{
  "id": 39838,
  "title": "How can i process so big dataset?",
  "url": "/competitions/kkbox-churn-prediction-challenge/discussion/39838",
  "author_name": "",
  "post_date": "2017-09-22T01:55:45.667939200Z",
  "votes": 2,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Its my first time that processing so big dataset. How much memory do i need for the competition? 16GB? 32GB? Is there any way to solve this memeoy problem ?</p>",
  "messages": [
    {
      "id": "223356",
      "postDate": "09/22/2017 01:55:45",
      "content": "<p>Its my first time that processing so big dataset. How much memory do i need for the competition? 16GB? 32GB? Is there any way to solve this memeoy problem ?</p>",
      "rawMarkdown": "Its my first time that processing so big dataset. How much memory do i need for the competition? 16GB? 32GB? Is there any way to solve this memeoy problem ?",
      "votes": null
    },
    {
      "id": "223424",
      "postDate": "09/22/2017 03:40:02",
      "content": "<p>There are some tricks such as computing the specific features of a smaller set of users at one time or removing the unnecessary features when you load the data. You can find how other people deal with it in Kernels page.</p>",
      "rawMarkdown": "There are some tricks such as computing the specific features of a smaller set of users at one time or removing the unnecessary features when you load the data. You can find how other people deal with it in Kernels page.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 223424,
      "author_name": "cmchen",
      "author_url": "",
      "post_date": "09/22/2017 03:40:02",
      "content": "<p>There are some tricks such as computing the specific features of a smaller set of users at one time or removing the unnecessary features when you load the data. You can find how other people deal with it in Kernels page.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "223356": "Its my first time that processing so big dataset. How much memory do i need for the competition? 16GB? 32GB? Is there any way to solve this memeoy problem ?",
    "223424": "There are some tricks such as computing the specific features of a smaller set of users at one time or removing the unnecessary features when you load the data. You can find how other people deal with it in Kernels page."
  },
  "source": "meta"
}