{
  "id": 74173,
  "title": "A row in the test set has the a 'first_active_month' missing",
  "url": "/competitions/elo-merchant-category-recommendation/discussion/74173",
  "author_name": "",
  "post_date": "2018-12-09T16:48:56.635547700Z",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>I just discovered this ( a bit late perhaps but better than never I suppose), and I was wondering if people had bothered to handle this, and if so, how they had done it. </p>\n\n<p>This is Card_Id = 'C_ID_c27b4f80f7'</p>",
  "messages": [
    {
      "id": "436130",
      "postDate": "12/09/2018 16:48:56",
      "content": "<p>I just discovered this ( a bit late perhaps but better than never I suppose), and I was wondering if people had bothered to handle this, and if so, how they had done it. </p>\n\n<p>This is Card_Id = 'C_ID_c27b4f80f7'</p>",
      "rawMarkdown": "I just discovered this ( a bit late perhaps but better than never I suppose), and I was wondering if people had bothered to handle this, and if so, how they had done it. \n\nThis is Card_Id = 'C_ID_c27b4f80f7'",
      "votes": null
    },
    {
      "id": "436134",
      "postDate": "12/09/2018 16:51:36",
      "content": "<p>I replace it with the earliest data from the test set.</p>",
      "rawMarkdown": "I replace it with the earliest data from the test set.",
      "votes": null
    },
    {
      "id": "436138",
      "postDate": "12/09/2018 16:55:51",
      "content": "<p>So did I, but isn't that kindof arbitrary?</p>",
      "rawMarkdown": "So did I, but isn't that kindof arbitrary?",
      "votes": null
    },
    {
      "id": "436148",
      "postDate": "12/09/2018 17:24:20",
      "content": "<p>Though to be fair, I'm not sure how much of an impact a more measured analysis of one missing data point will make</p>",
      "rawMarkdown": "Though to be fair, I'm not sure how much of an impact a more measured analysis of one missing data point will make",
      "votes": null
    },
    {
      "id": "436153",
      "postDate": "12/09/2018 17:39:19",
      "content": "<p>Yea, if it would be even 1%, I'd bother with it. As it is only 1 sample - it should't have much impact.</p>",
      "rawMarkdown": "Yea, if it would be even 1%, I'd bother with it. As it is only 1 sample - it should't have much impact.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 436134,
      "author_name": "artgor",
      "author_url": "",
      "post_date": "12/09/2018 16:51:36",
      "content": "<p>I replace it with the earliest data from the test set.</p>",
      "votes": null,
      "replies": [
        {
          "id": 436138,
          "author_name": "delayedkarma",
          "author_url": "",
          "post_date": "12/09/2018 16:55:51",
          "content": "<p>So did I, but isn't that kindof arbitrary?</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 436148,
          "author_name": "delayedkarma",
          "author_url": "",
          "post_date": "12/09/2018 17:24:20",
          "content": "<p>Though to be fair, I'm not sure how much of an impact a more measured analysis of one missing data point will make</p>",
          "votes": null,
          "replies": []
        },
        {
          "id": 436153,
          "author_name": "artgor",
          "author_url": "",
          "post_date": "12/09/2018 17:39:19",
          "content": "<p>Yea, if it would be even 1%, I'd bother with it. As it is only 1 sample - it should't have much impact.</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "436130": "I just discovered this ( a bit late perhaps but better than never I suppose), and I was wondering if people had bothered to handle this, and if so, how they had done it. \n\nThis is Card_Id = 'C_ID_c27b4f80f7'",
    "436134": "I replace it with the earliest data from the test set.",
    "436138": "So did I, but isn't that kindof arbitrary?",
    "436148": "Though to be fair, I'm not sure how much of an impact a more measured analysis of one missing data point will make",
    "436153": "Yea, if it would be even 1%, I'd bother with it. As it is only 1 sample - it should't have much impact."
  },
  "source": "meta"
}