{
  "id": 540436,
  "title": "Welcome Kagglers!",
  "url": "/competitions/jane-street-real-time-market-data-forecasting/discussion/540436",
  "author_name": "kait o'neil",
  "post_date": "2024-10-14T14:50:24.157000",
  "votes": 46,
  "comment_count": 43,
  "views": 0,
  "content": "<p>We're excited about seeing your innovative entries in this Kaggle competition. Please give us any feedback you have and feel free to ask questions; we look forward to engaging with you :)</p>\n<p>Jane Streeters love competitive challenges and puzzles, and we hope you enjoy working on this one.</p>\n<p>Good luck!<br>\nThe Jane Street Team</p>",
  "messages": [
    {
      "id": 3017143,
      "postDate": "2024-10-14T14:50:24.157Z",
      "content": "<p>We're excited about seeing your innovative entries in this Kaggle competition. Please give us any feedback you have and feel free to ask questions; we look forward to engaging with you :)</p>\n<p>Jane Streeters love competitive challenges and puzzles, and we hope you enjoy working on this one.</p>\n<p>Good luck!<br>\nThe Jane Street Team</p>",
      "rawMarkdown": "We're excited about seeing your innovative entries in this Kaggle competition. Please give us any feedback you have and feel free to ask questions; we look forward to engaging with you :)\n\nJane Streeters love competitive challenges and puzzles, and we hope you enjoy working on this one.\n\nGood luck!\nThe Jane Street Team",
      "votes": 46
    },
    {
      "id": 3035984,
      "postDate": "2024-11-04T04:25:22.647Z",
      "content": "<p>The current test.parquet dataset only contains one batch of date_id and time_id. Does the leaderboard uses test data with multiple date_ids and time_ids? If it's scored one date_id and time_id at a time, this will affect the settings of the models and a lot of time lagged features are not able to be created. Can we assume the test set has long enough date_ids and time_ids (eg. &gt; 5) to use? Thanks a lot!</p>",
      "rawMarkdown": "The current test.parquet dataset only contains one batch of date_id and time_id. Does the leaderboard uses test data with multiple date_ids and time_ids? If it's scored one date_id and time_id at a time, this will affect the settings of the models and a lot of time lagged features are not able to be created. Can we assume the test set has long enough date_ids and time_ids (eg. > 5) to use? Thanks a lot!",
      "votes": 5
    },
    {
      "id": 3057363,
      "postDate": "2024-11-28T04:18:31.907Z",
      "content": "<p>Hi, I apologize if this is a basic question, but based on the sentence:<br>\n\"The goal of the competition is to forecast one of these responders, i.e., responder_6, for up to six months in the future.\"</p>\n<p>Is the model expected to perform 1-step prediction (predicting responder_6 for the next time period and using actual values iteratively for subsequent predictions) or n-step prediction (predicting responder_6 for all time periods up to six months into the future in one go)? If it’s n-step prediction, will the prediction length (n) be fixed or variable?</p>",
      "rawMarkdown": "Hi, I apologize if this is a basic question, but based on the sentence:\n\"The goal of the competition is to forecast one of these responders, i.e., responder_6, for up to six months in the future.\"\n\nIs the model expected to perform 1-step prediction (predicting responder_6 for the next time period and using actual values iteratively for subsequent predictions) or n-step prediction (predicting responder_6 for all time periods up to six months into the future in one go)? If it’s n-step prediction, will the prediction length (n) be fixed or variable?",
      "votes": 3,
      "replies": [
        {
          "id": 3063231,
          "postDate": "2024-12-04T09:35:14.260Z",
          "content": "<p>These are exactly my questions as well and they have absolutely not clarified it anywhere? They just have a sample with all zeros that has 38 datapoints.</p>",
          "rawMarkdown": "These are exactly my questions as well and they have absolutely not clarified it anywhere? They just have a sample with all zeros that has 38 datapoints.",
          "replies": [
            {
              "id": 3065629,
              "postDate": "2024-12-07T04:48:57.383Z",
              "content": "<p>I think the \"6 months\" refers to the evaluation period from January through July (Jan 13 - July 12 = 6 months). I think the API requires we predict responder_6 for that time_id only.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F15357304%2Fe547a1033f154cd62b48da1387d1e10a%2FScreen%20Shot%202024-12-06%20at%208.46.19%20PM.png?generation=1733546875077884&amp;alt=media\" alt=\"\"></p>",
              "rawMarkdown": "I think the \"6 months\" refers to the evaluation period from January through July (Jan 13 - July 12 = 6 months). I think the API requires we predict responder_6 for that time_id only.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F15357304%2Fe547a1033f154cd62b48da1387d1e10a%2FScreen%20Shot%202024-12-06%20at%208.46.19%20PM.png?generation=1733546875077884&alt=media)"
            }
          ]
        }
      ]
    },
    {
      "id": 3036100,
      "postDate": "2024-11-04T08:11:39.637Z",
      "content": "<p>Here I am! My first Kaggle competition!</p>",
      "rawMarkdown": "Here I am! My first Kaggle competition!",
      "votes": 3,
      "replies": [
        {
          "id": 3079070,
          "postDate": "2024-12-23T07:37:42.367Z",
          "content": "<p>Same for me! </p>",
          "rawMarkdown": "Same for me! "
        }
      ]
    },
    {
      "id": 3029505,
      "postDate": "2024-10-27T12:01:15.310Z",
