{
  "id": 93906,
  "title": "Can I use pre calculate data when I submit the final work?",
  "url": "/competitions/LANL-Earthquake-Prediction/discussion/93906",
  "author_name": "",
  "post_date": "2019-05-31T02:12:30.662790200Z",
  "votes": 2,
  "comment_count": 9,
  "views": 0,
  "content": "<p>Hi,</p>\n\n<p>I have some questions about final submission to ask. </p>\n\n<p>I believe most people load their own pre-calculated features (by adding datasets to the kernel) when we do the experiments. I am just wondering if we can still load the pre calculate training features in the final submitted version?  This will save time and there is no need to calculate the training features.</p>\n\n<p>By the way, besides \" submission.csv\",  the output files of my kernel have some other files generated by catboost, will the kaggle system identifies my submission.csv? Or I have to delete all irrelevant files?</p>\n\n<p>Thank you and good luck to your project!</p>",
  "messages": [
    {
      "id": "540081",
      "postDate": "05/31/2019 02:12:30",
      "content": "<p>Hi,</p>\n\n<p>I have some questions about final submission to ask. </p>\n\n<p>I believe most people load their own pre-calculated features (by adding datasets to the kernel) when we do the experiments. I am just wondering if we can still load the pre calculate training features in the final submitted version?  This will save time and there is no need to calculate the training features.</p>\n\n<p>By the way, besides \" submission.csv\",  the output files of my kernel have some other files generated by catboost, will the kaggle system identifies my submission.csv? Or I have to delete all irrelevant files?</p>\n\n<p>Thank you and good luck to your project!</p>",
      "rawMarkdown": "Hi,\n \nI have some questions about final submission to ask. \n\nI believe most people load their own pre-calculated features (by adding datasets to the kernel) when we do the experiments. I am just wondering if we can still load the pre calculate training features in the final submitted version?  This will save time and there is no need to calculate the training features.\n\nBy the way, besides \" submission.csv\",  the output files of my kernel have some other files generated by catboost, will the kaggle system identifies my submission.csv? Or I have to delete all irrelevant files?\n\nThank you and good luck to your project!",
      "votes": null
    },
    {
      "id": "540239",
      "postDate": "05/31/2019 07:48:22",
      "content": "<p>This is not a <a href=\"https://www.kaggle.com/docs/competitions#kernels-only-competitions\">Kernel-Only Competition</a>.\nOnly what you upload as submission file matters for the final leaderboard.  </p>\n\n<p>Keep in mind that if you finish \"in the money\" you must be able to reproduce the submission. See competition rules: <br>\n\"The delivered software code must be capable of generating the winning Submission and contain a description of resources required to build and/or run the executable code successfully\"</p>\n\n<p>Good luck! </p>",
      "rawMarkdown": "This is not a [Kernel-Only Competition](https://www.kaggle.com/docs/competitions#kernels-only-competitions).\nOnly what you upload as submission file matters for the final leaderboard.  \n\nKeep in mind that if you finish \"in the money\" you must be able to reproduce the submission. See competition rules:  \n\"The delivered software code must be capable of generating the winning Submission and contain a description of resources required to build and/or run the executable code successfully\"\n\nGood luck!",
      "votes": null
    },
    {
      "id": "540692",
      "postDate": "05/31/2019 21:25:25",
      "content": "<p>Sometimes I hope I'll never win a cash prize. I'd have to trawl through a hundred old kernels to retrace exactly what I did...</p>",
      "rawMarkdown": "Sometimes I hope I'll never win a cash prize. I'd have to trawl through a hundred old kernels to retrace exactly what I did...",
      "votes": null
    },
    {
      "id": "541135",
      "postDate": "06/01/2019 20:04:15",
      "content": "<p>Thanks for your reply! I still have the question about the submission. I used CatBoost and the output of the code contains many files generated by CatBoost besides my \"submission.csv\", I am wondering if system will identify my submission.csv or there is a place that I can choose which file to submit? Thank you!</p>",
      "rawMarkdown": "Thanks for your reply! I still have the question about the submission. I used CatBoost and the output of the code contains many files generated by CatBoost besides my \"submission.csv\", I am wondering if system will identify my submission.csv or there is a place that I can choose which file to submit? Thank you!",
      "votes": null
    },
    {
      "id": "541139",
      "postDate": "06/01/2019 20:28:45",
      "content": "<p>Running a kernel does not automatically submit any file to the competition. You need to manually select the file (the submission.csv in your case) and click \"submit to competition\". If the submission is successful you will see the score on the leaderboard.</p>\n\n<p>Hope it helps</p>",
