{
  "id": 733,
  "title": "Use randomforest in R",
  "url": "/competitions/ClaimPredictionChallenge/discussion/733",
  "author_name": "",
  "post_date": "2011-07-26T06:10:29.930000",
  "votes": 1,
  "comment_count": 0,
  "views": 3361,
  "content": "<p>Hi guys, I just started to learn R, so would be really great if you guys could help with this :-)</p>\r\n<p>was trying to use randomforest in R on the train set, but got this error:</p>\r\n<p>something likes &quot;randomforest doesn't support categorical feature with more than 32 nominal values&quot;</p>\r\n<p>1) Could you please share some R code that could transform a categorical feature with, say 32&#43; nominal values to a feature uses only the top N distinct values?<br>\r\n<br>\r\n2) How can I transform a numeric feature to nominal feature using &quot;binning&quot; in R<br>\r\n<br>\r\nThanks a lot!</p>",
  "messages": [
    {
      "id": 4742,
      "postDate": "2011-07-26T06:10:29.930Z",
      "content": "<p>Hi guys, I just started to learn R, so would be really great if you guys could help with this :-)</p>\r\n<p>was trying to use randomforest in R on the train set, but got this error:</p>\r\n<p>something likes &quot;randomforest doesn't support categorical feature with more than 32 nominal values&quot;</p>\r\n<p>1) Could you please share some R code that could transform a categorical feature with, say 32&#43; nominal values to a feature uses only the top N distinct values?<br>\r\n<br>\r\n2) How can I transform a numeric feature to nominal feature using &quot;binning&quot; in R<br>\r\n<br>\r\nThanks a lot!</p>",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "4742": ""
  }
}