{
  "id": 666770,
  "title": "Getting errors at submission ",
  "url": "/competitions/csiro-biomass/discussion/666770",
  "author_name": "",
  "post_date": "2026-01-09T02:04:32.112707700Z",
  "votes": null,
  "comment_count": 6,
  "views": 0,
  "content": "<p>Hello Everyone,</p>\n<p>Need your expertise and guidance in pointing me out what could be the problem that I’m facing while submitting the files for this competition.</p>\n<p>Here is my notebook for your reference- <a href=\"https://www.kaggle.com/code/vikasj1in/csiro-image2biomass-prediction-vk\" target=\"_blank\">https://www.kaggle.com/code/vikasj1in/csiro-image2biomass-prediction-vk</a></p>\n<p>I didn’t have any issues in my earlier submission though. Issue here is - I’m not getting any errors when executing this on kaggle environments.</p>\n<p>Thanking in advance!</p>",
  "messages": [
    {
      "id": "3388501",
      "postDate": "01/09/2026 02:04:32",
      "content": "<p>Hello Everyone,</p>\n<p>Need your expertise and guidance in pointing me out what could be the problem that I’m facing while submitting the files for this competition.</p>\n<p>Here is my notebook for your reference- <a href=\"https://www.kaggle.com/code/vikasj1in/csiro-image2biomass-prediction-vk\" target=\"_blank\">https://www.kaggle.com/code/vikasj1in/csiro-image2biomass-prediction-vk</a></p>\n<p>I didn’t have any issues in my earlier submission though. Issue here is - I’m not getting any errors when executing this on kaggle environments.</p>\n<p>Thanking in advance!</p>",
      "rawMarkdown": "Hello Everyone,\n\nNeed your expertise and guidance in pointing me out what could be the problem that I’m facing while submitting the files for this competition.\n\nHere is my notebook for your reference- https://www.kaggle.com/code/vikasj1in/csiro-image2biomass-prediction-vk\n\nI didn’t have any issues in my earlier submission though. Issue here is - I’m not getting any errors when executing this on kaggle environments.\n\nThanking in advance!",
      "votes": null
    },
    {
      "id": "3388513",
      "postDate": "01/09/2026 02:56:36",
      "content": "<p>I have probably have answered this at least a dozen posts and created one myself that talked about this error.  Sorry - but in a bad mood today!  </p>",
      "rawMarkdown": "I have probably have answered this at least a dozen posts and created one myself that talked about this error.  Sorry - but in a bad mood today!",
      "votes": null
    },
    {
      "id": "3388514",
      "postDate": "01/09/2026 02:58:16",
      "content": "<p>The sample_submission.csv file does not exist in the private test folder - this error kills you.  I think you tried to error trap - but what the heck - if you know it is not there (all those submission failure posts) don't still try to read it.  I have had mixed success in error trapping submissions.  It's always good practice but my suggestion - don't read a file a bunch of folks have told you does not exist.</p>\n<p><code>sample_submission = pd.read_csv(f'{data_path}/sample_submission.csv')</code></p>\n<p>Update:  I don't think your error trap did what it need to do.  It \"synthesized\" a new DataFrame but you really need a dataframe with content.  Maybe you read the images from the folder ?</p>",
      "rawMarkdown": "The sample_submission.csv file does not exist in the private test folder - this error kills you.  I think you tried to error trap - but what the heck - if you know it is not there (all those submission failure posts) don't still try to read it.  I have had mixed success in error trapping submissions.  It's always good practice but my suggestion - don't read a file a bunch of folks have told you does not exist.\n\n`sample_submission = pd.read_csv(f'{data_path}/sample_submission.csv')`\n\nUpdate:  I don't think your error trap did what it need to do.  It \"synthesized\" a new DataFrame but you really need a dataframe with content.  Maybe you read the images from the folder ?",
      "votes": null
    },
    {
      "id": "3388559",
      "postDate": "01/09/2026 05:25:25",
