{
  "id": 22136,
  "title": "Running linear regression with Null Values?",
  "url": "/competitions/grupo-bimbo-inventory-demand/discussion/22136",
  "author_name": "",
  "post_date": "2016-07-09T19:31:37.420Z",
  "votes": null,
  "comment_count": 1,
  "views": 457,
  "content": "<p>I changed my data set to this way and hope to run Linear regression with 3 weeks of data to predict the fourth one, and keep iterating.</p>\n\n<p>How would you guys deal with the Null values?  </p>",
  "messages": [
    {
      "id": "126559",
      "postDate": "07/09/2016 19:31:37",
      "content": "<p>I changed my data set to this way and hope to run Linear regression with 3 weeks of data to predict the fourth one, and keep iterating.</p>\n\n<p>How would you guys deal with the Null values?  </p>",
      "rawMarkdown": "I changed my data set to this way and hope to run Linear regression with 3 weeks of data to predict the fourth one, and keep iterating.\r\n\r\nHow would you guys deal with the Null values?",
      "votes": null
    },
    {
      "id": "126560",
      "postDate": "07/09/2016 19:40:58",
      "content": "<p>In this case it looks like your NaN's are the result of there being no Demand (i.e. no line item for that Semana/Cliente/Producto) for weeks 3-9, correct? I think I'd just fill them in with 0's to start and see how that goes. </p>",
      "rawMarkdown": "In this case it looks like your NaN's are the result of there being no Demand (i.e. no line item for that Semana/Cliente/Producto) for weeks 3-9, correct? I think I'd just fill them in with 0's to start and see how that goes.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 126560,
      "author_name": "brandenkmurray",
      "author_url": "",
      "post_date": "07/09/2016 19:40:58",
      "content": "<p>In this case it looks like your NaN's are the result of there being no Demand (i.e. no line item for that Semana/Cliente/Producto) for weeks 3-9, correct? I think I'd just fill them in with 0's to start and see how that goes. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "126559": "I changed my data set to this way and hope to run Linear regression with 3 weeks of data to predict the fourth one, and keep iterating.\r\n\r\nHow would you guys deal with the Null values?",
    "126560": "In this case it looks like your NaN's are the result of there being no Demand (i.e. no line item for that Semana/Cliente/Producto) for weeks 3-9, correct? I think I'd just fill them in with 0's to start and see how that goes."
  },
  "source": "meta"
}