![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F10210684%2F844915b4b3d26488f906b6f01f540e0f%2FScreenshot%202023-05-13%20at%2018.23.37.png?generation=1683969861251741&alt=media\")
i am stuck and really frustrated, i have no idea what the problem is, my submissions return \"scoring error\" since yesterday, while everything seem to work perfectly :/ is the problem in my notebook and i should investigate further or is it in the API ?? CAN SOMEONE CONFIRM PLEAASE
", "messages": [ { "id": "2248210", "postDate": "05/06/2023 15:29:28", "content": "i am stuck and really frustrated, i have no idea what the problem is, my submissions return \"scoring error\" since yesterday, while everything seem to work perfectly :/ is the problem in my notebook and i should investigate further or is it in the API ?? CAN SOMEONE CONFIRM PLEAASE
", "rawMarkdown": "i am stuck and really frustrated, i have no idea what the problem is, my submissions return \"scoring error\" since yesterday, while everything seem to work perfectly :/ is the problem in my notebook and i should investigate further or is it in the API ?? CAN SOMEONE CONFIRM PLEAASE", "votes": null }, { "id": "2248295", "postDate": "05/06/2023 17:29:55", "content": "Me too. What's happening?
", "rawMarkdown": "Me too. What's happening?", "votes": null }, { "id": "2248585", "postDate": "05/07/2023 02:47:57", "content": "I've just been getting notebook timeout errors but that's a separate issue that's on my end as far as I know. It appears that they changed something with the dataset at least from what other people are saying. Hard to find any additional information, I think they're still rerunning notebooks for the leaderboard update, and possibly still changing things with the API. I guess we just have to be patient and wait till an official release from the team in a day, I'm right here with you with the frustration though.
", "rawMarkdown": "I've just been getting notebook timeout errors but that's a separate issue that's on my end as far as I know. It appears that they changed something with the dataset at least from what other people are saying. Hard to find any additional information, I think they're still rerunning notebooks for the leaderboard update, and possibly still changing things with the API. I guess we just have to be patient and wait till an official release from the team in a day, I'm right here with you with the frustration though.", "votes": null }, { "id": "2250866", "postDate": "05/08/2023 21:07:37", "content": "the recent release talks only about LB update, nothing about submission errors.
", "rawMarkdown": "the recent release talks only about LB update, nothing about submission errors.", "votes": null }, { "id": "2250872", "postDate": "05/08/2023 21:11:20", "content": "in my case i've found that the problem starts in the second iteration, when test data is data of group_level 5-12, but i am pulling my hair to know what the problem is, why the code doesn't work for this group, and obviously NO help nor indication from the API,
", "rawMarkdown": "in my case i've found that the problem starts in the second iteration, when test data is data of group_level 5-12, but i am pulling my hair to know what the problem is, why the code doesn't work for this group, and obviously NO help nor indication from the API,", "votes": null }, { "id": "2252088", "postDate": "05/09/2023 21:56:22", "content": "I have a similar problem, the difference is that it occurs in the level_group 13-22 …. i'm really frustated
", "rawMarkdown": "I have a similar problem, the difference is that it occurs in the level_group 13-22 .... i'm really frustated", "votes": null }, { "id": "2253603", "postDate": "05/10/2023 10:19:43", "content": "@adrielcabral you're right, i found that too, and found also which line of code causes the error, but still don't know how to solve it, at first i thought that converting NANs to INT was the issue, but nah, i made sure that the same features in the same order occurs in the testdata (it costs me 15 wasted submissions) but vainly… did you manage to find something ??
", "rawMarkdown": "adrielcabral you're right, i found that too, and found also which line of code causes the error, but still don't know how to solve it, at first i thought that converting NANs to INT was the issue, but nah, i made sure that the same features in the same order occurs in the testdata (it costs me 15 wasted submissions) but vainly... did you manage to find something ??", "votes": null }, { "id": "2253631", "postDate": "05/10/2023 10:57:37", "content": "I had this problem when I first started the competition, for me the problem was that I did joining operation to combine features from different levels. This made that columns are missing when players skip certain levels while submitting.
\nThe solution is to pad out missing columns using column names from training when submitting. I hope that this helps.
