{
  "id": 75431,
  "title": "I find a Regularity , but don't know how to use it!",
  "url": "/competitions/elo-merchant-category-recommendation/discussion/75431",
  "author_name": "",
  "post_date": "2018-12-21T14:50:35.069117100Z",
  "votes": 8,
  "comment_count": 7,
  "views": 0,
  "content": "<p>First  filter and get only the outliers data in train.csv.\nSecond when setting feature_2 ==3 and feature_3 =1 then feature_1's value is 3.\nDo you have any tips?</p>",
  "messages": [
    {
      "id": "443395",
      "postDate": "12/21/2018 14:50:35",
      "content": "<p>First  filter and get only the outliers data in train.csv.\nSecond when setting feature_2 ==3 and feature_3 =1 then feature_1's value is 3.\nDo you have any tips?</p>",
      "rawMarkdown": "First  filter and get only the outliers data in train.csv.\nSecond when setting feature_2 ==3 and feature_3 =1 then feature_1's value is 3.\nDo you have any tips?",
      "votes": null
    },
    {
      "id": "443419",
      "postDate": "12/21/2018 15:48:13",
      "content": "<p>I find that whether outliers or not , when setting feature_2 = 3 and feature_3 =1,feature_1's value is always 3.And the combinations of the three features only have 14 different situations.But I don't know how to deal with these features in depth .\nHope it helps you!</p>",
      "rawMarkdown": "I find that whether outliers or not , when setting feature_2 = 3 and feature_3 =1,feature_1's value is always 3.And the combinations of the three features only have 14 different situations.But I don't know how to deal with these features in depth .\nHope it helps you!",
      "votes": null
    },
    {
      "id": "443494",
      "postDate": "12/21/2018 18:40:43",
      "content": "<p>Across the entire dataset, Feature 1 and feature 2 collectively determine feature 3. That is, every feature 1 + feature 2 combination has the same feature 3 value, regardless of outlier status.</p>",
      "rawMarkdown": "Across the entire dataset, Feature 1 and feature 2 collectively determine feature 3. That is, every feature 1 + feature 2 combination has the same feature 3 value, regardless of outlier status.",
      "votes": null
    },
    {
      "id": "443635",
      "postDate": "12/22/2018 01:44:26",
      "content": "<p>feature 1 + feature 2 combination has the same feature 3 value ?\nIt doesn't seem this.</p>",
      "rawMarkdown": "feature 1 + feature 2 combination has the same feature 3 value ?\nIt doesn't seem this.",
      "votes": null
    },
    {
      "id": "443636",
      "postDate": "12/22/2018 01:45:33",
      "content": "<p>Thank you!</p>",
      "rawMarkdown": "Thank you!",
      "votes": null
    },
    {
      "id": "443866",
      "postDate": "12/22/2018 16:11:01",
      "content": "<p>I think maybe because of only 14 different situations, so get_dummies for this features does't help.</p>",
      "rawMarkdown": "I think maybe because of only 14 different situations, so get_dummies for this features does't help.",
      "votes": null
    },
    {
      "id": "456081",
      "postDate": "01/15/2019 05:04:18",
      "content": "<p>So, may I think the reason for get_dummies not working is get_dummies can bring 2*5*3 situations? But only need 14 situations？</p>",
      "rawMarkdown": "So, may I think the reason for get_dummies not working is get_dummies can bring 2*5*3 situations? But only need 14 situations？",
      "votes": null
    },
    {
      "id": "456208",
      "postDate": "01/15/2019 11:01:08",
      "content": "<p>it's 5 * 3 * 2 = 30 </p>",
      "rawMarkdown": "it's 5 * 3 * 2 = 30",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 443419,
      "author_name": "wxxdepp",
      "author_url": "",
      "post_date": "12/21/2018 15:48:13",
      "content": "<p>I find that whether outliers or not , when setting feature_2 = 3 and feature_3 =1,feature_1's value is always 3.And the combinations of the three features only have 14 different situations.But I don't know how to deal with these features in depth .\nHope it helps you!</p>",
      "votes": null,
      "replies": [
        {
          "id": 443636,
          "author_name": "codlife",
          "author_url": "",
          "post_date": "12/22/2018 01:45:33",
          "content": "<p>Thank you!</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 443866,
          "author_name": "codlife",
          "author_url": "",
          "post_date": "12/22/2018 16:11:01",
          "content": "<p>I think maybe because of only 14 different situations, so get_dummies for this features does't help.</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 456081,
          "author_name": "loveyinuo",
          "author_url": "",
          "post_date": "01/15/2019 05:04:18",
          "content": "<p>So, may I think the reason for get_dummies not working is get_dummies can bring 2*5*3 situations? But only need 14 situations？</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 456208,
          "author_name": "codlife",
          "author_url": "",
          "post_date": "01/15/2019 11:01:08",
          "content": "<p>it's 5 * 3 * 2 = 30 </p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 443494,
      "author_name": "peterhurford",
      "author_url": "",
      "post_date": "12/21/2018 18:40:43",
      "content": "<p>Across the entire dataset, Feature 1 and feature 2 collectively determine feature 3. That is, every feature 1 + feature 2 combination has the same feature 3 value, regardless of outlier status.</p>",
      "votes": null,
      "replies": [
        {
          "id": 443635,
          "author_name": "codlife",
          "author_url": "",
          "post_date": "12/22/2018 01:44:26",
          "content": "<p>feature 1 + feature 2 combination has the same feature 3 value ?\nIt doesn't seem this.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "443395": "First  filter and get only the outliers data in train.csv.\nSecond when setting feature_2 ==3 and feature_3 =1 then feature_1's value is 3.\nDo you have any tips?",
    "443419": "I find that whether outliers or not , when setting feature_2 = 3 and feature_3 =1,feature_1's value is always 3.And the combinations of the three features only have 14 different situations.But I don't know how to deal with these features in depth .\nHope it helps you!",
    "443494": "Across the entire dataset, Feature 1 and feature 2 collectively determine feature 3. That is, every feature 1 + feature 2 combination has the same feature 3 value, regardless of outlier status.",
    "443635": "feature 1 + feature 2 combination has the same feature 3 value ?\nIt doesn't seem this.",
    "443636": "Thank you!",
    "443866": "I think maybe because of only 14 different situations, so get_dummies for this features does't help.",
    "456081": "So, may I think the reason for get_dummies not working is get_dummies can bring 2*5*3 situations? But only need 14 situations？",
    "456208": "it's 5 * 3 * 2 = 30"
  },
  "source": "meta"
}