{ "id": 398720, "title": "Yet another thread for 'Submission Scoring Error' [Solved]", "url": "/competitions/tlvmc-parkinsons-freezing-gait-prediction/discussion/398720", "author_name": "Sireesh Limbu", "post_date": "2023-03-31T12:22:50.234000", "votes": 4, "comment_count": 6, "views": 0, "content": "

Hello fellow Kagglers!

\n

Sorry for another post about submission error and such. I have been stuck for a while trying to figure out why my notebook gives this error (Notebook Threw Exception). After going through a few posts on similar error, I made all the changes but I still couldn't point the error here. My output (head) looks like this:
\n. Id StartHesitation Turn Walking
\n0 003f117e14_0 0.0 0.002 0.0
\n1 003f117e14_1 0.0 0.002 0.0
\n2 003f117e14_2 0.0 0.002 0.0
\n3 003f117e14_3 0.0 0.002 0.0
\n4 003f117e14_4 0.0 0.002 0.0

\n

and tail like this

\n

. Id StartHesitation Turn Walking
\n286365 02ab235146_281683 0.0 0.0 0.0
\n286366 02ab235146_281684 0.0 0.0 0.0
\n286367 02ab235146_281685 0.0 0.0 0.0
\n286368 02ab235146_281686 0.0 0.0 0.0
\n286369 02ab235146_281687 0.0 0.0 0.0

\n

Is it because the column types are float and not int because I thought that was no issue. Also the the decimal point limit is just 3 which worked in all cases in the posts I went through. None of the values exceed 0.9.

\n

Any sort of idea/tip would be greatly appreciated!

\n

TIA

\n

Update 1 : I made the test source more generic after @avivlevi815 pointed it out so that even if ID changes, they can be processed. Now I am getting 'Submission Scoring Error'. I have checked my number of rows and columns are correct. And I rounded my probabilities to 3 after going through the post on 'Submission Scoring Error' although the metric has been updated to cater for this issue. The values are between 0 and 1. No row has more than one column with non zero value.

\n

Update 2 [Solved]:

\n

So few things that I did to rectify the error :

\n
    \n

  1. The update 1 mentioned above to use a generic code that uses file with any name.csv inside the test folder that Kaggle tests from their end

    2. \n

  2. For some reason I had not made use of sample_submission.csv file to create the submission.csv. I just looked at the structure and tried to imitate it. I copied the small following code from @ammarnassanalhajali to use submission_sample.csv while creating the submission file. As @avivlevi815 rightly mentioned that both sample_submission.csv and test files are change in Kaggles backend for scoring. Also, index = False.

    \n
    sub = pd.read_csv(p+'sample_submission.csv')\nsub['t'] = 0\nsubmission = pd.merge(sub[['Id','t']], final_pred, how='left', on='Id').fillna(0.0)\nsubmission[['Id','StartHesitation', 'Turn' , 'Walking']].to_csv('submission.csv', index=False)\n
    3. \n
\n

Hope this helps someone going through the same issue.
\nHappy Kaggling!

", "messages": [ { "id": 2204219, "postDate": "2023-03-31T12:22:50.233Z", "content": "

Hello fellow Kagglers!

\n

Sorry for another post about submission error and such. I have been stuck for a while trying to figure out why my notebook gives this error (Notebook Threw Exception). After going through a few posts on similar error, I made all the changes but I still couldn't point the error here. My output (head) looks like this:
\n. Id StartHesitation Turn Walking
\n0 003f117e14_0 0.0 0.002 0.0
\n1 003f117e14_1 0.0 0.002 0.0
\n2 003f117e14_2 0.0 0.002 0.0
\n3 003f117e14_3 0.0 0.002 0.0
\n4 003f117e14_4 0.0 0.002 0.0

\n

and tail like this

\n

. Id StartHesitation Turn Walking
\n286365 02ab235146_281683 0.0 0.0 0.0
\n286366 02ab235146_281684 0.0 0.0 0.0
\n286367 02ab235146_281685 0.0 0.0 0.0
\n286368 02ab235146_281686 0.0 0.0 0.0
\n286369 02ab235146_281687 0.0 0.0 0.0

\n

Is it because the column types are float and not int because I thought that was no issue. Also the the decimal point limit is just 3 which worked in all cases in the posts I went through. None of the values exceed 0.9.