      "rawMarkdown": "Running a kernel does not automatically submit any file to the competition. You need to manually select the file (the submission.csv in your case) and click \"submit to competition\". If the submission is successful you will see the score on the leaderboard.\n\nHope it helps",
      "votes": null
    },
    {
      "id": "541141",
      "postDate": "06/01/2019 20:42:26",
      "content": "<p><a href=\"/zechengzhang1\">@zechengzhang1</a> Go to \"My Submissions\" tab and manually select which ones you want scored. \"My Submissions\" tab is the second tab on the top right corner.</p>",
      "rawMarkdown": "zechengzhang1 Go to \"My Submissions\" tab and manually select which ones you want scored. \"My Submissions\" tab is the second tab on the top right corner.",
      "votes": null
    },
    {
      "id": "541784",
      "postDate": "06/03/2019 03:12:15",
      "content": "<p>Thanks @Tim! The thing is: currently when I commit my code, my the output contains many files, I then choose the submission.csv manually and submit to test on public dataset; but I am wondering will the system identify my submission.csv after the system commits my final work? Thank you! </p>\n\n<p>By the way, do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!</p>",
      "rawMarkdown": "Thanks @Tim! The thing is: currently when I commit my code, my the output contains many files, I then choose the submission.csv manually and submit to test on public dataset; but I am wondering will the system identify my submission.csv after the system commits my final work? Thank you! \n\nBy the way, do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!",
      "votes": null
    },
    {
      "id": "541786",
      "postDate": "06/03/2019 03:15:20",
      "content": "<p>Hi, blue train, thanks for your help! You mean I also need to manually select the file (submission.csv) to submit after the system commits my final work? Thank you! </p>\n\n<p>By the way,  do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!</p>",
      "rawMarkdown": "Hi, blue train, thanks for your help! You mean I also need to manually select the file (submission.csv) to submit after the system commits my final work? Thank you! \n\nBy the way,  do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!",
      "votes": null
    },
    {
      "id": "541870",
      "postDate": "06/03/2019 06:20:43",
      "content": "<p>Yes, you need to click on the submission.csv file and search for \"submit to competition\" or somethin similar and click on that. If you don't see the score of your model on the public LB it means that you did NOT submit anything!</p>\n\n<p>As of Catboost, I cannot help on that sorry...</p>",
      "rawMarkdown": "Yes, you need to click on the submission.csv file and search for \"submit to competition\" or somethin similar and click on that. If you don't see the score of your model on the public LB it means that you did NOT submit anything!\n\nAs of Catboost, I cannot help on that sorry...",
      "votes": null
    },
    {
      "id": "542370",
      "postDate": "06/03/2019 19:11:13",
      "content": "<p>Thank you bluetrain. </p>",
      "rawMarkdown": "Thank you bluetrain.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 540239,
      "author_name": "stecasasso",
      "author_url": "",
      "post_date": "05/31/2019 07:48:22",
      "content": "<p>This is not a <a href=\"https://www.kaggle.com/docs/competitions#kernels-only-competitions\">Kernel-Only Competition</a>.\nOnly what you upload as submission file matters for the final leaderboard.  </p>\n\n<p>Keep in mind that if you finish \"in the money\" you must be able to reproduce the submission. See competition rules: <br>\n\"The delivered software code must be capable of generating the winning Submission and contain a description of resources required to build and/or run the executable code successfully\"</p>\n\n<p>Good luck! </p>",
      "votes": null,
      "replies": [
        {
          "id": 540692,
          "author_name": "bigironsphere",
          "author_url": "",
          "post_date": "05/31/2019 21:25:25",
          "content": "<p>Sometimes I hope I'll never win a cash prize. I'd have to trawl through a hundred old kernels to retrace exactly what I did...</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 541135,
          "author_name": "zechengzhang1",
          "author_url": "",
          "post_date": "06/01/2019 20:04:15",
          "content": "<p>Thanks for your reply! I still have the question about the submission. I used CatBoost and the output of the code contains many files generated by CatBoost besides my \"submission.csv\", I am wondering if system will identify my submission.csv or there is a place that I can choose which file to submit? Thank you!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 541139,
          "author_name": "stecasasso",
          "author_url": "",
          "post_date": "06/01/2019 20:28:45",
          "content": "<p>Running a kernel does not automatically submit any file to the competition. You need to manually select the file (the submission.csv in your case) and click \"submit to competition\". If the submission is successful you will see the score on the leaderboard.</p>\n\n<p>Hope it helps</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 541141,
