{ "id": 271906, "title": "[HELP] Notebook Threw Exception", "url": "/competitions/rsna-miccai-brain-tumor-radiogenomic-classification/discussion/271906", "author_name": "justhungryman", "post_date": "2021-09-13T05:51:42.336000", "votes": 2, "comment_count": 6, "views": 0, "content": "

There's no problem on kaggle notebook, but when I submit it, an error occurs.
\nIt doesn't tell me what the error is.
\nHow do I solve this?
\nMy code address is below.
\nhttps://www.kaggle.com/justhungryman/jun-voxel-fold

", "messages": [ { "id": 1511148, "postDate": "2021-09-13T06:56:00.583Z", "content": "
    \n
  1. You are not reading the submission file and performing predictions on that.
    \nBecause, when your notebook was submitted, it reruns it replacing the submission file.
    2. \n
  2. second reason might be the datasets you are using were private.
    3. \n
", "rawMarkdown": "1. You are not reading the submission file and performing predictions on that.\nBecause, when your notebook was submitted, it reruns it replacing the submission file.\n2. second reason might be the datasets you are using were private.", "votes": 1, "replies": [ { "id": 1511280, "postDate": "2021-09-13T09:31:31.010Z", "content": "

@tharun2001

\n
    \n
  1. After reading the file, sample_submission.csv, I modified the value there and submitted it, but the same problem occurs.
    2. \n
  2. I only used public dataset 😂
    3. \n
", "rawMarkdown": "@tharun2001\n\n1. After reading the file, sample_submission.csv, I modified the value there and submitted it, but the same problem occurs.\n2. I only used public dataset 😂" }, { "id": 1511418, "postDate": "2021-09-13T12:03:09.563Z", "content": "

I used a preprocessed test dataset to make predictions(Not the original one), this threw a notebook exception for me. I solved it by using the original test data.

", "rawMarkdown": "I used a preprocessed test dataset to make predictions(Not the original one), this threw a notebook exception for me. I solved it by using the original test data.", "votes": 1 } ] }, { "id": 1511101, "postDate": "2021-09-13T05:51:42.337Z", "content": "

There's no problem on kaggle notebook, but when I submit it, an error occurs.
\nIt doesn't tell me what the error is.
\nHow do I solve this?
\nMy code address is below.
\nhttps://www.kaggle.com/justhungryman/jun-voxel-fold

", "rawMarkdown": "There's no problem on kaggle notebook, but when I submit it, an error occurs.\nIt doesn't tell me what the error is.\nHow do I solve this?\nMy code address is below.\nhttps://www.kaggle.com/justhungryman/jun-voxel-fold", "votes": 2 }, { "id": 1512618, "postDate": "2021-09-14T12:22:21.757Z", "content": "

The problem was in the preprocessing code, which is already reported.
\nhttps://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop#1507829

\n

So I changed preprocessing method. Good Luck!

", "rawMarkdown": "The problem was in the preprocessing code, which is already reported.\nhttps://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop#1507829\n\nSo I changed preprocessing method. Good Luck!" }, { "id": 1512434, "postDate": "2021-09-14T08:51:55.960Z", "content": "

You should use the test data with the actual competition dataset (i.e convert it into dicom from png) notebook threw exception for me too but luckily it got solved for me. So, yeah, define a function to convert dcm to png for every test case. Hope this helps :D

", "rawMarkdown": "You should use the test data with the actual competition dataset (i.e convert it into dicom from png) notebook threw exception for me too but luckily it got solved for me. So, yeah, define a function to convert dcm to png for every test case. Hope this helps :D" }, { "id": 1511279, "postDate": "2021-09-13T09:31:19Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 1511148, "author_name": "tharun_01", "author_url": "", "post_date": "2021-09-13T06:56:00.583000", "content": "
    \n
  1. You are not reading the submission file and performing predictions on that.
    \nBecause, when your notebook was submitted, it reruns it replacing the submission file.
    2. \n
  2. second reason might be the datasets you are using were private.
    3. \n
", "votes": 1, "replies": [ { "id": 1511280, "author_name": "justhungryman", "author_url": "", "post_date": "2021-09-13T09:31:31.010000", "content": "

@tharun2001

\n
    \n
  1. After reading the file, sample_submission.csv, I modified the value there and submitted it, but the same problem occurs.
    2. \n
  2. I only used public dataset 😂
    3. \n
", "votes": 0, "replies": [] }, { "id": 1511418, "author_name": "tharun_01", "author_url": "", "post_date": "2021-09-13T12:03:09.563000", "content": "

I used a preprocessed test dataset to make predictions(Not the original one), this threw a notebook exception for me. I solved it by using the original test data.

", "votes": 1, "replies": [] } ] }, { "id": 1512618, "author_name": "justhungryman", "author_url": "", "post_date": "2021-09-14T12:22:21.757000", "content": "

The problem was in the preprocessing code, which is already reported.
\nhttps://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop#1507829

\n

So I changed preprocessing method. Good Luck!

", "votes": 0, "replies": [] }, { "id": 1512434, "author_name": "Eddwait", "author_url": "", "post_date": "2021-09-14T08:51:55.960000", "content": "

You should use the test data with the actual competition dataset (i.e convert it into dicom from png) notebook threw exception for me too but luckily it got solved for me. So, yeah, define a function to convert dcm to png for every test case. Hope this helps :D

", "votes": 0, "replies": [] }, { "id": 1511279, "author_name": "", "author_url": "", "post_date": "2021-09-13T09:31:19", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "1511148": "1. You are not reading the submission file and performing predictions on that.\nBecause, when your notebook was submitted, it reruns it replacing the submission file.\n2. second reason might be the datasets you are using were private.", "1511101": "There's no problem on kaggle notebook, but when I submit it, an error occurs.\nIt doesn't tell me what the error is.\nHow do I solve this?\nMy code address is below.\nhttps://www.kaggle.com/justhungryman/jun-voxel-fold", "1512618": "The problem was in the preprocessing code, which is already reported.\nhttps://www.kaggle.com/ren4yu/normalized-voxels-align-planes-and-crop#1507829\n\nSo I changed preprocessing method. Good Luck!", "1512434": "You should use the test data with the actual competition dataset (i.e convert it into dicom from png) notebook threw exception for me too but luckily it got solved for me. So, yeah, define a function to convert dcm to png for every test case. Hope this helps :D", "1511279": "" } }