{
  "id": 189513,
  "title": "task_container_id is not monotonically increasing for each user.",
  "url": "/competitions/riiid-test-answer-prediction/discussion/189513",
  "author_name": "",
  "post_date": "2020-10-07T20:27:07.136317600Z",
  "votes": 2,
  "comment_count": 1,
  "views": 0,
  "content": "<p><strong>Update</strong></p>\n<p>Just saw that <a href=\"https://www.kaggle.com/zhaochow\" target=\"_blank\">@zhaochow</a> already mentioned this issue here <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465</a></p>\n<p>and another discussion <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351</a></p>\n<p>Hi, </p>\n<p>I found something confusing.</p>\n<p>From Data tab page <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/data\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/data</a>, it claims that</p>\n<pre><code>task_container_id: (int16) Id code for the batch of questions or lectures. For example, a user might see three questions in a row before seeing the explanations for any of them. Those three would all share a task_container_id. Monotonically increasing for each user.\n</code></pre>\n<p>However, for user_id = <code>115</code>, I get <code>task_container_id</code> like</p>\n<pre><code>115\n[1, 2, 0, 3, 4, 5, 6, 7, 8, 9, 10, ....]\n</code></pre>\n<p>And for user_id = <code>24418</code>, <code>task_container_id</code> are</p>\n<pre><code>[0, 1, 2, 3, 4, 4, 4, ...,20, 22, 23, 24, 21, 25, ...... , 169, 171, 172, 173, 170, 174, ....\n</code></pre>\n<p>I verified the timestamps being increasing for the list of task_container_id per users.</p>\n<p>This is the code sample to get the information</p>\n<pre><code>    import pandas\n\n    train_df = pd.read_csv(\n        '/kaggle/input/riiid-test-answer-prediction/train.csv', iterator=True,\n    )\n\n    user_actions = {}\n\n    df = train_df.get_chunk(10000)\n    groups = df.groupby('user_id')\n\n    for user_id, info in groups:\n\n        data = list(groups.get_group(user_id)[['task_container_id']].itertuples(index=False, name=None))\n        user_actions[user_id] = data\n</code></pre>\n<p><a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a> : Do you know what happens here? Could you contact the host to clarify this claim about <code>task_container_id</code>?</p>",
  "messages": [
    {
      "id": "1041576",
      "postDate": "10/07/2020 20:27:07",
      "content": "<p><strong>Update</strong></p>\n<p>Just saw that <a href=\"https://www.kaggle.com/zhaochow\" target=\"_blank\">@zhaochow</a> already mentioned this issue here <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465</a></p>\n<p>and another discussion <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351</a></p>\n<p>Hi, </p>\n<p>I found something confusing.</p>\n<p>From Data tab page <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/data\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/data</a>, it claims that</p>\n<pre><code>task_container_id: (int16) Id code for the batch of questions or lectures. For example, a user might see three questions in a row before seeing the explanations for any of them. Those three would all share a task_container_id. Monotonically increasing for each user.\n</code></pre>\n<p>However, for user_id = <code>115</code>, I get <code>task_container_id</code> like</p>\n<pre><code>115\n[1, 2, 0, 3, 4, 5, 6, 7, 8, 9, 10, ....]\n</code></pre>\n<p>And for user_id = <code>24418</code>, <code>task_container_id</code> are</p>\n<pre><code>[0, 1, 2, 3, 4, 4, 4, ...,20, 22, 23, 24, 21, 25, ...... , 169, 171, 172, 173, 170, 174, ....\n</code></pre>\n<p>I verified the timestamps being increasing for the list of task_container_id per users.</p>\n<p>This is the code sample to get the information</p>\n<pre><code>    import pandas\n\n    train_df = pd.read_csv(\n        '/kaggle/input/riiid-test-answer-prediction/train.csv', iterator=True,\n    )\n\n    user_actions = {}\n\n    df = train_df.get_chunk(10000)\n    groups = df.groupby('user_id')\n\n    for user_id, info in groups:\n\n        data = list(groups.get_group(user_id)[['task_container_id']].itertuples(index=False, name=None))\n        user_actions[user_id] = data\n</code></pre>\n<p><a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a> : Do you know what happens here? Could you contact the host to clarify this claim about <code>task_container_id</code>?</p>",