          "author_name": "teeyee314",
          "author_url": "",
          "post_date": "06/01/2019 20:42:26",
          "content": "<p><a href=\"/zechengzhang1\">@zechengzhang1</a> Go to \"My Submissions\" tab and manually select which ones you want scored. \"My Submissions\" tab is the second tab on the top right corner.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 541784,
          "author_name": "zechengzhang1",
          "author_url": "",
          "post_date": "06/03/2019 03:12:15",
          "content": "<p>Thanks @Tim! The thing is: currently when I commit my code, my the output contains many files, I then choose the submission.csv manually and submit to test on public dataset; but I am wondering will the system identify my submission.csv after the system commits my final work? Thank you! </p>\n\n<p>By the way, do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 541786,
          "author_name": "zechengzhang1",
          "author_url": "",
          "post_date": "06/03/2019 03:15:20",
          "content": "<p>Hi, blue train, thanks for your help! You mean I also need to manually select the file (submission.csv) to submit after the system commits my final work? Thank you! </p>\n\n<p>By the way,  do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 541870,
          "author_name": "stecasasso",
          "author_url": "",
          "post_date": "06/03/2019 06:20:43",
          "content": "<p>Yes, you need to click on the submission.csv file and search for \"submit to competition\" or somethin similar and click on that. If you don't see the score of your model on the public LB it means that you did NOT submit anything!</p>\n\n<p>As of Catboost, I cannot help on that sorry...</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 542370,
          "author_name": "zechengzhang1",
          "author_url": "",
          "post_date": "06/03/2019 19:11:13",
          "content": "<p>Thank you bluetrain. </p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "540081": "Hi,\n \nI have some questions about final submission to ask. \n\nI believe most people load their own pre-calculated features (by adding datasets to the kernel) when we do the experiments. I am just wondering if we can still load the pre calculate training features in the final submitted version?  This will save time and there is no need to calculate the training features.\n\nBy the way, besides \" submission.csv\",  the output files of my kernel have some other files generated by catboost, will the kaggle system identifies my submission.csv? Or I have to delete all irrelevant files?\n\nThank you and good luck to your project!",
    "540239": "This is not a [Kernel-Only Competition](https://www.kaggle.com/docs/competitions#kernels-only-competitions).\nOnly what you upload as submission file matters for the final leaderboard.  \n\nKeep in mind that if you finish \"in the money\" you must be able to reproduce the submission. See competition rules:  \n\"The delivered software code must be capable of generating the winning Submission and contain a description of resources required to build and/or run the executable code successfully\"\n\nGood luck!",
    "540692": "Sometimes I hope I'll never win a cash prize. I'd have to trawl through a hundred old kernels to retrace exactly what I did...",
    "541135": "Thanks for your reply! I still have the question about the submission. I used CatBoost and the output of the code contains many files generated by CatBoost besides my \"submission.csv\", I am wondering if system will identify my submission.csv or there is a place that I can choose which file to submit? Thank you!",
    "541139": "Running a kernel does not automatically submit any file to the competition. You need to manually select the file (the submission.csv in your case) and click \"submit to competition\". If the submission is successful you will see the score on the leaderboard.\n\nHope it helps",
    "541141": "zechengzhang1 Go to \"My Submissions\" tab and manually select which ones you want scored. \"My Submissions\" tab is the second tab on the top right corner.",
    "541784": "Thanks @Tim! The thing is: currently when I commit my code, my the output contains many files, I then choose the submission.csv manually and submit to test on public dataset; but I am wondering will the system identify my submission.csv after the system commits my final work? Thank you! \n\nBy the way, do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!",
    "541786": "Hi, blue train, thanks for your help! You mean I also need to manually select the file (submission.csv) to submit after the system commits my final work? Thank you! \n\nBy the way,  do you know how to prevent CatBoostRegressor generating a lot of useless output files? All my irrelevant output files come from CatBoostRegressor. Thank you!",
    "541870": "Yes, you need to click on the submission.csv file and search for \"submit to competition\" or somethin similar and click on that. If you don't see the score of your model on the public LB it means that you did NOT submit anything!\n\nAs of Catboost, I cannot help on that sorry...",
    "542370": "Thank you bluetrain."
  },
  "source": "meta"
}