{
  "id": 131141,
  "title": "No R ?",
  "url": "/competitions/deepfake-detection-challenge/discussion/131141",
  "author_name": "",
  "post_date": "2020-02-18T13:18:01.488381700Z",
  "votes": null,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Is this competition only for Python users ? I can't create R notebooks or scripts. Does anyone else have the same problem ?</p>",
  "messages": [
    {
      "id": "749235",
      "postDate": "02/18/2020 13:18:01",
      "content": "<p>Is this competition only for Python users ? I can't create R notebooks or scripts. Does anyone else have the same problem ?</p>",
      "rawMarkdown": "Is this competition only for Python users ? I can't create R notebooks or scripts. Does anyone else have the same problem ?",
      "votes": null
    },
    {
      "id": "749264",
      "postDate": "02/18/2020 14:03:27",
      "content": "<p>You would require huge computation resources for this competition. R would be very slow and would take way longer to run kernels unless until you can write everything using <code>Rcpp</code> package. Python appears to be better optimized for latency and throughput as compared to R and not all good libraries are supported in R. So every aspect points towards it makes sense to use python over R for this particular competition but again if you can write R in Cpp you are exceptional.</p>",
      "rawMarkdown": "You would require huge computation resources for this competition. R would be very slow and would take way longer to run kernels unless until you can write everything using `Rcpp` package. Python appears to be better optimized for latency and throughput as compared to R and not all good libraries are supported in R. So every aspect points towards it makes sense to use python over R for this particular competition but again if you can write R in Cpp you are exceptional.",
      "votes": null
    },
    {
      "id": "749299",
      "postDate": "02/18/2020 14:51:49",
      "content": "<p>I was hoping to use R to work directly on the images (extracted faces) instead of the videos.\nI would have mainly used two libraries for R, the first is <code>magick</code> for the image treatments and the second is <code>tensorflow</code> for the model. Do you think it will still exceed the kernel limits ?</p>",
      "rawMarkdown": "I was hoping to use R to work directly on the images (extracted faces) instead of the videos.\nI would have mainly used two libraries for R, the first is `magick ` for the image treatments and the second is `tensorflow ` for the model. Do you think it will still exceed the kernel limits ?",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 749264,
      "author_name": "cyberia",
      "author_url": "",
      "post_date": "02/18/2020 14:03:27",
      "content": "<p>You would require huge computation resources for this competition. R would be very slow and would take way longer to run kernels unless until you can write everything using <code>Rcpp</code> package. Python appears to be better optimized for latency and throughput as compared to R and not all good libraries are supported in R. So every aspect points towards it makes sense to use python over R for this particular competition but again if you can write R in Cpp you are exceptional.</p>",
      "votes": null,
      "replies": [
        {
          "id": 749299,
          "author_name": "mohenni",
          "author_url": "",
          "post_date": "02/18/2020 14:51:49",
          "content": "<p>I was hoping to use R to work directly on the images (extracted faces) instead of the videos.\nI would have mainly used two libraries for R, the first is <code>magick</code> for the image treatments and the second is <code>tensorflow</code> for the model. Do you think it will still exceed the kernel limits ?</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "749235": "Is this competition only for Python users ? I can't create R notebooks or scripts. Does anyone else have the same problem ?",
    "749264": "You would require huge computation resources for this competition. R would be very slow and would take way longer to run kernels unless until you can write everything using `Rcpp` package. Python appears to be better optimized for latency and throughput as compared to R and not all good libraries are supported in R. So every aspect points towards it makes sense to use python over R for this particular competition but again if you can write R in Cpp you are exceptional.",
    "749299": "I was hoping to use R to work directly on the images (extracted faces) instead of the videos.\nI would have mainly used two libraries for R, the first is `magick ` for the image treatments and the second is `tensorflow ` for the model. Do you think it will still exceed the kernel limits ?"
  },
  "source": "meta"
}