{
  "id": 65800,
  "title": "is this happen to any one?",
  "url": "/competitions/ga-customer-revenue-prediction/discussion/65800",
  "author_name": "",
  "post_date": "2018-09-15T04:20:08.272033100Z",
  "votes": 2,
  "comment_count": 4,
  "views": 0,
  "content": "<p>When I am using <strong>GPU</strong> run <a href=\"https://www.kaggle.com/sudalairajkumar/simple-exploration-baseline-ga-customer-revenue\">SRK Kernel</a> It gives me dimension features <strong>25</strong> and when I used <strong>CPU</strong> it will return <strong>55</strong> why this happens anyone can please give me the answer ?</p>",
  "messages": [
    {
      "id": "387496",
      "postDate": "09/15/2018 04:20:08",
      "content": "<p>When I am using <strong>GPU</strong> run <a href=\"https://www.kaggle.com/sudalairajkumar/simple-exploration-baseline-ga-customer-revenue\">SRK Kernel</a> It gives me dimension features <strong>25</strong> and when I used <strong>CPU</strong> it will return <strong>55</strong> why this happens anyone can please give me the answer ?</p>",
      "rawMarkdown": "When I am using **GPU** run [SRK Kernel][1] It gives me dimension features **25** and when I used **CPU** it will return **55** why this happens anyone can please give me the answer ?\n\n\n  [1]: https://www.kaggle.com/sudalairajkumar/simple-exploration-baseline-ga-customer-revenue",
      "votes": null
    },
    {
      "id": "387825",
      "postDate": "09/15/2018 18:12:50",
      "content": "<p>This is very unique issue Ashish. You should also post it on the Kaggle forum with screenshots also tag someone like SRK or Giba coz they can help you in figuring out the solution or the reason for same.</p>",
      "rawMarkdown": "This is very unique issue Ashish. You should also post it on the Kaggle forum with screenshots also tag someone like SRK or Giba coz they can help you in figuring out the solution or the reason for same.",
      "votes": null
    },
    {
      "id": "387895",
      "postDate": "09/15/2018 22:04:53",
      "content": "<p>I believe this is because some of the columns are actually json objects that contain multiple \"columns.\" When you flatten all the JSON objects, the full dataset has 55 columns.</p>",
      "rawMarkdown": "I believe this is because some of the columns are actually json objects that contain multiple \"columns.\" When you flatten all the JSON objects, the full dataset has 55 columns.",
      "votes": null
    },
    {
      "id": "388185",
      "postDate": "09/16/2018 14:00:57",
      "content": "<p>@Ashish: Do you mind providing the snapshot of the 25 features you are getting?</p>",
      "rawMarkdown": "Ashish: Do you mind providing the snapshot of the 25 features you are getting?",
      "votes": null
    },
    {
      "id": "395182",
      "postDate": "09/28/2018 05:16:11",
      "content": "<p>I was also having a similar issue. In pandas documentation, it says, data should be \"dict or list of dicts\"\nSo I converted the line to: column_as_df = json_normalize(df[column]<strong>.tolist()</strong>)\nAnd it worked.</p>",
      "rawMarkdown": "I was also having a similar issue. In pandas documentation, it says, data should be \"dict or list of dicts\"\nSo I converted the line to: column_as_df = json_normalize(df[column]**.tolist()**)\nAnd it worked.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 387825,
      "author_name": "ishaan45",
      "author_url": "",
      "post_date": "09/15/2018 18:12:50",
      "content": "<p>This is very unique issue Ashish. You should also post it on the Kaggle forum with screenshots also tag someone like SRK or Giba coz they can help you in figuring out the solution or the reason for same.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 387895,
      "author_name": "dwiwad",
      "author_url": "",
      "post_date": "09/15/2018 22:04:53",
      "content": "<p>I believe this is because some of the columns are actually json objects that contain multiple \"columns.\" When you flatten all the JSON objects, the full dataset has 55 columns.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 388185,
      "author_name": "pratikkgandhi",
      "author_url": "",
      "post_date": "09/16/2018 14:00:57",
      "content": "<p>@Ashish: Do you mind providing the snapshot of the 25 features you are getting?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 395182,
      "author_name": "ug2409",
      "author_url": "",
      "post_date": "09/28/2018 05:16:11",
      "content": "<p>I was also having a similar issue. In pandas documentation, it says, data should be \"dict or list of dicts\"\nSo I converted the line to: column_as_df = json_normalize(df[column]<strong>.tolist()</strong>)\nAnd it worked.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "387496": "When I am using **GPU** run [SRK Kernel][1] It gives me dimension features **25** and when I used **CPU** it will return **55** why this happens anyone can please give me the answer ?\n\n\n  [1]: https://www.kaggle.com/sudalairajkumar/simple-exploration-baseline-ga-customer-revenue",
    "387825": "This is very unique issue Ashish. You should also post it on the Kaggle forum with screenshots also tag someone like SRK or Giba coz they can help you in figuring out the solution or the reason for same.",
    "387895": "I believe this is because some of the columns are actually json objects that contain multiple \"columns.\" When you flatten all the JSON objects, the full dataset has 55 columns.",
    "388185": "Ashish: Do you mind providing the snapshot of the 25 features you are getting?",
    "395182": "I was also having a similar issue. In pandas documentation, it says, data should be \"dict or list of dicts\"\nSo I converted the line to: column_as_df = json_normalize(df[column]**.tolist()**)\nAnd it worked."
  },
  "source": "meta"
}