", "rawMarkdown": "I had this problem when I first started the competition, for me the problem was that I did joining operation to combine features from different levels. This made that columns are missing when players skip certain levels while submitting. \n\nThe solution is to pad out missing columns using column names from training when submitting. I hope that this helps.", "votes": null }, { "id": "2253678", "postDate": "05/10/2023 11:47:14", "content": "i've just found the same thing… where were you 3 days ago😅😅?
", "rawMarkdown": "i've just found the same thing... where were you 3 days ago😅😅?", "votes": null }, { "id": "2253681", "postDate": "05/10/2023 11:48:18", "content": "Sorry for the late reply :D
", "rawMarkdown": "Sorry for the late reply :D", "votes": null }, { "id": "2253683", "postDate": "05/10/2023 11:49:06", "content": "@adrielcabral , it turns out to be the level variable, there are missing levels in the group 13-12.
", "rawMarkdown": "adrielcabral , it turns out to be the level variable, there are missing levels in the group 13-12.", "votes": null }, { "id": "2253987", "postDate": "05/10/2023 15:15:45", "content": "i trying handle missing columns with this code:
\ndf = df.reindex(columns=df_13_22_columns)\ndf = df.replace(np.nan, )\nbut without success
\nthanks for u help !!! i'll keep trying
", "rawMarkdown": "i trying handle missing columns with this code:\n\n```python\ndf = df.reindex(columns=df_13_22_columns)\ndf = df.replace(np.nan, 0)\n```\n\nbut without success\n\nthanks for u help !!! i'll keep trying", "votes": null }, { "id": "2254001", "postDate": "05/10/2023 15:29:26", "content": "try using 'category' type of pandas , it preserves the categories you want even when they are missing
", "rawMarkdown": "try using 'category' type of pandas , it preserves the categories you want even when they are missing", "votes": null }, { "id": "2254010", "postDate": "05/10/2023 15:39:18", "content": "thanks !!! i'll try
", "rawMarkdown": "thanks !!! i'll try", "votes": null }, { "id": "2254201", "postDate": "05/10/2023 18:06:44", "content": "nothing yet… i'm feeling so stupid, the error continues
", "rawMarkdown": "nothing yet... i'm feeling so stupid, the error continues", "votes": null }, { "id": "2254262", "postDate": "05/10/2023 19:12:22", "content": "try to create a dataset containing all anomalies you could think of, and adjust your function to handle all the erros… be aware that other types of errors may occur in the private testset.
", "rawMarkdown": "try to create a dataset containing all anomalies you could think of, and adjust your function to handle all the erros... be aware that other types of errors may occur in the private testset.", "votes": null }, { "id": "2254345", "postDate": "05/10/2023 21:39:38", "content": "I'll keep trying :). Do you think this could cause a Gap between CV and LB? because we are using columns in the training that may not be present in the test data, that is, we are using \"false data\" in that missing columns which can be harmful
", "rawMarkdown": "I'll keep trying :). Do you think this could cause a Gap between CV and LB? because we are using columns in the training that may not be present in the test data, that is, we are using \"false data\" in that missing columns which can be harmful", "votes": null }, { "id": "2254380", "postDate": "05/10/2023 22:41:03", "content": "actually the same columns should appear in both train and test set, their values change but they should exist, let's say for example you calculate the elapsed time by level for each session_id, and in the testset, there is no level 19 for some players, the column elapsed_time_level_19 should appear and take value 0 because players didn't play that level then spent 0 time in it, we're not using any false data here.
