{
  "id": 137356,
  "title": "Disappointment",
  "url": "/competitions/bengaliai-cv19/discussion/137356",
  "author_name": "",
  "post_date": "2020-03-20T10:29:28.034161400Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>First of all, congratulations to the winners and all the participants.</p>\n\n<p>I would like to show my disappointment with this competition and the organizers. The choice of parquet files instead another format limited the participation only to Python users. For who code in R, this format wasn't available.</p>\n\n<p>I asked for help and opened an issue in the Kaggle github account to update the arrow package which gives functionalities to manage these files, but I never received support for this inconvenient. I felt motivated with this challenge and expected to submit a solution, but I only could publish a spin-off notebook (look <a href=\"https://www.kaggle.com/jmartindelasierra/class-activation-mapping-in-handwritten-bengali\">here</a> and upvote if you find it useful).</p>\n\n<p>Please, next time provide more standard data formats. If they aren't, please be sure that there will be enough support for the different programmer communities.</p>\n\n<p>Thank you.\nJesus</p>",
  "messages": [
    {
      "id": "780506",
      "postDate": "03/20/2020 10:29:28",
      "content": "<p>First of all, congratulations to the winners and all the participants.</p>\n\n<p>I would like to show my disappointment with this competition and the organizers. The choice of parquet files instead another format limited the participation only to Python users. For who code in R, this format wasn't available.</p>\n\n<p>I asked for help and opened an issue in the Kaggle github account to update the arrow package which gives functionalities to manage these files, but I never received support for this inconvenient. I felt motivated with this challenge and expected to submit a solution, but I only could publish a spin-off notebook (look <a href=\"https://www.kaggle.com/jmartindelasierra/class-activation-mapping-in-handwritten-bengali\">here</a> and upvote if you find it useful).</p>\n\n<p>Please, next time provide more standard data formats. If they aren't, please be sure that there will be enough support for the different programmer communities.</p>\n\n<p>Thank you.\nJesus</p>",
      "rawMarkdown": "First of all, congratulations to the winners and all the participants.\n\nI would like to show my disappointment with this competition and the organizers. The choice of parquet files instead another format limited the participation only to Python users. For who code in R, this format wasn't available.\n\nI asked for help and opened an issue in the Kaggle github account to update the arrow package which gives functionalities to manage these files, but I never received support for this inconvenient. I felt motivated with this challenge and expected to submit a solution, but I only could publish a spin-off notebook (look [here](https://www.kaggle.com/jmartindelasierra/class-activation-mapping-in-handwritten-bengali) and upvote if you find it useful).\n\nPlease, next time provide more standard data formats. If they aren't, please be sure that there will be enough support for the different programmer communities.\n\nThank you.\nJesus",
      "votes": null
    },
    {
      "id": "781247",
      "postDate": "03/21/2020 04:30:38",
      "content": "<p>Wasn't there enough public kernels to show how to get it working with parquet files?</p>",
      "rawMarkdown": "Wasn't there enough public kernels to show how to get it working with parquet files?",
      "votes": null
    },
    {
      "id": "821499",
      "postDate": "04/26/2020 08:00:10",
      "content": "<p><a href=\"/jmartindelasierra\">@jmartindelasierra</a> In many computer vision problems, I think python is a better choice than R. But I agree about the data format, general image format like <code>.png</code> or <code>.jpeg</code> would enough and better, I think. Because it is just a simple handwritten image, using <code>parquet</code> is a bit fancy I think.   </p>",
      "rawMarkdown": "jmartindelasierra In many computer vision problems, I think python is a better choice than R. But I agree about the data format, general image format like `.png` or `.jpeg` would enough and better, I think. Because it is just a simple handwritten image, using `parquet` is a bit fancy I think.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 781247,
      "author_name": "kurianbenoy",
      "author_url": "",
      "post_date": "03/21/2020 04:30:38",
      "content": "<p>Wasn't there enough public kernels to show how to get it working with parquet files?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 821499,
      "author_name": "ipythonx",
      "author_url": "",
      "post_date": "04/26/2020 08:00:10",
      "content": "<p><a href=\"/jmartindelasierra\">@jmartindelasierra</a> In many computer vision problems, I think python is a better choice than R. But I agree about the data format, general image format like <code>.png</code> or <code>.jpeg</code> would enough and better, I think. Because it is just a simple handwritten image, using <code>parquet</code> is a bit fancy I think.   </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "780506": "First of all, congratulations to the winners and all the participants.\n\nI would like to show my disappointment with this competition and the organizers. The choice of parquet files instead another format limited the participation only to Python users. For who code in R, this format wasn't available.\n\nI asked for help and opened an issue in the Kaggle github account to update the arrow package which gives functionalities to manage these files, but I never received support for this inconvenient. I felt motivated with this challenge and expected to submit a solution, but I only could publish a spin-off notebook (look [here](https://www.kaggle.com/jmartindelasierra/class-activation-mapping-in-handwritten-bengali) and upvote if you find it useful).\n\nPlease, next time provide more standard data formats. If they aren't, please be sure that there will be enough support for the different programmer communities.\n\nThank you.\nJesus",
    "781247": "Wasn't there enough public kernels to show how to get it working with parquet files?",
    "821499": "jmartindelasierra In many computer vision problems, I think python is a better choice than R. But I agree about the data format, general image format like `.png` or `.jpeg` would enough and better, I think. Because it is just a simple handwritten image, using `parquet` is a bit fancy I think."
  },
  "source": "meta"
}