\n

Any sort of idea/tip would be greatly appreciated!

\n

TIA

\n

Update 1 : I made the test source more generic after @avivlevi815 pointed it out so that even if ID changes, they can be processed. Now I am getting 'Submission Scoring Error'. I have checked my number of rows and columns are correct. And I rounded my probabilities to 3 after going through the post on 'Submission Scoring Error' although the metric has been updated to cater for this issue. The values are between 0 and 1. No row has more than one column with non zero value.

\n

Update 2 [Solved]:

\n

So few things that I did to rectify the error :

\n
    \n

  1. The update 1 mentioned above to use a generic code that uses file with any name.csv inside the test folder that Kaggle tests from their end

    2. \n

  2. For some reason I had not made use of sample_submission.csv file to create the submission.csv. I just looked at the structure and tried to imitate it. I copied the small following code from @ammarnassanalhajali to use submission_sample.csv while creating the submission file. As @avivlevi815 rightly mentioned that both sample_submission.csv and test files are change in Kaggles backend for scoring. Also, index = False.

    \n
    sub = pd.read_csv(p+'sample_submission.csv')\nsub['t'] = 0\nsubmission = pd.merge(sub[['Id','t']], final_pred, how='left', on='Id').fillna(0.0)\nsubmission[['Id','StartHesitation', 'Turn' , 'Walking']].to_csv('submission.csv', index=False)\n
    3. \n
\n

Hope this helps someone going through the same issue.
\nHappy Kaggling!

", "rawMarkdown": "Hello fellow Kagglers!\n\nSorry for another post about submission error and such. I have been stuck for a while trying to figure out why my notebook gives this error (Notebook Threw Exception). After going through a few posts on similar error, I made all the changes but I still couldn't point the error here. My output (head) looks like this:\n. Id\t StartHesitation Turn\t Walking\n0\t003f117e14_0 \t0.0\t 0.002\t 0.0\n1\t003f117e14_1\t 0.0\t 0.002\t 0.0\n2\t003f117e14_2\t 0.0\t 0.002\t 0.0\n3\t003f117e14_3\t 0.0\t 0.002\t 0.0\n4\t003f117e14_4 \t0.0\t 0.002\t 0.0\n\nand tail like this\n\n.\t Id\t StartHesitation Turn\t Walking\n286365\t02ab235146_281683\t0.0\t 0.0\t 0.0\n286366\t02ab235146_281684\t0.0\t 0.0\t 0.0\n286367\t02ab235146_281685\t0.0 0.0\t 0.0\n286368\t02ab235146_281686\t0.0\t 0.0\t 0.0\n286369\t02ab235146_281687\t0.0\t 0.0\t 0.0\n\nIs it because the column types are float and not int because I thought that was no issue. Also the the decimal point limit is just 3 which worked in all cases in the posts I went through. None of the values exceed 0.9.\n\nAny sort of idea/tip would be greatly appreciated!\n\nTIA\n\n\nUpdate 1 : I made the test source more generic after @avivlevi815 pointed it out so that even if ID changes, they can be processed. Now I am getting 'Submission Scoring Error'. I have checked my number of rows and columns are correct. And I rounded my probabilities to 3 after going through the post on 'Submission Scoring Error' although the metric has been updated to cater for this issue. The values are between 0 and 1. No row has more than one column with non zero value.\n\nUpdate 2 [Solved]:\n\nSo few things that I did to rectify the error : \n1. The update 1 mentioned above to use a generic code that uses file with any name.csv inside the test folder that Kaggle tests from their end\n2. For some reason I had not made use of sample_submission.csv file to create the submission.csv. I just looked at the structure and tried to imitate it. I copied the small following code from @ammarnassanalhajali to use submission_sample.csv while creating the submission file. As @avivlevi815 rightly mentioned that both sample_submission.csv and test files are change in Kaggles backend for scoring. Also, index = False.\n\n sub = pd.read_csv(p+'sample_submission.csv')\n sub['t'] = 0\n submission = pd.merge(sub[['Id','t']], final_pred, how='left', on='Id').fillna(0.0)\n submission[['Id','StartHesitation', 'Turn' , 'Walking']].to_csv('submission.csv', index=False)\n\nHope this helps someone going through the same issue.\nHappy Kaggling!", "votes": 4 }, { "id": 2204380, "postDate": "2023-03-31T14:15:07.833Z", "content": "

