time, the submission.csv and test images are replaced with a much larger dataset with more than just one image. With that being said,  these are the main things to try.\n-  Try with more than just one image in your submission.csv file. This can be either a duplication of the same entry or it's better to use something like the train dataset for scoring/prediction. This will expose a lot of issues in how you are preparing the submission.csv file and help you root cause the issue.\n- Make sure your filename is 'submission.csv' and you have the right column names and store the csv like this:\n```\n submission_df.to_csv('submission.csv', index=False)\n\n```\n\nHope this helps you figure out what is wrong with your code and resolve the errors seen during submission. "
  }
}