![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F6189766%2F1887dab6d6adf2935c07b9fac0590f33%2F.PNG?generation=1686631611118609&alt=media\")
Is it just me who's suffering from submission scoring error right now??
\nWill this be re-scored again when the error's fixed? Or do I have to submit it again?
\nI do not want to waste my submission try…
\nI'm currently using jo wilder 3.7
\nI found out the problem and I think it may help you.
\nWhile predicting and submitting, the data only comes with one session id. I was making some meta data with certain columns, if this column is under some condition. And I realized that in my train / test set, there's a bunch of other session id so that the feature columns will still remains even with nan values. But if featurizing with a single session id with the condition not applicable, those features disappeared.
\nConsider this case
\ncolumn = \n\ncond = df[column] != \ngb = df[cond].groupby([, ])\nfeature = gb[].mean()\nIf cond is all False then feature will not be correctly generated. This happens while running jo_wilder because there is only one session id each time.
\nI hope this can help.
", "messages": [ { "id": "2300049", "postDate": "06/13/2023 02:30:07", "content": "Is it just me who's suffering from submission scoring error right now??
\nWill this be re-scored again when the error's fixed? Or do I have to submit it again?
\nI do not want to waste my submission try…
\nI'm currently using jo wilder 3.7
\nI found out the problem and I think it may help you.
\nWhile predicting and submitting, the data only comes with one session id. I was making some meta data with certain columns, if this column is under some condition. And I realized that in my train / test set, there's a bunch of other session id so that the feature columns will still remains even with nan values. But if featurizing with a single session id with the condition not applicable, those features disappeared.
\nConsider this case
\ncolumn = \n\ncond = df[column] != \ngb = df[cond].groupby([, ])\nfeature = gb[].mean()\nIf cond is all False then feature will not be correctly generated. This happens while running jo_wilder because there is only one session id each time.
\nI hope this can help.
", "rawMarkdown": "Is it just me who's suffering from submission scoring error right now??\n\nWill this be re-scored again when the error's fixed? Or do I have to submit it again?\n\nI do not want to waste my submission try...\n\nI'm currently using jo wilder 3.7\n\n\n----\nI found out the problem and I think it may help you.\n\nWhile predicting and submitting, the data only comes with one session id. I was making some meta data with certain columns, if this column is under some condition. And I realized that in my train / test set, there's a bunch of other session id so that the feature columns will still remains even with nan values. But if featurizing with a single session id with the condition not applicable, those features disappeared.\n\nConsider this case\n\n\n```python\ncolumn = 'some_column'\n\ncond = df[column] != 0\ngb = df[cond].groupby(['session_id', 'level_group'])\nfeature = gb['elapsed_time'].mean()\n```\n\n\n\nIf cond is all False then feature will not be correctly generated. This happens while running jo_wilder because there is only one session id each time. \n\nI hope this can help.", "votes": null }, { "id": "2300055", "postDate": "06/13/2023 02:38:01", "content": "Scoring error same for me! My submission cannot get score since last night, just running and running over 15h by now!
", "rawMarkdown": "Scoring error same for me! My submission cannot get score since last night, just running and running over 15h by now!", "votes": null }, { "id": "2300062", "postDate": "06/13/2023 02:48:32", "content": "did you try to submit again with the ones you already succeeded?
", "rawMarkdown": "did you try to submit again with the ones you already succeeded?", "votes": null }, { "id": "2300066", "postDate": "06/13/2023 02:50:47", "content": "I'm experiencing the same issue. The scoring has been running for more than 12 hours.
", "rawMarkdown": "I'm experiencing the same issue. The scoring has been running for more than 12 hours.", "votes": null }, { "id": "2300069", "postDate": "06/13/2023 02:52:13", "content": "Really exhausted in this competition…
", "rawMarkdown": "Really exhausted in this competition...", "votes": null }, { "id": "2300082", "postDate": "06/13/2023 03:04:42", "content": "I’m with you…
", "rawMarkdown": "I’m with you...", "votes": null }, { "id": "2300188", "postDate": "06/13/2023 04:35:33", "content": "Did you check if rerun the ones you've succeeded not working as well??
", "rawMarkdown": "Did you check if rerun the ones you've succeeded not working as well??", "votes": null }, { "id": "2300209", "postDate": "06/13/2023 04:53:14", "content": "
I have.
\nIf you have submitted a successful submission, but after re-scored, it becomes a submission scoring error.
\nThat's most likely a bug in your original submit script that wasn't shown as a failure (at least I did)
\nAfter I found those bug and modified them, I can submit normally.
