{
  "id": 175384,
  "title": "Submission CSV Not Found",
  "url": "/competitions/osic-pulmonary-fibrosis-progression/discussion/175384",
  "author_name": "",
  "post_date": "2020-08-18T04:09:52.947001400Z",
  "votes": null,
  "comment_count": 8,
  "views": 0,
  "content": "<p>Hello all,</p>\n<p>I have written up this basic model to start with and I was really happy to not have any errors when I went to save and commit it. However, I'm getting the dreaded <code>Submission CSV Not Found</code> error that some other users are getting. I have no idea why as it outputs a completely valid csv file and there are no errors in the log. <br>\nI've tried putting a try/except clause around the creation of it (no errors) and I tried turning it into a script to find errors that way (no errors). I'm not running any of the images so I highly doubt that I'm running out of memory or CPU compute.</p>\n<p>I've attached my kernel below for you to look at. <br>\n<a href=\"https://www.kaggle.com/dannyk2018/simple-linear-model/\" target=\"_blank\">https://www.kaggle.com/dannyk2018/simple-linear-model/</a></p>\n<p>Thanks for your help in advance,<br>\nDanny</p>",
  "messages": [
    {
      "id": "974816",
      "postDate": "08/18/2020 04:09:52",
      "content": "<p>Hello all,</p>\n<p>I have written up this basic model to start with and I was really happy to not have any errors when I went to save and commit it. However, I'm getting the dreaded <code>Submission CSV Not Found</code> error that some other users are getting. I have no idea why as it outputs a completely valid csv file and there are no errors in the log. <br>\nI've tried putting a try/except clause around the creation of it (no errors) and I tried turning it into a script to find errors that way (no errors). I'm not running any of the images so I highly doubt that I'm running out of memory or CPU compute.</p>\n<p>I've attached my kernel below for you to look at. <br>\n<a href=\"https://www.kaggle.com/dannyk2018/simple-linear-model/\" target=\"_blank\">https://www.kaggle.com/dannyk2018/simple-linear-model/</a></p>\n<p>Thanks for your help in advance,<br>\nDanny</p>",
      "rawMarkdown": "Hello all,\n\nI have written up this basic model to start with and I was really happy to not have any errors when I went to save and commit it. However, I'm getting the dreaded `Submission CSV Not Found` error that some other users are getting. I have no idea why as it outputs a completely valid csv file and there are no errors in the log. \nI've tried putting a try/except clause around the creation of it (no errors) and I tried turning it into a script to find errors that way (no errors). I'm not running any of the images so I highly doubt that I'm running out of memory or CPU compute.\n\nI've attached my kernel below for you to look at. \nhttps://www.kaggle.com/dannyk2018/simple-linear-model/\n\nThanks for your help in advance,\nDanny",
      "votes": null
    },
    {
      "id": "975118",
      "postDate": "08/18/2020 06:50:21",
      "content": "<p>I had a look at your notebook, there might two things which are making your submissions invalid:</p>\n<ul>\n<li><p><code>kaggle_submission.to_csv('./submission.csv', index=False, float_format='%.1f')</code> should be <code>kaggle_submission.to_csv('submission.csv', index=False, float_format='%.1f')</code> (Hopefully tinkering with the output path would help, worked in my case)</p></li>\n<li><p>Internet should be switched off. You can't submit if your internet is switched on.</p></li>\n</ul>",
      "rawMarkdown": "I had a look at your notebook, there might two things which are making your submissions invalid:\n\n- `kaggle_submission.to_csv('./submission.csv', index=False, float_format='%.1f')` should be `kaggle_submission.to_csv('submission.csv', index=False, float_format='%.1f')` (Hopefully tinkering with the output path would help, worked in my case)\n\n- Internet should be switched off. You can't submit if your internet is switched on.",
      "votes": null
    },
    {
      "id": "975657",
      "postDate": "08/18/2020 12:04:15",
      "content": "<p>I'm having something similar happen??? I 'commit' my Notebook, then view it, 'Submit' my output…then wait and it 'fails'. I've tried various versions of output paths that I've seen in other Notebooks, but no change?? It seems like a fairly round-a-bout way of submitting results, but it obviously works…for some??</p>\n<p>PS: Your FVC values calculated appear at least 10x what a large human would produce…might need to check something there??? I might be wrong too…???</p>",
      "rawMarkdown": "I'm having something similar happen??? I 'commit' my Notebook, then view it, 'Submit' my output...then wait and it 'fails'. I've tried various versions of output paths that I've seen in other Notebooks, but no change?? It seems like a fairly round-a-bout way of submitting results, but it obviously works...for some??\n\nPS: Your FVC values calculated appear at least 10x what a large human would produce...might need to check something there??? I might be wrong too...???",
      "votes": null
    },
    {
      "id": "975797",
      "postDate": "08/18/2020 13:22:08",
      "content": "<p>Thanks for the reply! I originally had the path without the <code>./</code> and that wasn't working either. Maybe I'll try it again just to see</p>",
      "rawMarkdown": "Thanks for the reply! I originally had the path without the `./` and that wasn't working either. Maybe I'll try it again just to see",
      "votes": null
    },
    {
      "id": "975800",
      "postDate": "08/18/2020 13:24:32",
      "content": "<p>Yeah this seems like a decently common occurrence which is super frustrating especially cause there's nothing in the output log to suggest what it might be. Also yeah, my FVC values are crazy off, this was just to get a basic model flow running and then tune from there. Might be worth creating some more realistic values just to see if that's throwing it off.</p>",