", "rawMarkdown": "actually the same columns should appear in both train and test set, their values change but they should exist, let's say for example you calculate the elapsed time by level for each session_id, and in the testset, there is no level 19 for some players, the column elapsed_time_level_19 should appear and take value 0 because players didn't play that level then spent 0 time in it, we're not using any false data here.", "votes": null }, { "id": "2254412", "postDate": "05/11/2023 00:25:27", "content": "you're right, it does make sense
", "rawMarkdown": "you're right, it does make sense", "votes": null }, { "id": "2256129", "postDate": "05/12/2023 09:17:34", "content": "problem solved !! using df.reindex was correct but i was using it wrong, thank you very much @wouldyoujustfocus @woprime
@adrielcabral whew! what a relief, good luck with this painful competition😅
", "rawMarkdown": "adrielcabral whew! what a relief, good luck with this painful competition😅", "votes": null }, { "id": "2256182", "postDate": "05/12/2023 09:57:00", "content": "I'm just in this competition to learn new pandas commands and work on creativity with feature engineering 😊. I'm not even expecting to get on the bronze zone 😂😂 thanks ! good luck for u too @wouldyoujustfocus
", "rawMarkdown": "I'm just in this competition to learn new pandas commands and work on creativity with feature engineering 😊. I'm not even expecting to get on the bronze zone 😂😂 thanks ! good luck for u too @wouldyoujustfocus", "votes": null }, { "id": "2257351", "postDate": "05/13/2023 09:26:05", "content": "@wouldyoujustfocus I'm sorry for being late but my problems seems to be different from yours. I can get this output file as submission.csv and that is the same structure of my previous, successed submission. However I got scoring error. Why….
\n
The error presented to me was also a scoring error, but in the end it had nothing to do with the submission, but columns that were used in training, but were not present when created from the submission data. Even if yours is a 'scoring error', I would check if there were any changes in the code that were likely to introduce a lack of columns, which was my case.
\nA good start is to compare the code of a successful submissive and the one that is failing.
", "rawMarkdown": "The error presented to me was also a scoring error, but in the end it had nothing to do with the submission, but columns that were used in training, but were not present when created from the submission data. Even if yours is a 'scoring error', I would check if there were any changes in the code that were likely to introduce a lack of columns, which was my case.\n\nA good start is to compare the code of a successful submissive and the one that is failing.", "votes": null }, { "id": "2261608", "postDate": "05/16/2023 13:01:56", "content": "@adrielcabral Thank you so much. I successfully resolved the error! As you said, the cause is that the num of columns was different from train data and test data.
", "rawMarkdown": "adrielcabral Thank you so much. I successfully resolved the error! As you said, the cause is that the num of columns was different from train data and test data.", "votes": null } ], "comments": [ { "id": 2248295, "author_name": "itsunori", "author_url": "", "post_date": "05/06/2023 17:29:55", "content": "Me too. What's happening?
", "votes": null, "replies": [ { "id": 2250872, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/08/2023 21:11:20", "content": "in my case i've found that the problem starts in the second iteration, when test data is data of group_level 5-12, but i am pulling my hair to know what the problem is, why the code doesn't work for this group, and obviously NO help nor indication from the API,
", "votes": null, "replies": [ { "id": 2252088, "author_name": "adrielcabral", "author_url": "", "post_date": "05/09/2023 21:56:22", "content": "I have a similar problem, the difference is that it occurs in the level_group 13-22 …. i'm really frustated
", "votes": null, "replies": [ { "id": 2253603, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/10/2023 10:19:43", "content": "@adrielcabral you're right, i found that too, and found also which line of code causes the error, but still don't know how to solve it, at first i thought that converting NANs to INT was the issue, but nah, i made sure that the same features in the same order occurs in the testdata (it costs me 15 wasted submissions) but vainly… did you manage to find something ??
", "votes": null, "replies": [] } ] }, { "id": 2253683, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/10/2023 11:49:06", "content": "@adrielcabral , it turns out to be the level variable, there are missing levels in the group 13-12.
", "votes": null, "replies": [ { "id": 2253987, "author_name": "adrielcabral", "author_url": "", "post_date": "05/10/2023 15:15:45", "content": "i trying handle missing columns with this code:
\ndf = df.reindex(columns=df_13_22_columns)\ndf = df.replace(np.nan, )\nbut without success
\nthanks for u help !!! i'll keep trying
", "votes": null, "replies": [ { "id": 2254001, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/10/2023 15:29:26", "content": "try using 'category' type of pandas , it preserves the categories you want even when they are missing
", "votes": null, "replies": [ { "id": 2254010, "author_name": "adrielcabral", "author_url": "", "post_date": "05/10/2023 15:39:18", "content": "thanks !!! i'll try
", "votes": null, "replies": [ { "id": 2254201, "author_name": "adrielcabral", "author_url": "", "post_date": "05/10/2023 18:06:44", "content": "nothing yet… i'm feeling so stupid, the error continues
", "votes": null, "replies": [ { "id": 2254262, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/10/2023 19:12:22", "content": "try to create a dataset containing all anomalies you could think of, and adjust your function to handle all the erros… be aware that other types of errors may occur in the private testset.
