{
  "id": 4984,
  "title": "Domains of rows in dataset",
  "url": "/competitions/cause-effect-pairs/discussion/4984",
  "author_name": "",
  "post_date": "2013-07-01T12:30:36.053Z",
  "votes": null,
  "comment_count": 6,
  "views": 1842,
  "content": "<p>I'd like to ask about the domain each row comes from. I think it is essential to group rows from the same domain together, particularly it will help to better deal with categorical data. Any thoughts?</p>",
  "messages": [
    {
      "id": "26490",
      "postDate": "07/01/2013 12:30:36",
      "content": "<p>I'd like to ask about the domain each row comes from. I think it is essential to group rows from the same domain together, particularly it will help to better deal with categorical data. Any thoughts?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "26496",
      "postDate": "07/01/2013 13:37:50",
      "content": "<p>Something I am really interested in is how you guys deal with categorical/categorical or numerical/categorical data? Given the shape of the dataset they cannot even be encoded as dummy values(sparse array)</p>\r\n<p>How do you handle categorical data? I assume it <em>must</em> be handled in another way than numerical data</p>\r\n<p>Thanks</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "26536",
      "postDate": "07/01/2013 19:38:48",
      "content": "<p>[quote=Davit;26490]</p>\r\n<p>I'd like to ask about the domain each row comes from. I think it is essential to group rows from the same domain together, particularly it will help to better deal with categorical data. Any thoughts?</p>\r\n<p>[/quote]</p>\r\n<p>Information about which variable is numerical/categorical/binary is provided in xxx_publicinfo.csv</p>\r\n<p>Dealing with categorical variables has not been treated yet in the literature on cause-effect pairs to my knowledge, except for binary variables.&nbsp;</p>\r\n<p></p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "26544",
      "postDate": "07/01/2013 21:55:02",
      "content": "<p>Thanks for response. Besides information about which variable is numerical/categorical/binary it would be useful to know which rows come from the same domain like&nbsp;<em><strong>chemistry,&nbsp;climatology,&nbsp;ecology</strong></em>, etc. This is essential because rows\r\n from the same domain take consistent values(right?) and thus might be encoded together</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "26547",
      "postDate": "07/01/2013 22:06:32",
      "content": "<p>For the moment we are not releasing this information.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "26577",
      "postDate": "07/02/2013 07:24:27",
      "content": "<p>[quote=Isabelle;26536]except for binary variables.[/quote]</p>\r\n<p>Which paper is this?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "26594",
      "postDate": "07/02/2013 15:36:06",
      "content": "<p>[quote=mouse;26577]</p>\r\n<p>[quote=Isabelle;26536]except for binary variables.[/quote]</p>\r\n<p>Which paper is this?</p>\r\n<p>[/quote]</p>\r\n<p>http://kyb.mpg.de/fileadmin/user_upload/files/publications/attachments/AIMath2006-Sun_%5b0%5d.pdf</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 26496,
      "author_name": "davminer",
      "author_url": "",
      "post_date": "07/01/2013 13:37:50",
      "content": "<p>Something I am really interested in is how you guys deal with categorical/categorical or numerical/categorical data? Given the shape of the dataset they cannot even be encoded as dummy values(sparse array)</p>\r\n<p>How do you handle categorical data? I assume it <em>must</em> be handled in another way than numerical data</p>\r\n<p>Thanks</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 26536,
      "author_name": "iguyon",
      "author_url": "",
      "post_date": "07/01/2013 19:38:48",
      "content": "<p>[quote=Davit;26490]</p>\r\n<p>I'd like to ask about the domain each row comes from. I think it is essential to group rows from the same domain together, particularly it will help to better deal with categorical data. Any thoughts?</p>\r\n<p>[/quote]</p>\r\n<p>Information about which variable is numerical/categorical/binary is provided in xxx_publicinfo.csv</p>\r\n<p>Dealing with categorical variables has not been treated yet in the literature on cause-effect pairs to my knowledge, except for binary variables.&nbsp;</p>\r\n<p></p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 26544,
      "author_name": "davminer",
      "author_url": "",
      "post_date": "07/01/2013 21:55:02",
      "content": "<p>Thanks for response. Besides information about which variable is numerical/categorical/binary it would be useful to know which rows come from the same domain like&nbsp;<em><strong>chemistry,&nbsp;climatology,&nbsp;ecology</strong></em>, etc. This is essential because rows\r\n from the same domain take consistent values(right?) and thus might be encoded together</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 26547,
      "author_name": "iguyon",
      "author_url": "",
      "post_date": "07/01/2013 22:06:32",
      "content": "<p>For the moment we are not releasing this information.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 26577,
      "author_name": "bram278806",
      "author_url": "",
      "post_date": "07/02/2013 07:24:27",
      "content": "<p>[quote=Isabelle;26536]except for binary variables.[/quote]</p>\r\n<p>Which paper is this?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 26594,
      "author_name": "iguyon",
      "author_url": "",
      "post_date": "07/02/2013 15:36:06",
      "content": "<p>[quote=mouse;26577]</p>\r\n<p>[quote=Isabelle;26536]except for binary variables.[/quote]</p>\r\n<p>Which paper is this?</p>\r\n<p>[/quote]</p>\r\n<p>http://kyb.mpg.de/fileadmin/user_upload/files/publications/attachments/AIMath2006-Sun_%5b0%5d.pdf</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "26490": "",
    "26496": "",
    "26536": "",
    "26544": "",
    "26547": "",
    "26577": "",
    "26594": ""
  },
  "source": "meta"
}