{
  "id": 333668,
  "title": "Strange plot(D86_mean)",
  "url": "/competitions/amex-default-prediction/discussion/333668",
  "author_name": "",
  "post_date": "2022-06-27T16:34:33.574839600Z",
  "votes": 7,
  "comment_count": 5,
  "views": 0,
  "content": "<p>It is a curious pattern, but difficult to consider.<br>\nThere seems to be a correlation with the count.</p>",
  "messages": [
    {
      "id": "1835303",
      "postDate": "06/27/2022 16:34:33",
      "content": "<p>It is a curious pattern, but difficult to consider.<br>\nThere seems to be a correlation with the count.</p>",
      "rawMarkdown": "It is a curious pattern, but difficult to consider.\nThere seems to be a correlation with the count.",
      "votes": null
    },
    {
      "id": "1835309",
      "postDate": "06/27/2022 16:36:36",
      "content": "<p>the image file</p>",
      "rawMarkdown": "the image file",
      "votes": null
    },
    {
      "id": "1835336",
      "postDate": "06/27/2022 17:05:16",
      "content": "<p>other images</p>",
      "rawMarkdown": "other images",
      "votes": null
    },
    {
      "id": "1835358",
      "postDate": "06/27/2022 17:26:19",
      "content": "<p>Thank you for sharing! This is quite interesting. I am surprised even std features show similar kind of patterns.</p>",
      "rawMarkdown": "Thank you for sharing! This is quite interesting. I am surprised even std features show similar kind of patterns.",
      "votes": null
    },
    {
      "id": "1835515",
      "postDate": "06/27/2022 19:42:27",
      "content": "<p>It's easy to explain:</p>\n<ul>\n<li>D_86 is always 0 or 1. If you take the mean of two such values, the result is 0, 0.5 or 1. If you take the mean of three such values, the result is 0, 0.33, 0.67 or 1. If you take the mean of four such values, the result is 0, 0.25, 0.5, 0.75 or 1. And so on…</li>\n<li>The same holds for R_25 and B_31, which are always 0 or 1.</li>\n<li>If you compute the standard deviation of a small array of zeros and ones, the set of possible outcomes is again finite…</li>\n</ul>",
      "rawMarkdown": "It's easy to explain:\n- D_86 is always 0 or 1. If you take the mean of two such values, the result is 0, 0.5 or 1. If you take the mean of three such values, the result is 0, 0.33, 0.67 or 1. If you take the mean of four such values, the result is 0, 0.25, 0.5, 0.75 or 1. And so on...\n- The same holds for R_25 and B_31, which are always 0 or 1.\n- If you compute the standard deviation of a small array of zeros and ones, the set of possible outcomes is again finite...",
      "votes": null
    },
    {
      "id": "1836692",
      "postDate": "06/29/2022 01:44:22",
      "content": "<p>every feature could be a derived feature or origin feature. If it is derived, for example, it may be created by some formula of other features, then there could be certain patterns. but digging too much looks not profitable</p>",
      "rawMarkdown": "every feature could be a derived feature or origin feature. If it is derived, for example, it may be created by some formula of other features, then there could be certain patterns. but digging too much looks not profitable",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1835309,
      "author_name": "miyawakiyoshifumi",
      "author_url": "",
      "post_date": "06/27/2022 16:36:36",
      "content": "<p>the image file</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1835336,
      "author_name": "miyawakiyoshifumi",
      "author_url": "",
      "post_date": "06/27/2022 17:05:16",
      "content": "<p>other images</p>",
      "votes": null,
      "replies": [
        {
          "id": 1835358,
          "author_name": "raphael1123",
          "author_url": "",
          "post_date": "06/27/2022 17:26:19",
          "content": "<p>Thank you for sharing! This is quite interesting. I am surprised even std features show similar kind of patterns.</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 1835515,
      "author_name": "ambrosm",
      "author_url": "",
      "post_date": "06/27/2022 19:42:27",
      "content": "<p>It's easy to explain:</p>\n<ul>\n<li>D_86 is always 0 or 1. If you take the mean of two such values, the result is 0, 0.5 or 1. If you take the mean of three such values, the result is 0, 0.33, 0.67 or 1. If you take the mean of four such values, the result is 0, 0.25, 0.5, 0.75 or 1. And so on…</li>\n<li>The same holds for R_25 and B_31, which are always 0 or 1.</li>\n<li>If you compute the standard deviation of a small array of zeros and ones, the set of possible outcomes is again finite…</li>\n</ul>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1836692,
      "author_name": "kongliangyu",
      "author_url": "",
      "post_date": "06/29/2022 01:44:22",
      "content": "<p>every feature could be a derived feature or origin feature. If it is derived, for example, it may be created by some formula of other features, then there could be certain patterns. but digging too much looks not profitable</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1835303": "It is a curious pattern, but difficult to consider.\nThere seems to be a correlation with the count.",
    "1835309": "the image file",
    "1835336": "other images",
    "1835358": "Thank you for sharing! This is quite interesting. I am surprised even std features show similar kind of patterns.",
    "1835515": "It's easy to explain:\n- D_86 is always 0 or 1. If you take the mean of two such values, the result is 0, 0.5 or 1. If you take the mean of three such values, the result is 0, 0.33, 0.67 or 1. If you take the mean of four such values, the result is 0, 0.25, 0.5, 0.75 or 1. And so on...\n- The same holds for R_25 and B_31, which are always 0 or 1.\n- If you compute the standard deviation of a small array of zeros and ones, the set of possible outcomes is again finite...",
    "1836692": "every feature could be a derived feature or origin feature. If it is derived, for example, it may be created by some formula of other features, then there could be certain patterns. but digging too much looks not profitable"
  },
  "source": "meta"
}