", "votes": null, "replies": [ { "id": 2254345, "author_name": "adrielcabral", "author_url": "", "post_date": "05/10/2023 21:39:38", "content": "I'll keep trying :). Do you think this could cause a Gap between CV and LB? because we are using columns in the training that may not be present in the test data, that is, we are using \"false data\" in that missing columns which can be harmful
", "votes": null, "replies": [ { "id": 2254380, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/10/2023 22:41:03", "content": "actually the same columns should appear in both train and test set, their values change but they should exist, let's say for example you calculate the elapsed time by level for each session_id, and in the testset, there is no level 19 for some players, the column elapsed_time_level_19 should appear and take value 0 because players didn't play that level then spent 0 time in it, we're not using any false data here.
", "votes": null, "replies": [ { "id": 2254412, "author_name": "adrielcabral", "author_url": "", "post_date": "05/11/2023 00:25:27", "content": "you're right, it does make sense
", "votes": null, "replies": [] } ] } ] } ] } ] } ] } ] } ] } ] }, { "id": 2257351, "author_name": "itsunori", "author_url": "", "post_date": "05/13/2023 09:26:05", "content": "@wouldyoujustfocus I'm sorry for being late but my problems seems to be different from yours. I can get this output file as submission.csv and that is the same structure of my previous, successed submission. However I got scoring error. Why….
\n
The error presented to me was also a scoring error, but in the end it had nothing to do with the submission, but columns that were used in training, but were not present when created from the submission data. Even if yours is a 'scoring error', I would check if there were any changes in the code that were likely to introduce a lack of columns, which was my case.
\nA good start is to compare the code of a successful submissive and the one that is failing.
", "votes": null, "replies": [ { "id": 2261608, "author_name": "itsunori", "author_url": "", "post_date": "05/16/2023 13:01:56", "content": "@adrielcabral Thank you so much. I successfully resolved the error! As you said, the cause is that the num of columns was different from train data and test data.
", "votes": null, "replies": [] } ] } ] } ] } ] }, { "id": 2248585, "author_name": "artberz", "author_url": "", "post_date": "05/07/2023 02:47:57", "content": "I've just been getting notebook timeout errors but that's a separate issue that's on my end as far as I know. It appears that they changed something with the dataset at least from what other people are saying. Hard to find any additional information, I think they're still rerunning notebooks for the leaderboard update, and possibly still changing things with the API. I guess we just have to be patient and wait till an official release from the team in a day, I'm right here with you with the frustration though.
", "votes": null, "replies": [ { "id": 2250866, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/08/2023 21:07:37", "content": "the recent release talks only about LB update, nothing about submission errors.
", "votes": null, "replies": [] } ] }, { "id": 2253631, "author_name": "woprime", "author_url": "", "post_date": "05/10/2023 10:57:37", "content": "I had this problem when I first started the competition, for me the problem was that I did joining operation to combine features from different levels. This made that columns are missing when players skip certain levels while submitting.
\nThe solution is to pad out missing columns using column names from training when submitting. I hope that this helps.
", "votes": null, "replies": [ { "id": 2253678, "author_name": "wouldyoujustfocus", "author_url": "", "post_date": "05/10/2023 11:47:14", "content": "i've just found the same thing… where were you 3 days ago😅😅?