Did you take into consideration that the test & sample submission files are replaced when run at Kaggles end? (after you submit)

", "rawMarkdown": "Did you take into consideration that the test & sample submission files are replaced when run at Kaggles end? (after you submit)", "votes": 1, "replies": [ { "id": 2204480, "postDate": "2023-03-31T15:45:30.857Z", "content": "

Hi again. I did read this but as I think of it now, I did not pay much heed to it. I mean we only take into account the test.csv right? At Kaggle's end they just replace this test.csv with another test.csv with different values? At least that is what I could make of it.
\nOr does it also mean that there will no longer be files named '02ab235146.csv' and '003f117e14.csv'? And that I should code it in a generic way that caters for whatever filename.csv it has? Sorry if this is a basic question I am new to this comp.

", "rawMarkdown": "Hi again. I did read this but as I think of it now, I did not pay much heed to it. I mean we only take into account the test.csv right? At Kaggle's end they just replace this test.csv with another test.csv with different values? At least that is what I could make of it. \nOr does it also mean that there will no longer be files named '02ab235146.csv' and '003f117e14.csv'? And that I should code it in a generic way that caters for whatever filename.csv it has? Sorry if this is a basic question I am new to this comp. \n", "replies": [ { "id": 2204743, "postDate": "2023-03-31T22:00:22.220Z", "rawMarkdown": "", "votes": 1, "isDeleted": true }, { "id": 2204806, "postDate": "2023-04-01T01:17:54.013Z", "content": "

Yes thanks! I did this update!

", "rawMarkdown": "Yes thanks! I did this update! " } ] } ] }, { "id": 2204742, "postDate": "2023-03-31T21:56:37.667Z", "rawMarkdown": "", "isDeleted": true, "replies": [ { "id": 2204805, "postDate": "2023-04-01T01:17:23.590Z", "content": "

Hey! Thanks for the tip! After the generic code update for test.csv I am now getting a \"submission scoring error\" instead. Which means something in the output file is amiss! My output still looks the same as above. I have done all the basic checks. I'm really not sure what the issue is?

", "rawMarkdown": "Hey! Thanks for the tip! After the generic code update for test.csv I am now getting a \"submission scoring error\" instead. Which means something in the output file is amiss! My output still looks the same as above. I have done all the basic checks. I'm really not sure what the issue is?", "votes": 1 } ] } ], "comments": [ { "id": 2204380, "author_name": "Aviv Levi", "author_url": "", "post_date": "2023-03-31T14:15:07.833000", "content": "

Did you take into consideration that the test & sample submission files are replaced when run at Kaggles end? (after you submit)

", "votes": 1, "replies": [ { "id": 2204480, "author_name": "Sireesh Limbu", "author_url": "", "post_date": "2023-03-31T15:45:30.857000", "content": "

Hi again. I did read this but as I think of it now, I did not pay much heed to it. I mean we only take into account the test.csv right? At Kaggle's end they just replace this test.csv with another test.csv with different values? At least that is what I could make of it.
\nOr does it also mean that there will no longer be files named '02ab235146.csv' and '003f117e14.csv'? And that I should code it in a generic way that caters for whatever filename.csv it has? Sorry if this is a basic question I am new to this comp.