      "content": "<p>That fix not enough - still get a scoring error.  I would normally than suggest making the predictions against the train data rather than the test.  It's late - finding the code a little hard to follow.  I like to find the mistakes with this method on my local machine, but your pipeline would be a chore to recreate locally.  You can often find issues predicting train on local machines very quickly, while the turn-around time on kaggle is much slower.</p>\n<p>So I changed 3 lines to make that happen</p>\n<pre><code>train_df = pd.read_csv(f'{data_path}/train.csv')\ntest_df = pd.read_csv(f'{data_path}/train.csv')\nsample_submission = pd.read_csv(f'{data_path}/train.csv')\n</code></pre>\n<p>Bedtime here - so later to see what errors pop out.</p>",
      "rawMarkdown": "That fix not enough - still get a scoring error.  I would normally than suggest making the predictions against the train data rather than the test.  It's late - finding the code a little hard to follow.  I like to find the mistakes with this method on my local machine, but your pipeline would be a chore to recreate locally.  You can often find issues predicting train on local machines very quickly, while the turn-around time on kaggle is much slower.\n\nSo I changed 3 lines to make that happen\n```python\n\ntrain_df = pd.read_csv(f'{data_path}/train.csv')\ntest_df = pd.read_csv(f'{data_path}/train.csv')\nsample_submission = pd.read_csv(f'{data_path}/train.csv')\n```\n\nBedtime here - so later to see what errors pop out.",
      "votes": null
    },
    {
      "id": "3388806",
      "postDate": "01/09/2026 16:10:31",
      "content": "<p>Hmm - this is the first time running train did not produce an error in this competition.  Need to look deeper.</p>",
      "rawMarkdown": "Hmm - this is the first time running train did not produce an error in this competition.  Need to look deeper.",
      "votes": null
    },
    {
      "id": "3388811",
      "postDate": "01/09/2026 16:23:39",
      "content": "<p>Went back to look at your notebook - it appears you find the source of the error.  </p>\n<p>Always nice to come back to your post and let folks know it's solved.  </p>",
      "rawMarkdown": "Went back to look at your notebook - it appears you find the source of the error.  \n\nAlways nice to come back to your post and let folks know it's solved.",
      "votes": null
    },
    {
      "id": "3389002",
      "postDate": "01/10/2026 01:55:30",
      "content": "<p>Thanks for your input. it worked fine and i could successfully submit the work. Much appreciated!</p>",
      "rawMarkdown": "Thanks for your input. it worked fine and i could successfully submit the work. Much appreciated!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3388513,
      "author_name": "pcjimmmy",
      "author_url": "",
      "post_date": "01/09/2026 02:56:36",
      "content": "<p>I have probably have answered this at least a dozen posts and created one myself that talked about this error.  Sorry - but in a bad mood today!  </p>",
      "votes": null,
      "replies": [
        {
          "id": 3388514,
          "author_name": "pcjimmmy",
          "author_url": "",
          "post_date": "01/09/2026 02:58:16",
          "content": "<p>The sample_submission.csv file does not exist in the private test folder - this error kills you.  I think you tried to error trap - but what the heck - if you know it is not there (all those submission failure posts) don't still try to read it.  I have had mixed success in error trapping submissions.  It's always good practice but my suggestion - don't read a file a bunch of folks have told you does not exist.</p>\n<p><code>sample_submission = pd.read_csv(f'{data_path}/sample_submission.csv')</code></p>\n<p>Update:  I don't think your error trap did what it need to do.  It \"synthesized\" a new DataFrame but you really need a dataframe with content.  Maybe you read the images from the folder ?</p>",
          "votes": null,
          "replies": [
            {
              "id": 3388559,
              "author_name": "pcjimmmy",
              "author_url": "",
              "post_date": "01/09/2026 05:25:25",
