{
  "id": 356653,
  "title": "Question about large data",
  "url": "/competitions/tabular-playground-series-oct-2022/discussion/356653",
  "author_name": "",
  "post_date": "2022-10-01T11:28:43.556180200Z",
  "votes": 4,
  "comment_count": 3,
  "views": 0,
  "content": "<p>Are we supposed to use all training sets (0 to 9) to develop our model?</p>\n<p>Thanks you</p>",
  "messages": [
    {
      "id": "1965465",
      "postDate": "10/01/2022 11:28:43",
      "content": "<p>Are we supposed to use all training sets (0 to 9) to develop our model?</p>\n<p>Thanks you</p>",
      "rawMarkdown": "Are we supposed to use all training sets (0 to 9) to develop our model?\n\nThanks you",
      "votes": null
    },
    {
      "id": "1965667",
      "postDate": "10/01/2022 13:25:55",
      "content": "<p>Yes, I think the entire training set (10 tables) need to be used for the model. Hope it helps <a href=\"https://www.kaggle.com/pythonlove371\" target=\"_blank\">@pythonlove371</a> </p>",
      "rawMarkdown": "Yes, I think the entire training set (10 tables) need to be used for the model. Hope it helps @pythonlove371",
      "votes": null
    },
    {
      "id": "1965713",
      "postDate": "10/01/2022 13:38:42",
      "content": "<p>You can definitely get started on just a single training set and try to make improvements there (e.g., feature engineering, etc.), and then increase the number of files you use for training later.</p>",
      "rawMarkdown": "You can definitely get started on just a single training set and try to make improvements there (e.g., feature engineering, etc.), and then increase the number of files you use for training later.",
      "votes": null
    },
    {
      "id": "1967595",
      "postDate": "10/02/2022 15:38:16",
      "content": "<p>I have created a dataset in parquet format, which is smaller size and requires 5 seconds loading time <a href=\"https://www.kaggle.com/datasets/alvinleenh/tps-rocket-league-data-float16-parquet-format\" target=\"_blank\">here</a><br>\nHope this helps to speed up the loading time.</p>",
      "rawMarkdown": "I have created a dataset in parquet format, which is smaller size and requires 5 seconds loading time [here](https://www.kaggle.com/datasets/alvinleenh/tps-rocket-league-data-float16-parquet-format)\nHope this helps to speed up the loading time.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1965667,
      "author_name": "ravi20076",
      "author_url": "",
      "post_date": "10/01/2022 13:25:55",
      "content": "<p>Yes, I think the entire training set (10 tables) need to be used for the model. Hope it helps <a href=\"https://www.kaggle.com/pythonlove371\" target=\"_blank\">@pythonlove371</a> </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1965713,
      "author_name": "inversion",
      "author_url": "",
      "post_date": "10/01/2022 13:38:42",
      "content": "<p>You can definitely get started on just a single training set and try to make improvements there (e.g., feature engineering, etc.), and then increase the number of files you use for training later.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1967595,
      "author_name": "alvinleenh",
      "author_url": "",
      "post_date": "10/02/2022 15:38:16",
      "content": "<p>I have created a dataset in parquet format, which is smaller size and requires 5 seconds loading time <a href=\"https://www.kaggle.com/datasets/alvinleenh/tps-rocket-league-data-float16-parquet-format\" target=\"_blank\">here</a><br>\nHope this helps to speed up the loading time.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1965465": "Are we supposed to use all training sets (0 to 9) to develop our model?\n\nThanks you",
    "1965667": "Yes, I think the entire training set (10 tables) need to be used for the model. Hope it helps @pythonlove371",
    "1965713": "You can definitely get started on just a single training set and try to make improvements there (e.g., feature engineering, etc.), and then increase the number of files you use for training later.",
    "1967595": "I have created a dataset in parquet format, which is smaller size and requires 5 seconds loading time [here](https://www.kaggle.com/datasets/alvinleenh/tps-rocket-league-data-float16-parquet-format)\nHope this helps to speed up the loading time."
  },
  "source": "meta"
}