{
  "id": 190698,
  "title": "Groups in test_iter",
  "url": "/competitions/riiid-test-answer-prediction/discussion/190698",
  "author_name": "",
  "post_date": "2020-10-13T03:38:17.799946900Z",
  "votes": 1,
  "comment_count": 4,
  "views": 0,
  "content": "<p>Can we assume that there will be data from only one 'group_num' in every iteration of test_iter() ?</p>",
  "messages": [
    {
      "id": "1047900",
      "postDate": "10/13/2020 03:38:17",
      "content": "<p>Can we assume that there will be data from only one 'group_num' in every iteration of test_iter() ?</p>",
      "rawMarkdown": "Can we assume that there will be data from only one 'group_num' in every iteration of test_iter() ?",
      "votes": null
    },
    {
      "id": "1047970",
      "postDate": "10/13/2020 05:14:30",
      "content": "<p>Yep, I believe that's how it should be structured..</p>",
      "rawMarkdown": "Yep, I believe that's how it should be structured..",
      "votes": null
    },
    {
      "id": "1048565",
      "postDate": "10/13/2020 15:49:08",
      "content": "<p>Yes.                                </p>",
      "rawMarkdown": "Yes.",
      "votes": null
    },
    {
      "id": "1048649",
      "postDate": "10/13/2020 17:19:53",
      "content": "<p>Thanks! :)</p>",
      "rawMarkdown": "Thanks! :)",
      "votes": null
    },
    {
      "id": "1050631",
      "postDate": "10/15/2020 15:39:37",
      "content": "<p>Q1. <a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a> does this also mean that for <em>every</em> iteration on <code>test_iter()</code> we only need the first row to obtain the previous batch's responses and solutions (and some of the first rows can be empty lists)? </p>\n<p>The data description reads:</p>\n<p><code>prior_group_responses</code> (string) provides all of the user_answer entries for previous group in a string representation of a list in the <em>first row of the group</em>. All other rows in each group are null. If you are using Python, you will likely want to call eval on the non-null rows. <strong>Some rows may be null, or empty lists.</strong></p>\n<p>The first row for any group cannot be null but can only be an empty list. </p>\n<p>Q2. Probably a silly question: By <em>previous group</em> (in the data description) do we always mean the preceding batch of test_iter()? Is <code>group_num</code> is monotonically increasing?</p>",
      "rawMarkdown": "Q1. @sohier does this also mean that for *every* iteration on `test_iter()` we only need the first row to obtain the previous batch's responses and solutions (and some of the first rows can be empty lists)? \n\nThe data description reads:\n\n`prior_group_responses` (string) provides all of the user_answer entries for previous group in a string representation of a list in the *first row of the group*. All other rows in each group are null. If you are using Python, you will likely want to call eval on the non-null rows. **Some rows may be null, or empty lists.**\n\nThe first row for any group cannot be null but can only be an empty list. \n\nQ2. Probably a silly question: By *previous group* (in the data description) do we always mean the preceding batch of test_iter()? Is `group_num` is monotonically increasing?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1047970,
      "author_name": "adityaecdrid",
      "author_url": "",
      "post_date": "10/13/2020 05:14:30",
      "content": "<p>Yep, I believe that's how it should be structured..</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1048565,
      "author_name": "sohier",
      "author_url": "",
      "post_date": "10/13/2020 15:49:08",
      "content": "<p>Yes.                                </p>",
      "votes": null,
      "replies": [
        {
          "id": 1048649,
          "author_name": "abhimanyud",
          "author_url": "",
          "post_date": "10/13/2020 17:19:53",
          "content": "<p>Thanks! :)</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 1050631,
          "author_name": "doctorkael",
          "author_url": "",
          "post_date": "10/15/2020 15:39:37",
          "content": "<p>Q1. <a href=\"https://www.kaggle.com/sohier\" target=\"_blank\">@sohier</a> does this also mean that for <em>every</em> iteration on <code>test_iter()</code> we only need the first row to obtain the previous batch's responses and solutions (and some of the first rows can be empty lists)? </p>\n<p>The data description reads:</p>\n<p><code>prior_group_responses</code> (string) provides all of the user_answer entries for previous group in a string representation of a list in the <em>first row of the group</em>. All other rows in each group are null. If you are using Python, you will likely want to call eval on the non-null rows. <strong>Some rows may be null, or empty lists.</strong></p>\n<p>The first row for any group cannot be null but can only be an empty list. </p>\n<p>Q2. Probably a silly question: By <em>previous group</em> (in the data description) do we always mean the preceding batch of test_iter()? Is <code>group_num</code> is monotonically increasing?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1047900": "Can we assume that there will be data from only one 'group_num' in every iteration of test_iter() ?",
    "1047970": "Yep, I believe that's how it should be structured..",
    "1048565": "Yes.",
    "1048649": "Thanks! :)",
    "1050631": "Q1. @sohier does this also mean that for *every* iteration on `test_iter()` we only need the first row to obtain the previous batch's responses and solutions (and some of the first rows can be empty lists)? \n\nThe data description reads:\n\n`prior_group_responses` (string) provides all of the user_answer entries for previous group in a string representation of a list in the *first row of the group*. All other rows in each group are null. If you are using Python, you will likely want to call eval on the non-null rows. **Some rows may be null, or empty lists.**\n\nThe first row for any group cannot be null but can only be an empty list. \n\nQ2. Probably a silly question: By *previous group* (in the data description) do we always mean the preceding batch of test_iter()? Is `group_num` is monotonically increasing?"
  },
  "source": "meta"
}