{
  "id": 298250,
  "title": "Faster dataframes with Polars instead of Pandas",
  "url": "/competitions/g-research-crypto-forecasting/discussion/298250",
  "author_name": "Roland Luethy",
  "post_date": "2022-01-01T17:25:59.649000",
  "votes": 2,
  "comment_count": 5,
  "views": 0,
  "content": "<p>Polars is a relatively new library for handling dataframes. It can be significantly faster than Pandas according to this benchmark: <a href=\"http://https//h2oai.github.io/db-benchmark/\" target=\"_blank\">http://https//h2oai.github.io/db-benchmark/</a></p>\n<p>To try Polars, I created a submission using Polars instead of Pandas: </p>\n<p><a href=\"https://www.kaggle.com/rluethy/crypto-prediction-xgb-using-polars\" target=\"_blank\">https://www.kaggle.com/rluethy/crypto-prediction-xgb-using-polars</a></p>\n<p>And the Kaggle dataset with the python wheel to install the library offline:</p>\n<p><a href=\"https://www.kaggle.com/rluethy/polars-fast-dataframe-library\" target=\"_blank\">https://www.kaggle.com/rluethy/polars-fast-dataframe-library</a></p>",
  "messages": [
    {
      "id": 1635525,
      "postDate": "2022-01-01T20:52:57.730Z",
      "content": "<p>Thank you for sharing. I believe Rust has a great potential to be used in DS/ML/AI field, it's wonderful to see Rust-based libraries gaining momentum in the community. </p>",
      "rawMarkdown": "Thank you for sharing. I believe Rust has a great potential to be used in DS/ML/AI field, it's wonderful to see Rust-based libraries gaining momentum in the community. ",
      "votes": 1,
      "replies": [
        {
          "id": 1636190,
          "postDate": "2022-01-02T17:09:16.873Z",
          "content": "<p><strong>Hi, <a href=\"https://www.kaggle.com/vlpavlov\" target=\"_blank\">@vlpavlov</a>  can you share more informations or articles  about Rust-Based Libraries</strong>.</p>",
          "rawMarkdown": "**Hi, @vlpavlov  can you share more informations or articles  about Rust-Based Libraries**."
        },
        {
          "id": 1636414,
          "postDate": "2022-01-02T21:05:56.333Z",
          "content": "<p>Well, Polars is written in Rust :)</p>",
          "rawMarkdown": "Well, Polars is written in Rust :)"
        }
      ]
    },
    {
      "id": 1635480,
      "postDate": "2022-01-01T19:40:02.853Z",
      "content": "<h2>Hi, <a href=\"https://www.kaggle.com/rluethy\" target=\"_blank\">@rluethy</a> , Thank your for sharing details about Polars library.</h2>\n<h2>As my knowledge  Datatable  library also faster and better dataframes as compare to Panda Library.</h2>\n<h2>please check this notebook <a href=\"https://www.kaggle.com/tunguz/eda-with-python-datatable\" target=\"_blank\">click here</a></h2>",
      "rawMarkdown": "## Hi, @rluethy , Thank your for sharing details about Polars library.  \n## As my knowledge  Datatable  library also faster and better dataframes as compare to Panda Library.\n## please check this notebook [click here](https://www.kaggle.com/tunguz/eda-with-python-datatable)",
      "votes": 1
    },
    {
      "id": 1635347,
      "postDate": "2022-01-01T17:25:59.650Z",
      "content": "<p>Polars is a relatively new library for handling dataframes. It can be significantly faster than Pandas according to this benchmark: <a href=\"http://https//h2oai.github.io/db-benchmark/\" target=\"_blank\">http://https//h2oai.github.io/db-benchmark/</a></p>\n<p>To try Polars, I created a submission using Polars instead of Pandas: </p>\n<p><a href=\"https://www.kaggle.com/rluethy/crypto-prediction-xgb-using-polars\" target=\"_blank\">https://www.kaggle.com/rluethy/crypto-prediction-xgb-using-polars</a></p>\n<p>And the Kaggle dataset with the python wheel to install the library offline:</p>\n<p><a href=\"https://www.kaggle.com/rluethy/polars-fast-dataframe-library\" target=\"_blank\">https://www.kaggle.com/rluethy/polars-fast-dataframe-library</a></p>",
      "rawMarkdown": "Polars is a relatively new library for handling dataframes. It can be significantly faster than Pandas according to this benchmark: http://https//h2oai.github.io/db-benchmark/\n\nTo try Polars, I created a submission using Polars instead of Pandas: \n\nhttps://www.kaggle.com/rluethy/crypto-prediction-xgb-using-polars\n\nAnd the Kaggle dataset with the python wheel to install the library offline:\n\nhttps://www.kaggle.com/rluethy/polars-fast-dataframe-library\n\n",
      "votes": 2
    },
    {
      "id": 1635550,
      "postDate": "2022-01-01T21:44:35.707Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 1635525,
      "author_name": "Vlad Pavlov",
      "author_url": "",
      "post_date": "2022-01-01T20:52:57.730000",
      "content": "<p>Thank you for sharing. I believe Rust has a great potential to be used in DS/ML/AI field, it's wonderful to see Rust-based libraries gaining momentum in the community. </p>",
      "votes": 1,
      "replies": [
        {
          "id": 1636190,
          "author_name": "Muhammad Ismail",
          "author_url": "",
          "post_date": "2022-01-02T17:09:16.873000",
          "content": "<p><strong>Hi, <a href=\"https://www.kaggle.com/vlpavlov\" target=\"_blank\">@vlpavlov</a>  can you share more informations or articles  about Rust-Based Libraries</strong>.</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1636414,
          "author_name": "Vlad Pavlov",
          "author_url": "",
          "post_date": "2022-01-02T21:05:56.333000",
          "content": "<p>Well, Polars is written in Rust :)</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1635480,
      "author_name": "Muhammad Ismail",
      "author_url": "",
      "post_date": "2022-01-01T19:40:02.853000",
      "content": "<h2>Hi, <a href=\"https://www.kaggle.com/rluethy\" target=\"_blank\">@rluethy</a> , Thank your for sharing details about Polars library.</h2>\n<h2>As my knowledge  Datatable  library also faster and better dataframes as compare to Panda Library.</h2>\n<h2>please check this notebook <a href=\"https://www.kaggle.com/tunguz/eda-with-python-datatable\" target=\"_blank\">click here</a></h2>",
      "votes": 1,
      "replies": []
    },
    {
      "id": 1635550,
      "author_name": "",
      "author_url": "",
      "post_date": "2022-01-01T21:44:35.707000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1635525": "Thank you for sharing. I believe Rust has a great potential to be used in DS/ML/AI field, it's wonderful to see Rust-based libraries gaining momentum in the community. ",
    "1635480": "## Hi, @rluethy , Thank your for sharing details about Polars library.  \n## As my knowledge  Datatable  library also faster and better dataframes as compare to Panda Library.\n## please check this notebook [click here](https://www.kaggle.com/tunguz/eda-with-python-datatable)",
    "1635347": "Polars is a relatively new library for handling dataframes. It can be significantly faster than Pandas according to this benchmark: http://https//h2oai.github.io/db-benchmark/\n\nTo try Polars, I created a submission using Polars instead of Pandas: \n\nhttps://www.kaggle.com/rluethy/crypto-prediction-xgb-using-polars\n\nAnd the Kaggle dataset with the python wheel to install the library offline:\n\nhttps://www.kaggle.com/rluethy/polars-fast-dataframe-library\n\n",
    "1635550": ""
  }
}