I'm not trying it.
\nIn the end, I received an unfamiliar message called \"Kaggle error.\"
Hey guys, I've found my error case. Check if the case is applicable to yours. Thanks.
", "rawMarkdown": "Hey guys, I've found my error case. Check if the case is applicable to yours. Thanks.", "votes": null }, { "id": "2300577", "postDate": "06/13/2023 09:09:37", "content": "I think your case might be different from my case…
", "rawMarkdown": "I think your case might be different from my case...", "votes": null } ], "comments": [ { "id": 2300055, "author_name": "brook1114", "author_url": "", "post_date": "06/13/2023 02:38:01", "content": "Scoring error same for me! My submission cannot get score since last night, just running and running over 15h by now!
", "votes": null, "replies": [ { "id": 2300062, "author_name": "gilgarad", "author_url": "", "post_date": "06/13/2023 02:48:32", "content": "did you try to submit again with the ones you already succeeded?
", "votes": null, "replies": [] } ] }, { "id": 2300066, "author_name": "occccn", "author_url": "", "post_date": "06/13/2023 02:50:47", "content": "I'm experiencing the same issue. The scoring has been running for more than 12 hours.
", "votes": null, "replies": [ { "id": 2300069, "author_name": "brook1114", "author_url": "", "post_date": "06/13/2023 02:52:13", "content": "Really exhausted in this competition…
", "votes": null, "replies": [ { "id": 2300082, "author_name": "occccn", "author_url": "", "post_date": "06/13/2023 03:04:42", "content": "I’m with you…
", "votes": null, "replies": [ { "id": 2300188, "author_name": "gilgarad", "author_url": "", "post_date": "06/13/2023 04:35:33", "content": "Did you check if rerun the ones you've succeeded not working as well??
", "votes": null, "replies": [ { "id": 2300243, "author_name": "occccn", "author_url": "", "post_date": "06/13/2023 05:15:37", "content": "I'm not trying it.
\nIn the end, I received an unfamiliar message called \"Kaggle error.\"
I have.
\nIf you have submitted a successful submission, but after re-scored, it becomes a submission scoring error.
\nThat's most likely a bug in your original submit script that wasn't shown as a failure (at least I did)
\nAfter I found those bug and modified them, I can submit normally.
I think your case might be different from my case…
", "votes": null, "replies": [] } ] }, { "id": 2300277, "author_name": "gilgarad", "author_url": "", "post_date": "06/13/2023 05:38:18", "content": "Hey guys, I've found my error case. Check if the case is applicable to yours. Thanks.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2300049": "Is it just me who's suffering from submission scoring error right now??\n\nWill this be re-scored again when the error's fixed? Or do I have to submit it again?\n\nI do not want to waste my submission try...\n\nI'm currently using jo wilder 3.7\n\n\n----\nI found out the problem and I think it may help you.\n\nWhile predicting and submitting, the data only comes with one session id. I was making some meta data with certain columns, if this column is under some condition. And I realized that in my train / test set, there's a bunch of other session id so that the feature columns will still remains even with nan values. But if featurizing with a single session id with the condition not applicable, those features disappeared.\n\nConsider this case\n\n\n```python\ncolumn = 'some_column'\n\ncond = df[column] != 0\ngb = df[cond].groupby(['session_id', 'level_group'])\nfeature = gb['elapsed_time'].mean()\n```\n\n\n\nIf cond is all False then feature will not be correctly generated. This happens while running jo_wilder because there is only one session id each time. \n\nI hope this can help.", "2300055": "Scoring error same for me! My submission cannot get score since last night, just running and running over 15h by now!", "2300062": "did you try to submit again with the ones you already succeeded?", "2300066": "I'm experiencing the same issue. The scoring has been running for more than 12 hours.", "2300069": "Really exhausted in this competition...", "2300082": "I’m with you...", "2300188": "Did you check if rerun the ones you've succeeded not working as well??", "2300209": "\n\nI have.\nIf you have submitted a successful submission, but after re-scored, it becomes a submission scoring error.\nThat's most likely a bug in your original submit script that wasn't shown as a failure (at least I did)\nAfter I found those bug and modified them, I can submit normally.", "2300243": "I'm not trying it.\nIn the end, I received an unfamiliar message called \"Kaggle error.\"", "2300277": "Hey guys, I've found my error case. Check if the case is applicable to yours. Thanks.", "2300577": "I think your case might be different from my case..." }, "source": "meta" }