![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5387387%2F8dcffba636054032fb16e91eaf0401af%2FRSNA%202023-01-26.png?generation=1674749766949063&alt=media\")
Can anybody shed any light on why my submission is failing when the notebook I'm trying to submit is executing correctly?
\n
Contents of kaggle/working are:
\n![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F5387387%2F5e3578f4f52ee37f9761e833bb05a944%2Fkaggle%202023-01-26.png?generation=1674749822021649&alt=media\")
My submission.csv file looks like this
\nprediction_id cancer
\n10008_L 0.030661125
\n10008_R 0.011916713
The notebook URL is https://www.kaggle.com/code/julianmacnamara/rsna-screening-mammography-breast-cancer-detection
\nMany thanks in advance for any help anybody can offer
", "messages": [ { "id": 2118749, "postDate": "2023-01-28T08:40:03.213Z", "content": "I revisited the code and realised I'd inadvertently made 'prediction_id' an index and I should have included \"as_index=False\" in my .groupby statement as shown below
\n`# Select only the 'prediction_id' and 'cancer' columns
\nresulting_df = merged_df[['prediction_id', 'cancer']]
resulting_df = resulting_df.groupby('prediction_id', as_index=False).mean()
\nresulting_df = resulting_df.sort_index()`
\nHowever, when I ran this I still got a \"Submission score error\"
\nI was able to resolve this by following Abdoue's suggestion below
", "rawMarkdown": "I revisited the code and realised I'd inadvertently made 'prediction_id' an index and I should have included \"as_index=False\" in my .groupby statement as shown below\n\n`# Select only the 'prediction_id' and 'cancer' columns\nresulting_df = merged_df[['prediction_id', 'cancer']]\n\nresulting_df = resulting_df.groupby('prediction_id', as_index=False).mean()\n\nresulting_df = resulting_df.sort_index()`\n\nHowever, when I ran this I still got a \"Submission score error\"\n\nI was able to resolve this by following Abdoue's suggestion below", "votes": 1 }, { "id": 2116664, "postDate": "2023-01-26T16:27:44.217Z", "content": "Can anybody shed any light on why my submission is failing when the notebook I'm trying to submit is executing correctly?
\n
Contents of kaggle/working are:
\n
My submission.csv file looks like this
\nprediction_id cancer
\n10008_L 0.030661125
\n10008_R 0.011916713
The notebook URL is https://www.kaggle.com/code/julianmacnamara/rsna-screening-mammography-breast-cancer-detection
\nMany thanks in advance for any help anybody can offer
", "rawMarkdown": "Can anybody shed any light on why my submission is failing when the notebook I'm trying to submit is executing correctly?\n\n\n\nContents of kaggle/working are:\n\n\n\nMy submission.csv file looks like this\n\nprediction_id\tcancer\n10008_L\t0.030661125\n10008_R\t0.011916713\n\nThe notebook URL is https://www.kaggle.com/code/julianmacnamara/rsna-screening-mammography-breast-cancer-detection\n\nMany thanks in advance for any help anybody can offer\n\n", "votes": 1 }, { "id": 2117480, "postDate": "2023-01-27T10:01:45.227Z", "content": "Many thanks to everbody for their kind words and suggestions. I've tried them all as well as checking this but all to no avail
\n\"Insanity is doing the same thing over and over again and expecting different results.\" may be mis-attributed to Einstein but I still find it true in this case. However I've run out of things to try so am going to retire gracefully which is a shame. I was never going to win but, as a novice, I'd like to have seen how my ideas stacked up against other competitors
", "rawMarkdown": "Many thanks to everbody for their kind words and suggestions. I've tried them all as well as checking [this](https://www.kaggle.com/code-competition-debugging) but all to no avail\n\n \"Insanity is doing the same thing over and over again and expecting different results.\" may be mis-attributed to Einstein but I still find it true in this case. However I've run out of things to try so am going to retire gracefully which is a shame. I was never going to win but, as a novice, I'd like to have seen how my ideas stacked up against other competitors", "votes": 2 }, { "id": 2117960, "postDate": "2023-01-27T17:02:33Z", "content": "What was the error message?
