{
  "id": 139975,
  "title": "Are day and month cyclical?",
  "url": "/competitions/cat-in-the-dat-ii/discussion/139975",
  "author_name": "",
  "post_date": "2020-03-30T20:40:01.159872900Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>I know it's late, but the official description states that it contains \"(potentially) cyclical features\". I tried encoding day and month as cyclic features, but each of their correlation went down. Does this mean day and month are not actually cyclical?</p>",
  "messages": [
    {
      "id": "792000",
      "postDate": "03/30/2020 20:40:01",
      "content": "<p>I know it's late, but the official description states that it contains \"(potentially) cyclical features\". I tried encoding day and month as cyclic features, but each of their correlation went down. Does this mean day and month are not actually cyclical?</p>",
      "rawMarkdown": "I know it's late, but the official description states that it contains \"(potentially) cyclical features\". I tried encoding day and month as cyclic features, but each of their correlation went down. Does this mean day and month are not actually cyclical?",
      "votes": null
    },
    {
      "id": "796602",
      "postDate": "04/03/2020 18:18:35",
      "content": "<p>Yes, they are cyclical.</p>\n\n<p>Although the predictors day and month are cyclical, they are provided by Kaggle encoded as ordinal features.\nMonday is the day after Sunday. So, in a sense, Sunday and Monday are closer than Sunday and Wednesday. Nevertheless, as the day variable is numbered from 1 to 7 (to represent the days of the week), Monday equals 1, Wednesday 3 and Sunday 7. In this way, Sunday and Wednesday are closer than Sunday and Monday. The month variable, taking values going from 1 to 12 to number the months of the year, has the same peculiarity. </p>\n\n<p>I explain this in more details in the notebook I have produced on this competition. It is available <a href=\"https://www.kaggle.com/robinmunier/categorical-feature-encoding-challenge-ii/notebook\">here</a></p>\n\n<p>Hope it helps</p>",
      "rawMarkdown": "Yes, they are cyclical.\n\nAlthough the predictors day and month are cyclical, they are provided by Kaggle encoded as ordinal features.\nMonday is the day after Sunday. So, in a sense, Sunday and Monday are closer than Sunday and Wednesday. Nevertheless, as the day variable is numbered from 1 to 7 (to represent the days of the week), Monday equals 1, Wednesday 3 and Sunday 7. In this way, Sunday and Wednesday are closer than Sunday and Monday. The month variable, taking values going from 1 to 12 to number the months of the year, has the same peculiarity. \n\nI explain this in more details in the notebook I have produced on this competition. It is available [here](https://www.kaggle.com/robinmunier/categorical-feature-encoding-challenge-ii/notebook)\n\nHope it helps",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 796602,
      "author_name": "robinmunier",
      "author_url": "",
      "post_date": "04/03/2020 18:18:35",
      "content": "<p>Yes, they are cyclical.</p>\n\n<p>Although the predictors day and month are cyclical, they are provided by Kaggle encoded as ordinal features.\nMonday is the day after Sunday. So, in a sense, Sunday and Monday are closer than Sunday and Wednesday. Nevertheless, as the day variable is numbered from 1 to 7 (to represent the days of the week), Monday equals 1, Wednesday 3 and Sunday 7. In this way, Sunday and Wednesday are closer than Sunday and Monday. The month variable, taking values going from 1 to 12 to number the months of the year, has the same peculiarity. </p>\n\n<p>I explain this in more details in the notebook I have produced on this competition. It is available <a href=\"https://www.kaggle.com/robinmunier/categorical-feature-encoding-challenge-ii/notebook\">here</a></p>\n\n<p>Hope it helps</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "792000": "I know it's late, but the official description states that it contains \"(potentially) cyclical features\". I tried encoding day and month as cyclic features, but each of their correlation went down. Does this mean day and month are not actually cyclical?",
    "796602": "Yes, they are cyclical.\n\nAlthough the predictors day and month are cyclical, they are provided by Kaggle encoded as ordinal features.\nMonday is the day after Sunday. So, in a sense, Sunday and Monday are closer than Sunday and Wednesday. Nevertheless, as the day variable is numbered from 1 to 7 (to represent the days of the week), Monday equals 1, Wednesday 3 and Sunday 7. In this way, Sunday and Wednesday are closer than Sunday and Monday. The month variable, taking values going from 1 to 12 to number the months of the year, has the same peculiarity. \n\nI explain this in more details in the notebook I have produced on this competition. It is available [here](https://www.kaggle.com/robinmunier/categorical-feature-encoding-challenge-ii/notebook)\n\nHope it helps"
  },
  "source": "meta"
}