{
  "id": 14130,
  "title": "training/test set variable difference randomForest (R)",
  "url": "/competitions/expedia-personalized-sort/discussion/14130",
  "author_name": "apples",
  "post_date": "2015-05-09T16:38:54.123000",
  "votes": 0,
  "comment_count": 0,
  "views": 930,
  "content": "<p>Hello guys,</p>\n<p>I have a rather basic question regarding the randomForest package in R.</p>\n<p>I see that everybody adds new variables to the training set (some even up to 30!).</p>\n<p>However, how do you <em>test</em> your model with all these new variables, since randomForest&nbsp; will give an error message when you don't use the same variables in the test set:</p>\n<p><strong>random forest Error in eval(expr, envir, enclos) : object X not found</strong></p>\n\n<p>Sorry for my ignorance!</p>",
  "messages": [
    {
      "id": 77714,
      "postDate": "2015-05-09T16:38:54.123Z",
      "content": "<p>Hello guys,</p>\n<p>I have a rather basic question regarding the randomForest package in R.</p>\n<p>I see that everybody adds new variables to the training set (some even up to 30!).</p>\n<p>However, how do you <em>test</em> your model with all these new variables, since randomForest&nbsp; will give an error message when you don't use the same variables in the test set:</p>\n<p><strong>random forest Error in eval(expr, envir, enclos) : object X not found</strong></p>\n\n<p>Sorry for my ignorance!</p>"
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "77714": ""
  }
}