\nIf it was \"Submission score error\" I had the same problem. Solved it by mapping my predictions onto a dataframe created directly from the provided test_csv. This ensures that all values of 'prediction_id' are there, if for whatever reason a prediction doesn't make it through the pipeline.
\nsub_df = pd.concat([prediction_df, full_df], join=) \n\nsub_df = sub_df.sort_values(by=[]).reset_index(drop=) \n\nsub_df = sub_df.fillna() \nMany thanks Abdou
\nI don't understand why this should work but it does
Do you have other files except submission.csv in the output folder? If so, try to make sure only submission.csv in that folder.
", "rawMarkdown": "Do you have other files except submission.csv in the output folder? If so, try to make sure only submission.csv in that folder." }, { "id": 2117096, "postDate": "2023-01-27T01:51:37.130Z", "content": "Check on submissions (at the right of \"Leaderboard\"/Rules/Team), there is a little more information about why it has failed.
\nUsually messages:
\n-Submission score error
\n-Notebook timeout
\n-Notebook run out of memory
\n-Notebook Threw exception
Hi Alfredo
\nThanks for making me improve my understanding on indexing
\nAll the best
\nJulian
", "rawMarkdown": "Hi Alfredo\n\nThanks for making me improve my understanding on indexing\n\nAll the best\n\nJulian" } ] }, { "id": 2116943, "postDate": "2023-01-26T20:37:21.427Z", "content": "I hope anyone could answer that and help you with the submission.
", "rawMarkdown": "I hope anyone could answer that and help you with the submission.", "replies": [ { "id": 2116952, "postDate": "2023-01-26T20:46:34.087Z", "content": "Thank you. Much appreciated
\nAll the best
\nJulian
", "rawMarkdown": "Thank you. Much appreciated \n\nAll the best \n\nJulian ", "votes": 1 } ] } ], "comments": [ { "id": 2118749, "author_name": "Julian Macnamara", "author_url": "", "post_date": "2023-01-28T08:40:03.213000", "content": "I revisited the code and realised I'd inadvertently made 'prediction_id' an index and I should have included \"as_index=False\" in my .groupby statement as shown below
\n`# Select only the 'prediction_id' and 'cancer' columns
\nresulting_df = merged_df[['prediction_id', 'cancer']]
resulting_df = resulting_df.groupby('prediction_id', as_index=False).mean()
\nresulting_df = resulting_df.sort_index()`
\nHowever, when I ran this I still got a \"Submission score error\"
\nI was able to resolve this by following Abdoue's suggestion below
", "votes": 1, "replies": [] }, { "id": 2117480, "author_name": "Julian Macnamara", "author_url": "", "post_date": "2023-01-27T10:01:45.227000", "content": "Many thanks to everbody for their kind words and suggestions. I've tried them all as well as checking this but all to no avail
\n\"Insanity is doing the same thing over and over again and expecting different results.\" may be mis-attributed to Einstein but I still find it true in this case. However I've run out of things to try so am going to retire gracefully which is a shame. I was never going to win but, as a novice, I'd like to have seen how my ideas stacked up against other competitors
", "votes": 2, "replies": [] }, { "id": 2117960, "author_name": "AbdouE", "author_url": "", "post_date": "2023-01-27T17:02:33", "content": "What was the error message?
\nIf it was \"Submission score error\" I had the same problem. Solved it by mapping my predictions onto a dataframe created directly from the provided test_csv. This ensures that all values of 'prediction_id' are there, if for whatever reason a prediction doesn't make it through the pipeline.
\nsub_df = pd.concat([prediction_df, full_df], join=) \n\nsub_df = sub_df.sort_values(by=[]).reset_index(drop=) \n\nsub_df = sub_df.fillna() \nMany thanks Abdou
\nI don't understand why this should work but it does
Do you have other files except submission.csv in the output folder? If so, try to make sure only submission.csv in that folder.
