{
  "id": 479338,
  "title": "Problem with dataset",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/479338",
  "author_name": "",
  "post_date": "2024-02-24T06:45:34.578394600Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>That's the first time I am working with big data. I couldn't get the point of whether I have to use all datasets from the train folder or not.</p>",
  "messages": [
    {
      "id": "2666116",
      "postDate": "02/24/2024 06:45:34",
      "content": "<p>That's the first time I am working with big data. I couldn't get the point of whether I have to use all datasets from the train folder or not.</p>",
      "rawMarkdown": "That's the first time I am working with big data. I couldn't get the point of whether I have to use all datasets from the train folder or not.",
      "votes": null
    },
    {
      "id": "2666382",
      "postDate": "02/24/2024 11:08:24",
      "content": "<p>I suggest that you start by depth 0 tables and have a working baseline then gradually add depth 1 and depth 2 tables </p>",
      "rawMarkdown": "I suggest that you start by depth 0 tables and have a working baseline then gradually add depth 1 and depth 2 tables",
      "votes": null
    },
    {
      "id": "2674440",
      "postDate": "02/29/2024 10:28:08",
      "content": "<p>To add something: this data is much larger than what Kaggle, and probably your laptop, has in terms of RAM. Using Pandas might be difficult, so it's better to try Polars. </p>",
      "rawMarkdown": "To add something: this data is much larger than what Kaggle, and probably your laptop, has in terms of RAM. Using Pandas might be difficult, so it's better to try Polars.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2666382,
      "author_name": "fadylabib",
      "author_url": "",
      "post_date": "02/24/2024 11:08:24",
      "content": "<p>I suggest that you start by depth 0 tables and have a working baseline then gradually add depth 1 and depth 2 tables </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2674440,
      "author_name": "marekgp",
      "author_url": "",
      "post_date": "02/29/2024 10:28:08",
      "content": "<p>To add something: this data is much larger than what Kaggle, and probably your laptop, has in terms of RAM. Using Pandas might be difficult, so it's better to try Polars. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2666116": "That's the first time I am working with big data. I couldn't get the point of whether I have to use all datasets from the train folder or not.",
    "2666382": "I suggest that you start by depth 0 tables and have a working baseline then gradually add depth 1 and depth 2 tables",
    "2674440": "To add something: this data is much larger than what Kaggle, and probably your laptop, has in terms of RAM. Using Pandas might be difficult, so it's better to try Polars."
  },
  "source": "meta"
}