              "content": "<p>That fix not enough - still get a scoring error.  I would normally than suggest making the predictions against the train data rather than the test.  It's late - finding the code a little hard to follow.  I like to find the mistakes with this method on my local machine, but your pipeline would be a chore to recreate locally.  You can often find issues predicting train on local machines very quickly, while the turn-around time on kaggle is much slower.</p>\n<p>So I changed 3 lines to make that happen</p>\n<pre><code>train_df = pd.read_csv(f'{data_path}/train.csv')\ntest_df = pd.read_csv(f'{data_path}/train.csv')\nsample_submission = pd.read_csv(f'{data_path}/train.csv')\n</code></pre>\n<p>Bedtime here - so later to see what errors pop out.</p>",
              "votes": null,
              "replies": [
                {
                  "id": 3388806,
                  "author_name": "pcjimmmy",
                  "author_url": "",
                  "post_date": "01/09/2026 16:10:31",
                  "content": "<p>Hmm - this is the first time running train did not produce an error in this competition.  Need to look deeper.</p>",
                  "votes": null,
                  "replies": []
                }
              ]
            }
          ]
        }
      ]
    },
    {
      "id": 3388811,
      "author_name": "pcjimmmy",
      "author_url": "",
      "post_date": "01/09/2026 16:23:39",
      "content": "<p>Went back to look at your notebook - it appears you find the source of the error.  </p>\n<p>Always nice to come back to your post and let folks know it's solved.  </p>",
      "votes": null,
      "replies": [
        {
          "id": 3389002,
          "author_name": "vikasj1in",
          "author_url": "",
          "post_date": "01/10/2026 01:55:30",
          "content": "<p>Thanks for your input. it worked fine and i could successfully submit the work. Much appreciated!</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3388501": "Hello Everyone,\n\nNeed your expertise and guidance in pointing me out what could be the problem that I’m facing while submitting the files for this competition.\n\nHere is my notebook for your reference- https://www.kaggle.com/code/vikasj1in/csiro-image2biomass-prediction-vk\n\nI didn’t have any issues in my earlier submission though. Issue here is - I’m not getting any errors when executing this on kaggle environments.\n\nThanking in advance!",
    "3388513": "I have probably have answered this at least a dozen posts and created one myself that talked about this error.  Sorry - but in a bad mood today!",
    "3388514": "The sample_submission.csv file does not exist in the private test folder - this error kills you.  I think you tried to error trap - but what the heck - if you know it is not there (all those submission failure posts) don't still try to read it.  I have had mixed success in error trapping submissions.  It's always good practice but my suggestion - don't read a file a bunch of folks have told you does not exist.\n\n`sample_submission = pd.read_csv(f'{data_path}/sample_submission.csv')`\n\nUpdate:  I don't think your error trap did what it need to do.  It \"synthesized\" a new DataFrame but you really need a dataframe with content.  Maybe you read the images from the folder ?",
    "3388559": "That fix not enough - still get a scoring error.  I would normally than suggest making the predictions against the train data rather than the test.  It's late - finding the code a little hard to follow.  I like to find the mistakes with this method on my local machine, but your pipeline would be a chore to recreate locally.  You can often find issues predicting train on local machines very quickly, while the turn-around time on kaggle is much slower.\n\nSo I changed 3 lines to make that happen\n```python\n\ntrain_df = pd.read_csv(f'{data_path}/train.csv')\ntest_df = pd.read_csv(f'{data_path}/train.csv')\nsample_submission = pd.read_csv(f'{data_path}/train.csv')\n```\n\nBedtime here - so later to see what errors pop out.",
    "3388806": "Hmm - this is the first time running train did not produce an error in this competition.  Need to look deeper.",
    "3388811": "Went back to look at your notebook - it appears you find the source of the error.  \n\nAlways nice to come back to your post and let folks know it's solved.",
    "3389002": "Thanks for your input. it worked fine and i could successfully submit the work. Much appreciated!"
  },
  "source": "meta"
}