      "content": "<p>These posts <a href=\"https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542184\" target=\"_blank\">1</a> and <a href=\"https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542809\" target=\"_blank\">2</a> are asking about the <code>time_id</code> column. Could you confirm whether:</p>\n<ol>\n<li>Is it safe to assume that <code>time_id</code> will range from 0-967 in the forecasting phase?</li>\n<li>Do <code>time_id</code>s on different <code>date_id</code>s correspond to the same time of the day?</li>\n</ol>\n<p>Thanks a lot in advance.</p>",
      "rawMarkdown": "These posts [1](https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542184) and [2](https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542809) are asking about the `time_id` column. Could you confirm whether:\n\n1. Is it safe to assume that `time_id` will range from 0-967 in the forecasting phase?\n2. Do `time_id`s on different `date_id`s correspond to the same time of the day?\n\nThanks a lot in advance.",
      "votes": 3,
      "replies": [
        {
          "id": 3036206,
          "postDate": "2024-11-04T10:55:16.120Z",
          "content": "<p>Hello Aleix - </p>\n<ol>\n<li><p>I think it is not very likely that the range of time_ids will change in the forecasting phase, and we don't expect them to change. If something about the underlying market changes significantly then this might not be the case and we'll reply with that information if it comes to pass.</p></li>\n<li><p>There can be regime changes in the data where time_ids might switch to corresponding to different actual times of day, however we expect these regime changes to be very rare, and don't currently expect more such changes in the future. In the training data, however, it is not guaranteed that on two different date_ids, a specific time_id corresponds to the same time of day.</p></li>\n</ol>",
          "rawMarkdown": "Hello Aleix - \n\n1. I think it is not very likely that the range of time_ids will change in the forecasting phase, and we don't expect them to change. If something about the underlying market changes significantly then this might not be the case and we'll reply with that information if it comes to pass.\n\n2. There can be regime changes in the data where time_ids might switch to corresponding to different actual times of day, however we expect these regime changes to be very rare, and don't currently expect more such changes in the future. In the training data, however, it is not guaranteed that on two different date_ids, a specific time_id corresponds to the same time of day.",
          "votes": 5,
          "replies": [
            {
              "id": 3042028,
              "postDate": "2024-11-11T05:07:22.027Z",
              "content": "<p>Hi <a href=\"https://www.kaggle.com/mdesaijs\" target=\"_blank\">@mdesaijs</a>, thanks for the reply. I'm not sure if your answer was fully intentional since you didn't directly answer the question \"Is it safe to assume that time_id will range from 0-967 in the forecasting phase?\". You didn't say anything about time_id is being between 0 and 967 in current test set or the forecasting phase. You only said \"it won't change\", so should we infer that the number of time ids can be different than 968 right now.</p>",
              "rawMarkdown": "Hi @mdesaijs, thanks for the reply. I'm not sure if your answer was fully intentional since you didn't directly answer the question \"Is it safe to assume that time_id will range from 0-967 in the forecasting phase?\". You didn't say anything about time_id is being between 0 and 967 in current test set or the forecasting phase. You only said \"it won't change\", so should we infer that the number of time ids can be different than 968 right now.",
              "votes": 1
            }
          ]
        }
      ]
    },
    {
      "id": 3079104,
      "postDate": "2024-12-23T08:38:19.137Z",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/kaitoneil\" target=\"_blank\">@kaitoneil</a>, I had a quick question. So it says in the description that the goal is to forecast one of these responders, i.e., responder_6. Does this mean that the test will be asking specifically for predictions of responder_6, or that it could be asking for any one of the 8 responders. Thank you very much ! </p>",
      "rawMarkdown": "Hey @kaitoneil, I had a quick question. So it says in the description that the goal is to forecast one of these responders, i.e., responder_6. Does this mean that the test will be asking specifically for predictions of responder_6, or that it could be asking for any one of the 8 responders. Thank you very much ! ",
      "votes": 1,
      "replies": [
        {
          "id": 3079112,
          "postDate": "2024-12-23T08:48:26.567Z",
          "content": "<p>I also have this question. My current understanding is that you are judged by R squared score on this variable. </p>\n<p>Although this might not be how the leaderboard is ranked, as the all the top submissions have really low values in the \"scores\" column, values that would be not impressive if they represent R squared value. It leads me to believe that the score is some other metric of model performance.</p>",
          "rawMarkdown": "I also have this question. My current understanding is that you are judged by R squared score on this variable. \n\nAlthough this might not be how the leaderboard is ranked, as the all the top submissions have really low values in the \"scores\" column, values that would be not impressive if they represent R squared value. It leads me to believe that the score is some other metric of model performance."