", "votes": 0, "replies": [ { "id": 2204743, "author_name": "", "author_url": "", "post_date": "2023-03-31T22:00:22.220000", "content": "", "votes": 1, "replies": [] }, { "id": 2204806, "author_name": "Sireesh Limbu", "author_url": "", "post_date": "2023-04-01T01:17:54.013000", "content": "

Yes thanks! I did this update!

", "votes": 0, "replies": [] } ] } ] }, { "id": 2204742, "author_name": "", "author_url": "", "post_date": "2023-03-31T21:56:37.667000", "content": "", "votes": 0, "replies": [ { "id": 2204805, "author_name": "Sireesh Limbu", "author_url": "", "post_date": "2023-04-01T01:17:23.590000", "content": "

Hey! Thanks for the tip! After the generic code update for test.csv I am now getting a \"submission scoring error\" instead. Which means something in the output file is amiss! My output still looks the same as above. I have done all the basic checks. I'm really not sure what the issue is?

", "votes": 1, "replies": [] } ] } ], "raw_markdown_by_id": { "2204219": "Hello fellow Kagglers!\n\nSorry for another post about submission error and such. I have been stuck for a while trying to figure out why my notebook gives this error (Notebook Threw Exception). After going through a few posts on similar error, I made all the changes but I still couldn't point the error here. My output (head) looks like this:\n. Id\t StartHesitation Turn\t Walking\n0\t003f117e14_0 \t0.0\t 0.002\t 0.0\n1\t003f117e14_1\t 0.0\t 0.002\t 0.0\n2\t003f117e14_2\t 0.0\t 0.002\t 0.0\n3\t003f117e14_3\t 0.0\t 0.002\t 0.0\n4\t003f117e14_4 \t0.0\t 0.002\t 0.0\n\nand tail like this\n\n.\t Id\t StartHesitation Turn\t Walking\n286365\t02ab235146_281683\t0.0\t 0.0\t 0.0\n286366\t02ab235146_281684\t0.0\t 0.0\t 0.0\n286367\t02ab235146_281685\t0.0 0.0\t 0.0\n286368\t02ab235146_281686\t0.0\t 0.0\t 0.0\n286369\t02ab235146_281687\t0.0\t 0.0\t 0.0\n\nIs it because the column types are float and not int because I thought that was no issue. Also the the decimal point limit is just 3 which worked in all cases in the posts I went through. None of the values exceed 0.9.\n\nAny sort of idea/tip would be greatly appreciated!\n\nTIA\n\n\nUpdate 1 : I made the test source more generic after @avivlevi815 pointed it out so that even if ID changes, they can be processed. Now I am getting 'Submission Scoring Error'. I have checked my number of rows and columns are correct. And I rounded my probabilities to 3 after going through the post on 'Submission Scoring Error' although the metric has been updated to cater for this issue. The values are between 0 and 1. No row has more than one column with non zero value.\n\nUpdate 2 [Solved]:\n\nSo few things that I did to rectify the error : \n1. The update 1 mentioned above to use a generic code that uses file with any name.csv inside the test folder that Kaggle tests from their end\n2. For some reason I had not made use of sample_submission.csv file to create the submission.csv. I just looked at the structure and tried to imitate it. I copied the small following code from @ammarnassanalhajali to use submission_sample.csv while creating the submission file. As @avivlevi815 rightly mentioned that both sample_submission.csv and test files are change in Kaggles backend for scoring. Also, index = False.\n\n sub = pd.read_csv(p+'sample_submission.csv')\n sub['t'] = 0\n submission = pd.merge(sub[['Id','t']], final_pred, how='left', on='Id').fillna(0.0)\n submission[['Id','StartHesitation', 'Turn' , 'Walking']].to_csv('submission.csv', index=False)\n\nHope this helps someone going through the same issue.\nHappy Kaggling!", "2204380": "Did you take into consideration that the test & sample submission files are replaced when run at Kaggles end? (after you submit)", "2204742": "" } }