{
  "id": 478175,
  "title": "Does test_base overlaps case_id's with train_base?",
  "url": "/competitions/home-credit-credit-risk-model-stability/discussion/478175",
  "author_name": "",
  "post_date": "2024-02-19T13:28:12.369914100Z",
  "votes": 9,
  "comment_count": 4,
  "views": 0,
  "content": "<p>From data:<br>\ntest_base.csv (Note, the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv)<br>\nI don't understand what is saying? Overlapping?</p>",
  "messages": [
    {
      "id": "2658888",
      "postDate": "02/19/2024 13:28:12",
      "content": "<p>From data:<br>\ntest_base.csv (Note, the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv)<br>\nI don't understand what is saying? Overlapping?</p>",
      "rawMarkdown": "From data:\ntest_base.csv (Note, the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv)\nI don't understand what is saying? Overlapping?",
      "votes": null
    },
    {
      "id": "2659045",
      "postDate": "02/19/2024 15:09:44",
      "content": "<p>I don't get this message either. Just made a submission where left joined test with  train, then used train targets as scores — got 0 LB. It means that there is likely no overlap in <code>case_id</code> values.</p>",
      "rawMarkdown": "I don't get this message either. Just made a submission where left joined test with  train, then used train targets as scores — got 0 LB. It means that there is likely no overlap in `case_id` values.",
      "votes": null
    },
    {
      "id": "2659068",
      "postDate": "02/19/2024 15:28:37",
      "content": "<p>I believe it means that the size of the test set is 90% of the train set.</p>",
      "rawMarkdown": "I believe it means that the size of the test set is 90% of the train set.",
      "votes": null
    },
    {
      "id": "2659183",
      "postDate": "02/19/2024 16:47:17",
      "content": "<p>Yes, I think so, but is a very hard way of say it.</p>",
      "rawMarkdown": "Yes, I think so, but is a very hard way of say it.",
      "votes": null
    },
    {
      "id": "2659315",
      "postDate": "02/19/2024 18:43:42",
      "content": "<p>Yeah, I agree. I think they wanted to emphasize that the number of case ids constitutes 90% of the total number of training case ids, but other data files may be larger or smaller in comparison (for instance, it appears that there is more data in applprev in the test dataset).</p>",
      "rawMarkdown": "Yeah, I agree. I think they wanted to emphasize that the number of case ids constitutes 90% of the total number of training case ids, but other data files may be larger or smaller in comparison (for instance, it appears that there is more data in applprev in the test dataset).",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2659045,
      "author_name": "kononenko",
      "author_url": "",
      "post_date": "02/19/2024 15:09:44",
      "content": "<p>I don't get this message either. Just made a submission where left joined test with  train, then used train targets as scores — got 0 LB. It means that there is likely no overlap in <code>case_id</code> values.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 2659068,
      "author_name": "eivolkova",
      "author_url": "",
      "post_date": "02/19/2024 15:28:37",
      "content": "<p>I believe it means that the size of the test set is 90% of the train set.</p>",
      "votes": null,
      "replies": [
        {
          "id": 2659183,
          "author_name": "blindape",
          "author_url": "",
          "post_date": "02/19/2024 16:47:17",
          "content": "<p>Yes, I think so, but is a very hard way of say it.</p>",
          "votes": null,
          "replies": [
            {
              "id": 2659315,
              "author_name": "eivolkova",
              "author_url": "",
              "post_date": "02/19/2024 18:43:42",
              "content": "<p>Yeah, I agree. I think they wanted to emphasize that the number of case ids constitutes 90% of the total number of training case ids, but other data files may be larger or smaller in comparison (for instance, it appears that there is more data in applprev in the test dataset).</p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2658888": "From data:\ntest_base.csv (Note, the hidden test_base.csv contains approximately 90% of the numbers of case_id values of train_base.csv)\nI don't understand what is saying? Overlapping?",
    "2659045": "I don't get this message either. Just made a submission where left joined test with  train, then used train targets as scores — got 0 LB. It means that there is likely no overlap in `case_id` values.",
    "2659068": "I believe it means that the size of the test set is 90% of the train set.",
    "2659183": "Yes, I think so, but is a very hard way of say it.",
    "2659315": "Yeah, I agree. I think they wanted to emphasize that the number of case ids constitutes 90% of the total number of training case ids, but other data files may be larger or smaller in comparison (for instance, it appears that there is more data in applprev in the test dataset)."
  },
  "source": "meta"
}