", "votes": 0, "replies": [] }, { "id": 2117096, "author_name": "Alfredo Maussa", "author_url": "", "post_date": "2023-01-27T01:51:37.130000", "content": "Check on submissions (at the right of \"Leaderboard\"/Rules/Team), there is a little more information about why it has failed.
\nUsually messages:
\n-Submission score error
\n-Notebook timeout
\n-Notebook run out of memory
\n-Notebook Threw exception
Hi Alfredo
\nThanks for making me improve my understanding on indexing
\nAll the best
\nJulian
", "votes": 0, "replies": [] } ] }, { "id": 2116943, "author_name": "Marília Prata", "author_url": "", "post_date": "2023-01-26T20:37:21.427000", "content": "I hope anyone could answer that and help you with the submission.
", "votes": 0, "replies": [ { "id": 2116952, "author_name": "Julian Macnamara", "author_url": "", "post_date": "2023-01-26T20:46:34.087000", "content": "Thank you. Much appreciated
\nAll the best
\nJulian
", "votes": 1, "replies": [] } ] } ], "raw_markdown_by_id": { "2118749": "I revisited the code and realised I'd inadvertently made 'prediction_id' an index and I should have included \"as_index=False\" in my .groupby statement as shown below\n\n`# Select only the 'prediction_id' and 'cancer' columns\nresulting_df = merged_df[['prediction_id', 'cancer']]\n\nresulting_df = resulting_df.groupby('prediction_id', as_index=False).mean()\n\nresulting_df = resulting_df.sort_index()`\n\nHowever, when I ran this I still got a \"Submission score error\"\n\nI was able to resolve this by following Abdoue's suggestion below", "2116664": "Can anybody shed any light on why my submission is failing when the notebook I'm trying to submit is executing correctly?\n\n\n\nContents of kaggle/working are:\n\n\n\nMy submission.csv file looks like this\n\nprediction_id\tcancer\n10008_L\t0.030661125\n10008_R\t0.011916713\n\nThe notebook URL is https://www.kaggle.com/code/julianmacnamara/rsna-screening-mammography-breast-cancer-detection\n\nMany thanks in advance for any help anybody can offer\n\n", "2117480": "Many thanks to everbody for their kind words and suggestions. I've tried them all as well as checking [this](https://www.kaggle.com/code-competition-debugging) but all to no avail\n\n \"Insanity is doing the same thing over and over again and expecting different results.\" may be mis-attributed to Einstein but I still find it true in this case. However I've run out of things to try so am going to retire gracefully which is a shame. I was never going to win but, as a novice, I'd like to have seen how my ideas stacked up against other competitors", "2117960": "What was the error message? \n\nIf it was \"Submission score error\" I had the same problem. Solved it by mapping my predictions onto a dataframe created directly from the provided test_csv. This ensures that all values of 'prediction_id' are there, if for whatever reason a prediction doesn't make it through the pipeline. \n\n```python\nsub_df = pd.concat([prediction_df, full_df], join='outer') # This eliminates any missing prediction ids\n\nsub_df = sub_df.sort_values(by=['prediction_id']).reset_index(drop=True) # Reordering by prediction id as required\n\nsub_df = sub_df.fillna(0) # Fill missing prediction values\n```", "2117104": "Do you have other files except submission.csv in the output folder? If so, try to make sure only submission.csv in that folder.", "2117096": "Check on submissions (at the right of \"Leaderboard\"/Rules/Team), there is a little more information about why it has failed.\n\nUsually messages:\n-Submission score error\n-Notebook timeout\n-Notebook run out of memory\n-Notebook Threw exception\n\n* Not sure, but I wouldn't use \"patient_id\" as an index, considered resetting the index before generating the CSV file.", "2116943": "I hope anyone could answer that and help you with the submission." } }