      "rawMarkdown": "**Update**\n\nJust saw that @zhaochow already mentioned this issue here https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465\n\nand another discussion [https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351](https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351)\n\n\nHi, \n\nI found something confusing.\n\nFrom Data tab page [https://www.kaggle.com/c/riiid-test-answer-prediction/data](https://www.kaggle.com/c/riiid-test-answer-prediction/data), it claims that\n\n```\ntask_container_id: (int16) Id code for the batch of questions or lectures. For example, a user might see three questions in a row before seeing the explanations for any of them. Those three would all share a task_container_id. Monotonically increasing for each user.\n```\n\nHowever, for user_id = `115`, I get `task_container_id` like\n\n```\n115\n[1, 2, 0, 3, 4, 5, 6, 7, 8, 9, 10, ....]\n```\n\nAnd for user_id = `24418`, `task_container_id` are\n\n```\n[0, 1, 2, 3, 4, 4, 4, ...,20, 22, 23, 24, 21, 25, ...... , 169, 171, 172, 173, 170, 174, ....\n```\nI verified the timestamps being increasing for the list of task_container_id per users.\n\nThis is the code sample to get the information\n\n```\n\timport pandas\n\n\ttrain_df = pd.read_csv(\n\t    '/kaggle/input/riiid-test-answer-prediction/train.csv', iterator=True,\n\t)\n\n\tuser_actions = {}\n\n\tdf = train_df.get_chunk(10000)\n\tgroups = df.groupby('user_id')\n\n\tfor user_id, info in groups:\n\n\t    data = list(groups.get_group(user_id)[['task_container_id']].itertuples(index=False, name=None))\n\t    user_actions[user_id] = data\n```\n@sohier : Do you know what happens here? Could you contact the host to clarify this claim about `task_container_id`?",
      "votes": null
    },
    {
      "id": "1041581",
      "postDate": "10/07/2020 20:31:30",
      "content": "<p>Just saw that <a href=\"https://www.kaggle.com/zhaochow\" target=\"_blank\">@zhaochow</a> already mentioned this issue here <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465</a></p>",
      "rawMarkdown": "Just saw that @zhaochow already mentioned this issue here [https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465](https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465)",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1041581,
      "author_name": "yihdarshieh",
      "author_url": "",
      "post_date": "10/07/2020 20:31:30",
      "content": "<p>Just saw that <a href=\"https://www.kaggle.com/zhaochow\" target=\"_blank\">@zhaochow</a> already mentioned this issue here <a href=\"https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465\" target=\"_blank\">https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465</a></p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1041576": "**Update**\n\nJust saw that @zhaochow already mentioned this issue here https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465\n\nand another discussion [https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351](https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189351)\n\n\nHi, \n\nI found something confusing.\n\nFrom Data tab page [https://www.kaggle.com/c/riiid-test-answer-prediction/data](https://www.kaggle.com/c/riiid-test-answer-prediction/data), it claims that\n\n```\ntask_container_id: (int16) Id code for the batch of questions or lectures. For example, a user might see three questions in a row before seeing the explanations for any of them. Those three would all share a task_container_id. Monotonically increasing for each user.\n```\n\nHowever, for user_id = `115`, I get `task_container_id` like\n\n```\n115\n[1, 2, 0, 3, 4, 5, 6, 7, 8, 9, 10, ....]\n```\n\nAnd for user_id = `24418`, `task_container_id` are\n\n```\n[0, 1, 2, 3, 4, 4, 4, ...,20, 22, 23, 24, 21, 25, ...... , 169, 171, 172, 173, 170, 174, ....\n```\nI verified the timestamps being increasing for the list of task_container_id per users.\n\nThis is the code sample to get the information\n\n```\n\timport pandas\n\n\ttrain_df = pd.read_csv(\n\t    '/kaggle/input/riiid-test-answer-prediction/train.csv', iterator=True,\n\t)\n\n\tuser_actions = {}\n\n\tdf = train_df.get_chunk(10000)\n\tgroups = df.groupby('user_id')\n\n\tfor user_id, info in groups:\n\n\t    data = list(groups.get_group(user_id)[['task_container_id']].itertuples(index=False, name=None))\n\t    user_actions[user_id] = data\n```\n@sohier : Do you know what happens here? Could you contact the host to clarify this claim about `task_container_id`?",
    "1041581": "Just saw that @zhaochow already mentioned this issue here [https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465](https://www.kaggle.com/c/riiid-test-answer-prediction/discussion/189465)"
  },
  "source": "meta"
}