        }
      ]
    },
    {
      "id": 3051044,
      "postDate": "2024-11-20T19:15:30.867Z",
      "content": "<p>Hey guys, can someone tell is there any pre-requisite knowledge one needs to start with these kind of competition or something to read up beforehand?</p>",
      "rawMarkdown": "Hey guys, can someone tell is there any pre-requisite knowledge one needs to start with these kind of competition or something to read up beforehand?",
      "votes": 1,
      "replies": [
        {
          "id": 3072343,
          "postDate": "2024-12-15T03:56:57.803Z",
          "content": "<p>I think if you ask about this one, then it might be materials about python, machine-learning, pytorch maybe and kaggle procedure?</p>",
          "rawMarkdown": "I think if you ask about this one, then it might be materials about python, machine-learning, pytorch maybe and kaggle procedure?\n"
        }
      ]
    },
    {
      "id": 3026979,
      "postDate": "2024-10-24T12:01:24.493Z",
      "content": "<p>Hello, Can we get accurate confirmation if there will be extra symbol_id (20 more)? Or is it a information that can not be disclosed? For example, in a kaggle quantitative trading competition , the host confirmed that there would not be any extra symbol_id.</p>",
      "rawMarkdown": "Hello, Can we get accurate confirmation if there will be extra symbol_id (20 more)? Or is it a information that can not be disclosed? For example, in a kaggle quantitative trading competition , the host confirmed that there would not be any extra symbol_id.",
      "votes": 1,
      "replies": [
        {
          "id": 3027002,
          "postDate": "2024-10-24T12:33:57.857Z",
          "content": "<p>The symbol universe is derived from real-world products that trade. We don't have any expectation that it will change significantly, and it is highly unlikely to change very significantly. There could potentially be small changes possible (most likely individual symbols being added or removed, but even this seems somewhat unlikely). Your solution should ideally be prepared to anticipate this if it happens, though, and we'll also try to give updates if it seems more likely to happen for any reason.</p>",
          "rawMarkdown": "The symbol universe is derived from real-world products that trade. We don't have any expectation that it will change significantly, and it is highly unlikely to change very significantly. There could potentially be small changes possible (most likely individual symbols being added or removed, but even this seems somewhat unlikely). Your solution should ideally be prepared to anticipate this if it happens, though, and we'll also try to give updates if it seems more likely to happen for any reason.",
          "votes": 6,
          "replies": [
            {
              "id": 3027025,
              "postDate": "2024-10-24T13:00:01.627Z",
              "content": "<p>Thank you very much for your detailed response. </p>",
              "rawMarkdown": "Thank you very much for your detailed response. "
            }
          ]
        }
      ]
    },
    {
      "id": 3149434,
      "postDate": "2025-03-14T07:55:03.153Z",
      "content": "<p>Interesting Competition.</p>",
      "rawMarkdown": "Interesting Competition."
    },
    {
      "id": 3094452,
      "postDate": "2025-01-12T05:55:35.287Z",
      "content": "<p>Please give me some help as a novice, if my code has not run the result by the end of the time (assuming that it unfortunately reports an error), then should I take the wrong code as the final version, or take the version with the highest score as the final version?</p>",
      "rawMarkdown": "Please give me some help as a novice, if my code has not run the result by the end of the time (assuming that it unfortunately reports an error), then should I take the wrong code as the final version, or take the version with the highest score as the final version?"
    },
    {
      "id": 3085209,
      "postDate": "2024-12-31T21:11:30.107Z",
      "content": "<p>hi  <a href=\"https://www.kaggle.com/kaitoneil\" target=\"_blank\">@kaitoneil</a> <br>\nI'm having the following error<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3410058%2F12be37cf023bbcc3fc1c9491aeb652bc%2FScreenshot%202025-01-01%20at%202.38.35AM.png?generation=1735679449849825&amp;alt=media\" alt=\"\"></p>\n<p>Could you please help? Thanks!</p>",
      "rawMarkdown": "hi  @kaitoneil \nI'm having the following error\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3410058%2F12be37cf023bbcc3fc1c9491aeb652bc%2FScreenshot%202025-01-01%20at%202.38.35AM.png?generation=1735679449849825&alt=media)\n\nCould you please help? Thanks!"
    },
    {
      "id": 3072983,
      "postDate": "2024-12-15T21:54:09.797Z",
      "content": "<p>Hello, I am struggling to read / load the parquet dataset due to memory limitations and errors I receive when trying to read train or lags dataset. </p>\n<p>Can you advise best way to read the data? </p>",
      "rawMarkdown": "Hello, I am struggling to read / load the parquet dataset due to memory limitations and errors I receive when trying to read train or lags dataset. \n\nCan you advise best way to read the data? "
    },
    {
      "id": 3072323,
      "postDate": "2024-12-15T03:20:23.530Z",
      "content": "<p>Excited to be here!</p>",
      "rawMarkdown": "Excited to be here!"
    },
    {
      "id": 3072290,
      "postDate": "2024-12-15T02:15:15.807Z",
      "content": "<p>Nice. Looking forward to the competition!</p>",
      "rawMarkdown": "Nice. Looking forward to the competition!"
