{
  "id": 295191,
  "title": "Code Tab and feature engineering",
  "url": "/competitions/g-research-crypto-forecasting/discussion/295191",
  "author_name": "",
  "post_date": "2021-12-14T20:39:04.926272700Z",
  "votes": null,
  "comment_count": 1,
  "views": 0,
  "content": "<p>Has the code portion of this tab been discontinued? </p>\n<p>I am curious if anyone has come across more intuitive feature engineering than what I have seen passed around. </p>",
  "messages": [
    {
      "id": "1618278",
      "postDate": "12/14/2021 20:39:04",
      "content": "<p>Has the code portion of this tab been discontinued? </p>\n<p>I am curious if anyone has come across more intuitive feature engineering than what I have seen passed around. </p>",
      "rawMarkdown": "Has the code portion of this tab been discontinued? \n\nI am curious if anyone has come across more intuitive feature engineering than what I have seen passed around.",
      "votes": null
    },
    {
      "id": "1618701",
      "postDate": "12/15/2021 08:37:57",
      "content": "<p>Given the time serie API, I am affraid there is no 'easy' way to proceed. From experience in a previous competition I have built a general framework for feature engineering: <a href=\"https://www.kaggle.com/lucasmorin/on-line-feature-engineering\" target=\"_blank\">https://www.kaggle.com/lucasmorin/on-line-feature-engineering</a>. I am not sure it would qualify as 'simple' but I suspect it is not too far from 'minimal'.</p>",
      "rawMarkdown": "Given the time serie API, I am affraid there is no 'easy' way to proceed. From experience in a previous competition I have built a general framework for feature engineering: https://www.kaggle.com/lucasmorin/on-line-feature-engineering. I am not sure it would qualify as 'simple' but I suspect it is not too far from 'minimal'.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1618701,
      "author_name": "lucasmorin",
      "author_url": "",
      "post_date": "12/15/2021 08:37:57",
      "content": "<p>Given the time serie API, I am affraid there is no 'easy' way to proceed. From experience in a previous competition I have built a general framework for feature engineering: <a href=\"https://www.kaggle.com/lucasmorin/on-line-feature-engineering\" target=\"_blank\">https://www.kaggle.com/lucasmorin/on-line-feature-engineering</a>. I am not sure it would qualify as 'simple' but I suspect it is not too far from 'minimal'.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1618278": "Has the code portion of this tab been discontinued? \n\nI am curious if anyone has come across more intuitive feature engineering than what I have seen passed around.",
    "1618701": "Given the time serie API, I am affraid there is no 'easy' way to proceed. From experience in a previous competition I have built a general framework for feature engineering: https://www.kaggle.com/lucasmorin/on-line-feature-engineering. I am not sure it would qualify as 'simple' but I suspect it is not too far from 'minimal'."
  },
  "source": "meta"
}