      "rawMarkdown": "Yeah this seems like a decently common occurrence which is super frustrating especially cause there's nothing in the output log to suggest what it might be. Also yeah, my FVC values are crazy off, this was just to get a basic model flow running and then tune from there. Might be worth creating some more realistic values just to see if that's throwing it off.",
      "votes": null
    },
    {
      "id": "976172",
      "postDate": "08/18/2020 17:26:33",
      "content": "<p>Danny,</p>\n<p>What are you doing in this line? Does it presume that the test patients are in the train dataset?</p>\n<p>test=test[train.columns.drop(['FVC','Percent'])]</p>\n<p>-Rich</p>",
      "rawMarkdown": "Danny,\n\nWhat are you doing in this line? Does it presume that the test patients are in the train dataset?\n\ntest=test[train.columns.drop(['FVC','Percent'])]\n\n-Rich",
      "votes": null
    },
    {
      "id": "976234",
      "postDate": "08/18/2020 18:27:47",
      "content": "<p>Try running (on your end, just for debugging) the part of your code that produces the CSV file on the training data rather than the test data. This might kick up errors you wouldn't see otherwise, as there is a greater diversity of potential issues with the (larger) training data set than with the 5-line \"test\" set. I had a similar problem a while back and this is how I managed to solve it.</p>",
      "rawMarkdown": "Try running (on your end, just for debugging) the part of your code that produces the CSV file on the training data rather than the test data. This might kick up errors you wouldn't see otherwise, as there is a greater diversity of potential issues with the (larger) training data set than with the 5-line \"test\" set. I had a similar problem a while back and this is how I managed to solve it.",
      "votes": null
    },
    {
      "id": "980901",
      "postDate": "08/22/2020 01:50:42",
      "content": "<p>I just wanted to update anyone who might happen upon this same problem. I ended up going back over my notebook with a comb and recreated it. While I wasn't able to pin point the exact source of my problem, I believe it was an issue with how I was getting the ID values. I wasn't pulling them from <code>test.csv</code> as I should've been and as a result I think it silently failed. I still think a better error message would be a scoring issue in this case (which I know exists) but I digress. My advice to people in the future is to very carefully look at how you are creating your submission file.</p>",
      "rawMarkdown": "I just wanted to update anyone who might happen upon this same problem. I ended up going back over my notebook with a comb and recreated it. While I wasn't able to pin point the exact source of my problem, I believe it was an issue with how I was getting the ID values. I wasn't pulling them from `test.csv` as I should've been and as a result I think it silently failed. I still think a better error message would be a scoring issue in this case (which I know exists) but I digress. My advice to people in the future is to very carefully look at how you are creating your submission file.",
      "votes": null
    },
    {
      "id": "981370",
      "postDate": "08/22/2020 11:51:28",
      "content": "<p>I managed to solve my problem…a typo by me…'Patient_weeks' instead of 'Patient_week' as my column heading……..</p>",
      "rawMarkdown": "I managed to solve my problem...a typo by me...'Patient_weeks' instead of 'Patient_week' as my column heading........",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 975118,
      "author_name": "aadhavvignesh",
      "author_url": "",
      "post_date": "08/18/2020 06:50:21",
      "content": "<p>I had a look at your notebook, there might two things which are making your submissions invalid:</p>\n<ul>\n<li><p><code>kaggle_submission.to_csv('./submission.csv', index=False, float_format='%.1f')</code> should be <code>kaggle_submission.to_csv('submission.csv', index=False, float_format='%.1f')</code> (Hopefully tinkering with the output path would help, worked in my case)</p></li>\n<li><p>Internet should be switched off. You can't submit if your internet is switched on.</p></li>\n</ul>",
      "votes": null,
      "replies": [
        {
          "id": 975797,
          "author_name": "dannyk2018",
          "author_url": "",
          "post_date": "08/18/2020 13:22:08",
          "content": "<p>Thanks for the reply! I originally had the path without the <code>./</code> and that wasn't working either. Maybe I'll try it again just to see</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 975657,
      "author_name": "geoffcc",
      "author_url": "",
      "post_date": "08/18/2020 12:04:15",
      "content": "<p>I'm having something similar happen??? I 'commit' my Notebook, then view it, 'Submit' my output…then wait and it 'fails'. I've tried various versions of output paths that I've seen in other Notebooks, but no change?? It seems like a fairly round-a-bout way of submitting results, but it obviously works…for some??</p>\n<p>PS: Your FVC values calculated appear at least 10x what a large human would produce…might need to check something there??? I might be wrong too…???</p>",
      "votes": null,
      "replies": [
        {
          "id": 975800,
          "author_name": "dannyk2018",
          "author_url": "",
          "post_date": "08/18/2020 13:24:32",
          "content": "<p>Yeah this seems like a decently common occurrence which is super frustrating especially cause there's nothing in the output log to suggest what it might be. Also yeah, my FVC values are crazy off, this was just to get a basic model flow running and then tune from there. Might be worth creating some more realistic values just to see if that's throwing it off.</p>",