    },
    {
      "id": 3071813,
      "postDate": "2024-12-14T11:03:58.403Z",
      "content": "<p>Hi there,</p>\n<p>Could you clarify the range of the iterator \"i\" in the scoring function? Does it span from 0 to the size of the training data?</p>\n<p>Thank you for sharing the dataset in this competition. I've learned so much about time series modeling through this competition!</p>",
      "rawMarkdown": "Hi there,\n\nCould you clarify the range of the iterator \"i\" in the scoring function? Does it span from 0 to the size of the training data?\n\nThank you for sharing the dataset in this competition. I've learned so much about time series modeling through this competition!"
    },
    {
      "id": 3069032,
      "postDate": "2024-12-11T01:29:50.363Z",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/mdesaijs\" target=\"_blank\">@mdesaijs</a> , would it possible that in the forecasting stage, on the same date, the numbers of symbol id across different time id are not the same? (e.g. 38 symbols in time id 0 and 39 symbols in time id 1).</p>\n<p>Thanks!</p>",
      "rawMarkdown": "Hi @mdesaijs , would it possible that in the forecasting stage, on the same date, the numbers of symbol id across different time id are not the same? (e.g. 38 symbols in time id 0 and 39 symbols in time id 1).\n\nThanks!"
    },
    {
      "id": 3066733,
      "postDate": "2024-12-08T12:27:27.453Z",
      "content": "<p>Look forward to it! This is my first kaggle competition.</p>",
      "rawMarkdown": "Look forward to it! This is my first kaggle competition."
    },
    {
      "id": 3057682,
      "postDate": "2024-11-28T13:40:02.870Z",
      "content": "<p>Best of luck everyone!</p>",
      "rawMarkdown": "Best of luck everyone!"
    },
    {
      "id": 3054739,
      "postDate": "2024-11-25T04:43:18.883Z",
      "content": "<p>Thank you for such an interesting competition!</p>",
      "rawMarkdown": "Thank you for such an interesting competition!"
    },
    {
      "id": 3042394,
      "postDate": "2024-11-11T13:39:13.117Z",
      "content": "<p>Looking forward to it!</p>",
      "rawMarkdown": "Looking forward to it!"
    },
    {
      "id": 3041553,
      "postDate": "2024-11-10T14:06:45.620Z",
      "content": "<p>i am facing issue while doing submission it runs for 9 hours during submission and then it fails  and i have checked the output also but why it is failing during submission</p>",
      "rawMarkdown": "i am facing issue while doing submission it runs for 9 hours during submission and then it fails  and i have checked the output also but why it is failing during submission"
    },
    {
      "id": 3037435,
      "postDate": "2024-11-05T17:40:14.187Z",
      "content": "<p>I need to finish what I'm doing and get in here as soon as possible.</p>",
      "rawMarkdown": "I need to finish what I'm doing and get in here as soon as possible."
    },
    {
      "id": 3035009,
      "postDate": "2024-11-02T23:21:56.753Z",
      "content": "<p>Awesome, looking forward to the competition.</p>",
      "rawMarkdown": "Awesome, looking forward to the competition."
    },
    {
      "id": 3031828,
      "postDate": "2024-10-30T07:15:18.110Z",
      "content": "<p>So last time of the Jane Street competition is 3y ago?</p>",
      "rawMarkdown": "So last time of the Jane Street competition is 3y ago?"
    },
    {
      "id": 3030721,
      "postDate": "2024-10-28T19:58:37.340Z",
      "content": "<p>Ready to start</p>",
      "rawMarkdown": "Ready to start"
    },
    {
      "id": 3030712,
      "postDate": "2024-10-28T19:38:57.600Z",
      "content": "<p>Thank you so much for this challenge. its not just about the prize money but the challenge is very welcome!</p>",
      "rawMarkdown": "Thank you so much for this challenge. its not just about the prize money but the challenge is very welcome!"
    },
    {
      "id": 3026532,
      "postDate": "2024-10-23T23:49:34.197Z",
      "content": "<p>Excited to be here!</p>",
      "rawMarkdown": "Excited to be here!"
    },
    {
      "id": 3026450,
      "postDate": "2024-10-23T19:57:59.823Z",
      "content": "<p>Hello, this is exciting data to deep dive.</p>",
      "rawMarkdown": "Hello, this is exciting data to deep dive."
    },
    {
      "id": 3026144,
      "postDate": "2024-10-23T13:37:07.917Z",
      "content": "<p>Thank you, Jane Street Team!<br>\nI appreciate the opportunity provided by the team to sharpen my skills, and I’m looking forward to diving into the problem.</p>",
      "rawMarkdown": "Thank you, Jane Street Team!\nI appreciate the opportunity provided by the team to sharpen my skills, and I’m looking forward to diving into the problem."
