{
  "id": 244372,
  "title": "Why is there sub folders in the dataset?",
  "url": "/competitions/seti-breakthrough-listen/discussion/244372",
  "author_name": "Coin",
  "post_date": "2021-06-06T13:51:15.987000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Does the 16 sub folders of the dataset represents 16 different categories, or should we take them differently?</p>",
  "messages": [
    {
      "id": 1338488,
      "postDate": "2021-06-06T13:51:15.987Z",
      "content": "<p>Does the 16 sub folders of the dataset represents 16 different categories, or should we take them differently?</p>",
      "rawMarkdown": "Does the 16 sub folders of the dataset represents 16 different categories, or should we take them differently?",
      "votes": 1
    },
    {
      "id": 1338492,
      "postDate": "2021-06-06T13:53:52.750Z",
      "content": "<p>just sub folder is 1st char of hash value, and these are just hash value it is not category</p>",
      "rawMarkdown": "just sub folder is 1st char of hash value, and these are just hash value it is not category",
      "votes": 2,
      "replies": [
        {
          "id": 1338822,
          "postDate": "2021-06-06T18:46:37.927Z",
          "content": "<p>I think the split of the signal files into multiple sub folders was done just to improve the efficiency of searching for the file corresponding to a particular id. For example, the \"Bristol-Myers Squibb - Molecular Translation\" contest had millions of small image files, split into a 3-level folder hierarchy using the first 2 characters of the id as sub folder names. In that case I thought it might simplify things to just throw all the images into a single flat folder, but it turned out to be very inefficient to access the files that way. This might depend on the particular operating system or file system your machine is running.</p>",
          "rawMarkdown": "I think the split of the signal files into multiple sub folders was done just to improve the efficiency of searching for the file corresponding to a particular id. For example, the \"Bristol-Myers Squibb - Molecular Translation\" contest had millions of small image files, split into a 3-level folder hierarchy using the first 2 characters of the id as sub folder names. In that case I thought it might simplify things to just throw all the images into a single flat folder, but it turned out to be very inefficient to access the files that way. This might depend on the particular operating system or file system your machine is running.\n",
          "votes": 4
        },
        {
          "id": 1339253,
          "postDate": "2021-06-07T06:15:57.857Z",
          "content": "<p>Thanks for replying, I can make sense now.</p>",
          "rawMarkdown": "Thanks for replying, I can make sense now."
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 1338492,
      "author_name": "assign",
      "author_url": "",
      "post_date": "2021-06-06T13:53:52.750000",
      "content": "<p>just sub folder is 1st char of hash value, and these are just hash value it is not category</p>",
      "votes": 2,
      "replies": [
        {
          "id": 1338822,
          "author_name": "David J. Slate",
          "author_url": "",
          "post_date": "2021-06-06T18:46:37.927000",
          "content": "<p>I think the split of the signal files into multiple sub folders was done just to improve the efficiency of searching for the file corresponding to a particular id. For example, the \"Bristol-Myers Squibb - Molecular Translation\" contest had millions of small image files, split into a 3-level folder hierarchy using the first 2 characters of the id as sub folder names. In that case I thought it might simplify things to just throw all the images into a single flat folder, but it turned out to be very inefficient to access the files that way. This might depend on the particular operating system or file system your machine is running.</p>",
          "votes": 4,
          "replies": []
        },
        {
          "id": 1339253,
          "author_name": "Coin",
          "author_url": "",
          "post_date": "2021-06-07T06:15:57.857000",
          "content": "<p>Thanks for replying, I can make sense now.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "1338488": "Does the 16 sub folders of the dataset represents 16 different categories, or should we take them differently?",
    "1338492": "just sub folder is 1st char of hash value, and these are just hash value it is not category"
  }
}