{
  "id": 131535,
  "title": "How to change the dataset from .parquet to .feather type?",
  "url": "/competitions/bengaliai-cv19/discussion/131535",
  "author_name": "",
  "post_date": "2020-02-20T09:08:27.032626700Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Hi!</p>\n\n<p>I saw multiple people working on changing the dataset to a faster .feather type in order to shorten load time.\nI download the zip but I only have the .parquet files. Could you help me how to get .feather type files?</p>",
  "messages": [
    {
      "id": "751541",
      "postDate": "02/20/2020 09:08:27",
      "content": "<p>Hi!</p>\n\n<p>I saw multiple people working on changing the dataset to a faster .feather type in order to shorten load time.\nI download the zip but I only have the .parquet files. Could you help me how to get .feather type files?</p>",
      "rawMarkdown": "Hi!\n\nI saw multiple people working on changing the dataset to a faster .feather type in order to shorten load time.\nI download the zip but I only have the .parquet files. Could you help me how to get .feather type files?",
      "votes": null
    },
    {
      "id": "751616",
      "postDate": "02/20/2020 10:41:18",
      "content": "<p>You can have a script which reads the parquet files and saves them as feather files. Like the following\n<code>data = pd.read_parquet(filepath)</code>\n<code>data.to_feather(new_filepath.ftr)</code></p>\n\n<p>Conversion is just a one time thing.</p>",
      "rawMarkdown": "You can have a script which reads the parquet files and saves them as feather files. Like the following\n`data = pd.read_parquet(filepath)`\n`data.to_feather(new_filepath.ftr)`\n\nConversion is just a one time thing.",
      "votes": null
    },
    {
      "id": "751665",
      "postDate": "02/20/2020 11:37:14",
      "content": "<p>I get it.Thank you!</p>",
      "rawMarkdown": "I get it.Thank you!",
      "votes": null
    },
    {
      "id": "752094",
      "postDate": "02/20/2020 18:37:28",
      "content": "<p>But using .feather won't give you speed up during test prediction right? Then, how is it beneficial?</p>",
      "rawMarkdown": "But using .feather won't give you speed up during test prediction right? Then, how is it beneficial?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 751616,
      "author_name": "santosh16k",
      "author_url": "",
      "post_date": "02/20/2020 10:41:18",
      "content": "<p>You can have a script which reads the parquet files and saves them as feather files. Like the following\n<code>data = pd.read_parquet(filepath)</code>\n<code>data.to_feather(new_filepath.ftr)</code></p>\n\n<p>Conversion is just a one time thing.</p>",
      "votes": null,
      "replies": [
        {
          "id": 751665,
          "author_name": "whutddmm",
          "author_url": "",
          "post_date": "02/20/2020 11:37:14",
          "content": "<p>I get it.Thank you!</p>",
          "votes": null,
          "replies": []
        }
      ]
    },
    {
      "id": 752094,
      "author_name": "venky2506",
      "author_url": "",
      "post_date": "02/20/2020 18:37:28",
      "content": "<p>But using .feather won't give you speed up during test prediction right? Then, how is it beneficial?</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "751541": "Hi!\n\nI saw multiple people working on changing the dataset to a faster .feather type in order to shorten load time.\nI download the zip but I only have the .parquet files. Could you help me how to get .feather type files?",
    "751616": "You can have a script which reads the parquet files and saves them as feather files. Like the following\n`data = pd.read_parquet(filepath)`\n`data.to_feather(new_filepath.ftr)`\n\nConversion is just a one time thing.",
    "751665": "I get it.Thank you!",
    "752094": "But using .feather won't give you speed up during test prediction right? Then, how is it beneficial?"
  },
  "source": "meta"
}