          "votes": null,
          "replies": [
            {
              "id": 976172,
              "author_name": "richardepstein",
              "author_url": "",
              "post_date": "08/18/2020 17:26:33",
              "content": "<p>Danny,</p>\n<p>What are you doing in this line? Does it presume that the test patients are in the train dataset?</p>\n<p>test=test[train.columns.drop(['FVC','Percent'])]</p>\n<p>-Rich</p>",
              "votes": null,
              "replies": []
            }
          ]
        },
        {
          "id": 981370,
          "author_name": "geoffcc",
          "author_url": "",
          "post_date": "08/22/2020 11:51:28",
          "content": "<p>I managed to solve my problem…a typo by me…'Patient_weeks' instead of 'Patient_week' as my column heading……..</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 976234,
      "author_name": "archimedus",
      "author_url": "",
      "post_date": "08/18/2020 18:27:47",
      "content": "<p>Try running (on your end, just for debugging) the part of your code that produces the CSV file on the training data rather than the test data. This might kick up errors you wouldn't see otherwise, as there is a greater diversity of potential issues with the (larger) training data set than with the 5-line \"test\" set. I had a similar problem a while back and this is how I managed to solve it.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 980901,
      "author_name": "dannyk2018",
      "author_url": "",
      "post_date": "08/22/2020 01:50:42",
      "content": "<p>I just wanted to update anyone who might happen upon this same problem. I ended up going back over my notebook with a comb and recreated it. While I wasn't able to pin point the exact source of my problem, I believe it was an issue with how I was getting the ID values. I wasn't pulling them from <code>test.csv</code> as I should've been and as a result I think it silently failed. I still think a better error message would be a scoring issue in this case (which I know exists) but I digress. My advice to people in the future is to very carefully look at how you are creating your submission file.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "974816": "Hello all,\n\nI have written up this basic model to start with and I was really happy to not have any errors when I went to save and commit it. However, I'm getting the dreaded `Submission CSV Not Found` error that some other users are getting. I have no idea why as it outputs a completely valid csv file and there are no errors in the log. \nI've tried putting a try/except clause around the creation of it (no errors) and I tried turning it into a script to find errors that way (no errors). I'm not running any of the images so I highly doubt that I'm running out of memory or CPU compute.\n\nI've attached my kernel below for you to look at. \nhttps://www.kaggle.com/dannyk2018/simple-linear-model/\n\nThanks for your help in advance,\nDanny",
    "975118": "I had a look at your notebook, there might two things which are making your submissions invalid:\n\n- `kaggle_submission.to_csv('./submission.csv', index=False, float_format='%.1f')` should be `kaggle_submission.to_csv('submission.csv', index=False, float_format='%.1f')` (Hopefully tinkering with the output path would help, worked in my case)\n\n- Internet should be switched off. You can't submit if your internet is switched on.",
    "975657": "I'm having something similar happen??? I 'commit' my Notebook, then view it, 'Submit' my output...then wait and it 'fails'. I've tried various versions of output paths that I've seen in other Notebooks, but no change?? It seems like a fairly round-a-bout way of submitting results, but it obviously works...for some??\n\nPS: Your FVC values calculated appear at least 10x what a large human would produce...might need to check something there??? I might be wrong too...???",
    "975797": "Thanks for the reply! I originally had the path without the `./` and that wasn't working either. Maybe I'll try it again just to see",
    "975800": "Yeah this seems like a decently common occurrence which is super frustrating especially cause there's nothing in the output log to suggest what it might be. Also yeah, my FVC values are crazy off, this was just to get a basic model flow running and then tune from there. Might be worth creating some more realistic values just to see if that's throwing it off.",
    "976172": "Danny,\n\nWhat are you doing in this line? Does it presume that the test patients are in the train dataset?\n\ntest=test[train.columns.drop(['FVC','Percent'])]\n\n-Rich",
    "976234": "Try running (on your end, just for debugging) the part of your code that produces the CSV file on the training data rather than the test data. This might kick up errors you wouldn't see otherwise, as there is a greater diversity of potential issues with the (larger) training data set than with the 5-line \"test\" set. I had a similar problem a while back and this is how I managed to solve it.",
    "980901": "I just wanted to update anyone who might happen upon this same problem. I ended up going back over my notebook with a comb and recreated it. While I wasn't able to pin point the exact source of my problem, I believe it was an issue with how I was getting the ID values. I wasn't pulling them from `test.csv` as I should've been and as a result I think it silently failed. I still think a better error message would be a scoring issue in this case (which I know exists) but I digress. My advice to people in the future is to very carefully look at how you are creating your submission file.",
    "981370": "I managed to solve my problem...a typo by me...'Patient_weeks' instead of 'Patient_week' as my column heading........"
  },
  "source": "meta"
}