    },
    {
      "id": 3026143,
      "postDate": "2024-10-23T13:36:45.253Z",
      "rawMarkdown": "",
      "isDeleted": true
    },
    {
      "id": 3022895,
      "postDate": "2024-10-20T03:55:10.750Z",
      "content": "<p>Hi Kait,<br>\nWhat resources (books/articles/research papers) etc. do you recommend that will really help in the competition. I want to operate at a fundamental level with a strong foundation so that we are able to provide the best solution to the problem at hand.<br>\nLooking forward to your response!</p>",
      "rawMarkdown": "Hi Kait,\nWhat resources (books/articles/research papers) etc. do you recommend that will really help in the competition. I want to operate at a fundamental level with a strong foundation so that we are able to provide the best solution to the problem at hand.\nLooking forward to your response!",
      "votes": 2,
      "isDeleted": true,
      "replies": [
        {
          "id": 3024257,
          "postDate": "2024-10-21T13:16:25.193Z",
          "content": "<p>Hi Shubham! Thanks for reaching out. There are lots of great threads on the discussion boards that highlight resources - feel free to dive into those :)</p>\n<p>Good luck!</p>",
          "rawMarkdown": "Hi Shubham! Thanks for reaching out. There are lots of great threads on the discussion boards that highlight resources - feel free to dive into those :)\n\nGood luck!"
        },
        {
          "id": 3056384,
          "postDate": "2024-11-26T23:08:24.653Z",
          "content": "<p>Hi Shubham,</p>\n<p>if you have found some resources, please share. Also if you are open to learning together too</p>",
          "rawMarkdown": "Hi Shubham,\n\nif you have found some resources, please share. Also if you are open to learning together too"
        }
      ]
    },
    {
      "id": 3038474,
      "postDate": "2024-11-07T02:27:14.073Z",
      "content": "<p>Thank you!</p>",
      "rawMarkdown": "Thank you!"
    },
    {
      "id": 3029996,
      "postDate": "2024-10-28T03:28:11.713Z",
      "content": "<p>Thank you  Jane Street</p>",
      "rawMarkdown": "Thank you  Jane Street"
    }
  ],
  "comments": [
    {
      "id": 3035984,
      "author_name": "Salty Gold Fish",
      "author_url": "",
      "post_date": "2024-11-04T04:25:22.647000",
      "content": "<p>The current test.parquet dataset only contains one batch of date_id and time_id. Does the leaderboard uses test data with multiple date_ids and time_ids? If it's scored one date_id and time_id at a time, this will affect the settings of the models and a lot of time lagged features are not able to be created. Can we assume the test set has long enough date_ids and time_ids (eg. &gt; 5) to use? Thanks a lot!</p>",
      "votes": 5,
      "replies": []
    },
    {
      "id": 3057363,
      "author_name": "Syed Yasir",
      "author_url": "",
      "post_date": "2024-11-28T04:18:31.907000",
      "content": "<p>Hi, I apologize if this is a basic question, but based on the sentence:<br>\n\"The goal of the competition is to forecast one of these responders, i.e., responder_6, for up to six months in the future.\"</p>\n<p>Is the model expected to perform 1-step prediction (predicting responder_6 for the next time period and using actual values iteratively for subsequent predictions) or n-step prediction (predicting responder_6 for all time periods up to six months into the future in one go)? If it’s n-step prediction, will the prediction length (n) be fixed or variable?</p>",
      "votes": 3,
      "replies": [
        {
          "id": 3063231,
          "author_name": "sarashs",
          "author_url": "",
          "post_date": "2024-12-04T09:35:14.260000",
          "content": "<p>These are exactly my questions as well and they have absolutely not clarified it anywhere? They just have a sample with all zeros that has 38 datapoints.</p>",
          "votes": 0,
          "replies": [
            {
              "id": 3065629,
              "author_name": "Jim Beno",
              "author_url": "",
              "post_date": "2024-12-07T04:48:57.383000",
              "content": "<p>I think the \"6 months\" refers to the evaluation period from January through July (Jan 13 - July 12 = 6 months). I think the API requires we predict responder_6 for that time_id only.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F15357304%2Fe547a1033f154cd62b48da1387d1e10a%2FScreen%20Shot%202024-12-06%20at%208.46.19%20PM.png?generation=1733546875077884&amp;alt=media\" alt=\"\"></p>",
              "votes": 0,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 3036100,
      "author_name": "wbzdssfeq",
      "author_url": "",
      "post_date": "2024-11-04T08:11:39.637000",
      "content": "<p>Here I am! My first Kaggle competition!</p>",
      "votes": 3,
      "replies": [
        {
          "id": 3079070,
          "author_name": "Xboxgothacked",
          "author_url": "",
          "post_date": "2024-12-23T07:37:42.367000",
          "content": "<p>Same for me! </p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 3029505,
      "author_name": "AGG",
      "author_url": "",
      "post_date": "2024-10-27T12:01:15.310000",
      "content": "<p>These posts <a href=\"https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542184\" target=\"_blank\">1</a> and <a href=\"https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542809\" target=\"_blank\">2</a> are asking about the <code>time_id</code> column. Could you confirm whether:</p>\n<ol>\n<li>Is it safe to assume that <code>time_id</code> will range from 0-967 in the forecasting phase?</li>\n<li>Do <code>time_id</code>s on different <code>date_id</code>s correspond to the same time of the day?</li>\n</ol>\n<p>Thanks a lot in advance.</p>",
      "votes": 3,
      "replies": [
        {
          "id": 3036206,
          "author_name": "M Desai",
          "author_url": "",
          "post_date": "2024-11-04T10:55:16.120000",