", "votes": null, "replies": [ { "id": 2253681, "author_name": "woprime", "author_url": "", "post_date": "05/10/2023 11:48:18", "content": "Sorry for the late reply :D
", "votes": null, "replies": [] } ] }, { "id": 2256129, "author_name": "adrielcabral", "author_url": "", "post_date": "05/12/2023 09:17:34", "content": "problem solved !! using df.reindex was correct but i was using it wrong, thank you very much @wouldyoujustfocus @woprime
@adrielcabral whew! what a relief, good luck with this painful competition😅
", "votes": null, "replies": [ { "id": 2256182, "author_name": "adrielcabral", "author_url": "", "post_date": "05/12/2023 09:57:00", "content": "I'm just in this competition to learn new pandas commands and work on creativity with feature engineering 😊. I'm not even expecting to get on the bronze zone 😂😂 thanks ! good luck for u too @wouldyoujustfocus
", "votes": null, "replies": [] } ] } ] } ] } ], "raw_markdown_by_id": { "2248210": "i am stuck and really frustrated, i have no idea what the problem is, my submissions return \"scoring error\" since yesterday, while everything seem to work perfectly :/ is the problem in my notebook and i should investigate further or is it in the API ?? CAN SOMEONE CONFIRM PLEAASE", "2248295": "Me too. What's happening?", "2248585": "I've just been getting notebook timeout errors but that's a separate issue that's on my end as far as I know. It appears that they changed something with the dataset at least from what other people are saying. Hard to find any additional information, I think they're still rerunning notebooks for the leaderboard update, and possibly still changing things with the API. I guess we just have to be patient and wait till an official release from the team in a day, I'm right here with you with the frustration though.", "2250866": "the recent release talks only about LB update, nothing about submission errors.", "2250872": "in my case i've found that the problem starts in the second iteration, when test data is data of group_level 5-12, but i am pulling my hair to know what the problem is, why the code doesn't work for this group, and obviously NO help nor indication from the API,", "2252088": "I have a similar problem, the difference is that it occurs in the level_group 13-22 .... i'm really frustated", "2253603": "adrielcabral you're right, i found that too, and found also which line of code causes the error, but still don't know how to solve it, at first i thought that converting NANs to INT was the issue, but nah, i made sure that the same features in the same order occurs in the testdata (it costs me 15 wasted submissions) but vainly... did you manage to find something ??", "2253631": "I had this problem when I first started the competition, for me the problem was that I did joining operation to combine features from different levels. This made that columns are missing when players skip certain levels while submitting. \n\nThe solution is to pad out missing columns using column names from training when submitting. I hope that this helps.", "2253678": "i've just found the same thing... where were you 3 days ago😅😅?", "2253681": "Sorry for the late reply :D", "2253683": "adrielcabral , it turns out to be the level variable, there are missing levels in the group 13-12.", "2253987": "i trying handle missing columns with this code:\n\n```python\ndf = df.reindex(columns=df_13_22_columns)\ndf = df.replace(np.nan, 0)\n```\n\nbut without success\n\nthanks for u help !!! i'll keep trying", "2254001": "try using 'category' type of pandas , it preserves the categories you want even when they are missing", "2254010": "thanks !!! i'll try", "2254201": "nothing yet... i'm feeling so stupid, the error continues", "2254262": "try to create a dataset containing all anomalies you could think of, and adjust your function to handle all the erros... be aware that other types of errors may occur in the private testset.", "2254345": "I'll keep trying :). Do you think this could cause a Gap between CV and LB? because we are using columns in the training that may not be present in the test data, that is, we are using \"false data\" in that missing columns which can be harmful", "2254380": "actually the same columns should appear in both train and test set, their values change but they should exist, let's say for example you calculate the elapsed time by level for each session_id, and in the testset, there is no level 19 for some players, the column elapsed_time_level_19 should appear and take value 0 because players didn't play that level then spent 0 time in it, we're not using any false data here.", "2254412": "you're right, it does make sense", "2256129": "problem solved !! using `df.reindex` was correct but i was using it wrong, thank you very much @wouldyoujustfocus @woprime", "2256158": "adrielcabral whew! what a relief, good luck with this painful competition😅", "2256182": "I'm just in this competition to learn new pandas commands and work on creativity with feature engineering 😊. I'm not even expecting to get on the bronze zone 😂😂 thanks ! good luck for u too @wouldyoujustfocus", "2257351": "wouldyoujustfocus I'm sorry for being late but my problems seems to be different from yours. I can get this output file as submission.csv and that is the same structure of my previous, successed submission. However I got scoring error. Why....\n\n\n", "2258005": "The error presented to me was also a scoring error, but in the end it had nothing to do with the submission, but columns that were used in training, but were not present when created from the submission data. Even if yours is a 'scoring error', I would check if there were any changes in the code that were likely to introduce a lack of columns, which was my case.\n\nA good start is to compare the code of a successful submissive and the one that is failing.", "2261608": "adrielcabral Thank you so much. I successfully resolved the error! As you said, the cause is that the num of columns was different from train data and test data." }, "source": "meta" }