{
  "id": 359835,
  "title": "Question about test data",
  "url": "/competitions/tabular-playground-series-oct-2022/discussion/359835",
  "author_name": "Sharov Anton",
  "post_date": "2022-10-13T19:28:19.567000",
  "votes": 5,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hi.<br>\nAm I right that I need to infer my results for each team based only on one row of data ? Can I assume that test data are note some group of events (some chunk of game) which is randomly shuffled? Also, can I assume that each row in test id is a result of some aggregation?</p>\n<p>So, I think that based on game fragments (have common event_id) it doesn't have sense to make temporal (team or player spent time in attack ) features because for test they are irrelevant? In test I have only one snapshot based on which I should infer, is it right?</p>\n<p>Thanks in advance.</p>",
  "messages": [
    {
      "id": 1986082,
      "postDate": "2022-10-13T19:28:19.567Z",
      "content": "<p>Hi.<br>\nAm I right that I need to infer my results for each team based only on one row of data ? Can I assume that test data are note some group of events (some chunk of game) which is randomly shuffled? Also, can I assume that each row in test id is a result of some aggregation?</p>\n<p>So, I think that based on game fragments (have common event_id) it doesn't have sense to make temporal (team or player spent time in attack ) features because for test they are irrelevant? In test I have only one snapshot based on which I should infer, is it right?</p>\n<p>Thanks in advance.</p>",
      "rawMarkdown": "Hi.\nAm I right that I need to infer my results for each team based only on one row of data ? Can I assume that test data are note some group of events (some chunk of game) which is randomly shuffled? Also, can I assume that each row in test id is a result of some aggregation?\n\nSo, I think that based on game fragments (have common event_id) it doesn't have sense to make temporal (team or player spent time in attack ) features because for test they are irrelevant? In test I have only one snapshot based on which I should infer, is it right?\n\nThanks in advance.\n\n",
      "votes": 5
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1986082": "Hi.\nAm I right that I need to infer my results for each team based only on one row of data ? Can I assume that test data are note some group of events (some chunk of game) which is randomly shuffled? Also, can I assume that each row in test id is a result of some aggregation?\n\nSo, I think that based on game fragments (have common event_id) it doesn't have sense to make temporal (team or player spent time in attack ) features because for test they are irrelevant? In test I have only one snapshot based on which I should infer, is it right?\n\nThanks in advance.\n\n"
  }
}