          "content": "<p>Hello Aleix - </p>\n<ol>\n<li><p>I think it is not very likely that the range of time_ids will change in the forecasting phase, and we don't expect them to change. If something about the underlying market changes significantly then this might not be the case and we'll reply with that information if it comes to pass.</p></li>\n<li><p>There can be regime changes in the data where time_ids might switch to corresponding to different actual times of day, however we expect these regime changes to be very rare, and don't currently expect more such changes in the future. In the training data, however, it is not guaranteed that on two different date_ids, a specific time_id corresponds to the same time of day.</p></li>\n</ol>",
          "votes": 5,
          "replies": [
            {
              "id": 3042028,
              "author_name": "Gunes Evitan",
              "author_url": "",
              "post_date": "2024-11-11T05:07:22.027000",
              "content": "<p>Hi <a href=\"https://www.kaggle.com/mdesaijs\" target=\"_blank\">@mdesaijs</a>, thanks for the reply. I'm not sure if your answer was fully intentional since you didn't directly answer the question \"Is it safe to assume that time_id will range from 0-967 in the forecasting phase?\". You didn't say anything about time_id is being between 0 and 967 in current test set or the forecasting phase. You only said \"it won't change\", so should we infer that the number of time ids can be different than 968 right now.</p>",
              "votes": 1,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 3079104,
      "author_name": "Abel K",
      "author_url": "",
      "post_date": "2024-12-23T08:38:19.137000",
      "content": "<p>Hey <a href=\"https://www.kaggle.com/kaitoneil\" target=\"_blank\">@kaitoneil</a>, I had a quick question. So it says in the description that the goal is to forecast one of these responders, i.e., responder_6. Does this mean that the test will be asking specifically for predictions of responder_6, or that it could be asking for any one of the 8 responders. Thank you very much ! </p>",
      "votes": 1,
      "replies": [
        {
          "id": 3079112,
          "author_name": "Xboxgothacked",
          "author_url": "",
          "post_date": "2024-12-23T08:48:26.567000",
          "content": "<p>I also have this question. My current understanding is that you are judged by R squared score on this variable. </p>\n<p>Although this might not be how the leaderboard is ranked, as the all the top submissions have really low values in the \"scores\" column, values that would be not impressive if they represent R squared value. It leads me to believe that the score is some other metric of model performance.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 3051044,
      "author_name": "ALOK KUMAR SINGH",
      "author_url": "",
      "post_date": "2024-11-20T19:15:30.867000",
      "content": "<p>Hey guys, can someone tell is there any pre-requisite knowledge one needs to start with these kind of competition or something to read up beforehand?</p>",
      "votes": 1,
      "replies": [
        {
          "id": 3072343,
          "author_name": "*^-^* QinLilyHeA+",
          "author_url": "",
          "post_date": "2024-12-15T03:56:57.803000",
          "content": "<p>I think if you ask about this one, then it might be materials about python, machine-learning, pytorch maybe and kaggle procedure?</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 3026979,
      "author_name": "yuanzhe zhou",
      "author_url": "",
      "post_date": "2024-10-24T12:01:24.493000",
      "content": "<p>Hello, Can we get accurate confirmation if there will be extra symbol_id (20 more)? Or is it a information that can not be disclosed? For example, in a kaggle quantitative trading competition , the host confirmed that there would not be any extra symbol_id.</p>",
      "votes": 1,
      "replies": [
        {
          "id": 3027002,
          "author_name": "M Desai",
          "author_url": "",
          "post_date": "2024-10-24T12:33:57.857000",
          "content": "<p>The symbol universe is derived from real-world products that trade. We don't have any expectation that it will change significantly, and it is highly unlikely to change very significantly. There could potentially be small changes possible (most likely individual symbols being added or removed, but even this seems somewhat unlikely). Your solution should ideally be prepared to anticipate this if it happens, though, and we'll also try to give updates if it seems more likely to happen for any reason.</p>",
          "votes": 6,
          "replies": [
            {
              "id": 3027025,
              "author_name": "yuanzhe zhou",
              "author_url": "",
              "post_date": "2024-10-24T13:00:01.627000",
              "content": "<p>Thank you very much for your detailed response. </p>",
              "votes": 0,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 3149434,
      "author_name": "Minkyu Jo",
      "author_url": "",
      "post_date": "2025-03-14T07:55:03.153000",
      "content": "<p>Interesting Competition.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3094452,
      "author_name": "DongHang Zheng",
      "author_url": "",
      "post_date": "2025-01-12T05:55:35.287000",
      "content": "<p>Please give me some help as a novice, if my code has not run the result by the end of the time (assuming that it unfortunately reports an error), then should I take the wrong code as the final version, or take the version with the highest score as the final version?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3085209,
      "author_name": "the_astro_guy",
      "author_url": "",
      "post_date": "2024-12-31T21:11:30.107000",
      "content": "<p>hi  <a href=\"https://www.kaggle.com/kaitoneil\" target=\"_blank\">@kaitoneil</a> <br>\nI'm having the following error<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3410058%2F12be37cf023bbcc3fc1c9491aeb652bc%2FScreenshot%202025-01-01%20at%202.38.35AM.png?generation=1735679449849825&amp;alt=media\" alt=\"\"></p>\n<p>Could you please help? Thanks!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3072983,
      "author_name": "Kanan Taghiyev",
      "author_url": "",
      "post_date": "2024-12-15T21:54:09.797000",
      "content": "<p>Hello, I am struggling to read / load the parquet dataset due to memory limitations and errors I receive when trying to read train or lags dataset. </p>\n<p>Can you advise best way to read the data? </p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3072323,
      "author_name": "Jeffrey Zhang",
      "author_url": "",
      "post_date": "2024-12-15T03:20:23.530000",
      "content": "<p>Excited to be here!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3072290,
      "author_name": "Tommy Hsu",
      "author_url": "",
      "post_date": "2024-12-15T02:15:15.807000",
      "content": "<p>Nice. Looking forward to the competition!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3071813,
      "author_name": "yafs",
      "author_url": "",
      "post_date": "2024-12-14T11:03:58.403000",
      "content": "<p>Hi there,</p>\n<p>Could you clarify the range of the iterator \"i\" in the scoring function? Does it span from 0 to the size of the training data?</p>\n<p>Thank you for sharing the dataset in this competition. I've learned so much about time series modeling through this competition!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3069032,
      "author_name": "leadmaker",
      "author_url": "",
      "post_date": "2024-12-11T01:29:50.363000",
      "content": "<p>Hi <a href=\"https://www.kaggle.com/mdesaijs\" target=\"_blank\">@mdesaijs</a> , would it possible that in the forecasting stage, on the same date, the numbers of symbol id across different time id are not the same? (e.g. 38 symbols in time id 0 and 39 symbols in time id 1).</p>\n<p>Thanks!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3066733,
      "author_name": "Lulu Xu",
      "author_url": "",
      "post_date": "2024-12-08T12:27:27.453000",
      "content": "<p>Look forward to it! This is my first kaggle competition.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3057682,
      "author_name": "konstand",
      "author_url": "",
      "post_date": "2024-11-28T13:40:02.870000",
      "content": "<p>Best of luck everyone!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3054739,
      "author_name": "Adomas Fiseris",
      "author_url": "",
      "post_date": "2024-11-25T04:43:18.883000",
      "content": "<p>Thank you for such an interesting competition!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3042394,
      "author_name": "donzido",
      "author_url": "",
      "post_date": "2024-11-11T13:39:13.117000",
      "content": "<p>Looking forward to it!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3041553,
      "author_name": "Raviksh Singh Dikola",
      "author_url": "",
      "post_date": "2024-11-10T14:06:45.620000",
      "content": "<p>i am facing issue while doing submission it runs for 9 hours during submission and then it fails  and i have checked the output also but why it is failing during submission</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3037435,
      "author_name": "iddkjin",
      "author_url": "",
      "post_date": "2024-11-05T17:40:14.187000",
      "content": "<p>I need to finish what I'm doing and get in here as soon as possible.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3035009,
      "author_name": "Luke Grimes",
      "author_url": "",
      "post_date": "2024-11-02T23:21:56.753000",
      "content": "<p>Awesome, looking forward to the competition.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3031828,
      "author_name": "Wang Zilong 0718",
      "author_url": "",
      "post_date": "2024-10-30T07:15:18.110000",
      "content": "<p>So last time of the Jane Street competition is 3y ago?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3030721,
      "author_name": "LaLegumbre",
      "author_url": "",
      "post_date": "2024-10-28T19:58:37.340000",
      "content": "<p>Ready to start</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3030712,
      "author_name": "Wesly v. Aalst",
      "author_url": "",
      "post_date": "2024-10-28T19:38:57.600000",
      "content": "<p>Thank you so much for this challenge. its not just about the prize money but the challenge is very welcome!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3026532,
      "author_name": "Nhan Tran",
      "author_url": "",
      "post_date": "2024-10-23T23:49:34.197000",
      "content": "<p>Excited to be here!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3026450,
      "author_name": "blade",
      "author_url": "",
      "post_date": "2024-10-23T19:57:59.823000",
      "content": "<p>Hello, this is exciting data to deep dive.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3026144,
      "author_name": "Bhadra Mohit",
      "author_url": "",
      "post_date": "2024-10-23T13:37:07.917000",
      "content": "<p>Thank you, Jane Street Team!<br>\nI appreciate the opportunity provided by the team to sharpen my skills, and I’m looking forward to diving into the problem.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3026143,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-10-23T13:36:45.253000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3022895,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-10-20T03:55:10.750000",
      "content": "",
      "votes": 2,
      "replies": [
        {
          "id": 3024257,
          "author_name": "",
          "author_url": "",
          "post_date": "2024-10-21T13:16:25.193000",
          "content": "",
          "votes": 0,
          "replies": []
        },
        {
          "id": 3056384,
          "author_name": "",
          "author_url": "",
          "post_date": "2024-11-26T23:08:24.653000",
          "content": "",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 3038474,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-11-07T02:27:14.073000",
      "content": "",
      "votes": 0,
      "replies": []
    },
    {
      "id": 3029996,
      "author_name": "",
      "author_url": "",
      "post_date": "2024-10-28T03:28:11.713000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "3017143": "We're excited about seeing your innovative entries in this Kaggle competition. Please give us any feedback you have and feel free to ask questions; we look forward to engaging with you :)\n\nJane Streeters love competitive challenges and puzzles, and we hope you enjoy working on this one.\n\nGood luck!\nThe Jane Street Team",
    "3035984": "The current test.parquet dataset only contains one batch of date_id and time_id. Does the leaderboard uses test data with multiple date_ids and time_ids? If it's scored one date_id and time_id at a time, this will affect the settings of the models and a lot of time lagged features are not able to be created. Can we assume the test set has long enough date_ids and time_ids (eg. > 5) to use? Thanks a lot!",
    "3057363": "Hi, I apologize if this is a basic question, but based on the sentence:\n\"The goal of the competition is to forecast one of these responders, i.e., responder_6, for up to six months in the future.\"\n\nIs the model expected to perform 1-step prediction (predicting responder_6 for the next time period and using actual values iteratively for subsequent predictions) or n-step prediction (predicting responder_6 for all time periods up to six months into the future in one go)? If it’s n-step prediction, will the prediction length (n) be fixed or variable?",
    "3036100": "Here I am! My first Kaggle competition!",
    "3029505": "These posts [1](https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542184) and [2](https://www.kaggle.com/competitions/jane-street-real-time-market-data-forecasting/discussion/542809) are asking about the `time_id` column. Could you confirm whether:\n\n1. Is it safe to assume that `time_id` will range from 0-967 in the forecasting phase?\n2. Do `time_id`s on different `date_id`s correspond to the same time of the day?\n\nThanks a lot in advance.",
    "3079104": "Hey @kaitoneil, I had a quick question. So it says in the description that the goal is to forecast one of these responders, i.e., responder_6. Does this mean that the test will be asking specifically for predictions of responder_6, or that it could be asking for any one of the 8 responders. Thank you very much ! ",
    "3051044": "Hey guys, can someone tell is there any pre-requisite knowledge one needs to start with these kind of competition or something to read up beforehand?",
    "3026979": "Hello, Can we get accurate confirmation if there will be extra symbol_id (20 more)? Or is it a information that can not be disclosed? For example, in a kaggle quantitative trading competition , the host confirmed that there would not be any extra symbol_id.",
    "3149434": "Interesting Competition.",
    "3094452": "Please give me some help as a novice, if my code has not run the result by the end of the time (assuming that it unfortunately reports an error), then should I take the wrong code as the final version, or take the version with the highest score as the final version?",
    "3085209": "hi  @kaitoneil \nI'm having the following error\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3410058%2F12be37cf023bbcc3fc1c9491aeb652bc%2FScreenshot%202025-01-01%20at%202.38.35AM.png?generation=1735679449849825&alt=media)\n\nCould you please help? Thanks!",
    "3072983": "Hello, I am struggling to read / load the parquet dataset due to memory limitations and errors I receive when trying to read train or lags dataset. \n\nCan you advise best way to read the data? ",
    "3072323": "Excited to be here!",
    "3072290": "Nice. Looking forward to the competition!",
    "3071813": "Hi there,\n\nCould you clarify the range of the iterator \"i\" in the scoring function? Does it span from 0 to the size of the training data?\n\nThank you for sharing the dataset in this competition. I've learned so much about time series modeling through this competition!",
    "3069032": "Hi @mdesaijs , would it possible that in the forecasting stage, on the same date, the numbers of symbol id across different time id are not the same? (e.g. 38 symbols in time id 0 and 39 symbols in time id 1).\n\nThanks!",
    "3066733": "Look forward to it! This is my first kaggle competition.",
    "3057682": "Best of luck everyone!",
    "3054739": "Thank you for such an interesting competition!",
    "3042394": "Looking forward to it!",
    "3041553": "i am facing issue while doing submission it runs for 9 hours during submission and then it fails  and i have checked the output also but why it is failing during submission",
    "3037435": "I need to finish what I'm doing and get in here as soon as possible.",
    "3035009": "Awesome, looking forward to the competition.",
    "3031828": "So last time of the Jane Street competition is 3y ago?",
    "3030721": "Ready to start",
    "3030712": "Thank you so much for this challenge. its not just about the prize money but the challenge is very welcome!",
    "3026532": "Excited to be here!",
    "3026450": "Hello, this is exciting data to deep dive.",
    "3026144": "Thank you, Jane Street Team!\nI appreciate the opportunity provided by the team to sharpen my skills, and I’m looking forward to diving into the problem.",
    "3026143": "",
    "3022895": "Hi Kait,\nWhat resources (books/articles/research papers) etc. do you recommend that will really help in the competition. I want to operate at a fundamental level with a strong foundation so that we are able to provide the best solution to the problem at hand.\nLooking forward to your response!",
    "3038474": "Thank you!",
    "3029996": "Thank you  Jane Street"
  }
}