{
  "id": 588006,
  "title": "Data cleaning: arrival > departure",
  "url": "/competitions/aeroclub-recsys-2025/discussion/588006",
  "author_name": "",
  "post_date": "2025-07-03T20:21:58.092813400Z",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>(Minor issue, but i'll ask about it since i came across it)</p>\n<p>I'm trying to make sense of some weird, time-travelling, itineraries.</p>\n<p>Take this round trip, its outbound departing on \"2024-08-21T16:00:00 UTC+5\" and its inbound arriving on \"2024-08-21T15:10:00 UTC+5\".</p>\n<p>While segment-wise it makes sense since both flight times are ok, itinerary-wise it does not.</p>\n<p>Should i regard these itineraries as product-building errors and simply remove them, or should i be cautious these departure &amp; arrival dates might not be ok and avoid using them?</p>\n<p>Thx!!</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F951212%2Fa78ce8547573b3b097669f565b615d59%2Fflightrank_example.png?generation=1751574089758286&amp;alt=media\" alt=\"\"></p>",
  "messages": [
    {
      "id": "3240454",
      "postDate": "07/03/2025 20:21:58",
      "content": "<p>(Minor issue, but i'll ask about it since i came across it)</p>\n<p>I'm trying to make sense of some weird, time-travelling, itineraries.</p>\n<p>Take this round trip, its outbound departing on \"2024-08-21T16:00:00 UTC+5\" and its inbound arriving on \"2024-08-21T15:10:00 UTC+5\".</p>\n<p>While segment-wise it makes sense since both flight times are ok, itinerary-wise it does not.</p>\n<p>Should i regard these itineraries as product-building errors and simply remove them, or should i be cautious these departure &amp; arrival dates might not be ok and avoid using them?</p>\n<p>Thx!!</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F951212%2Fa78ce8547573b3b097669f565b615d59%2Fflightrank_example.png?generation=1751574089758286&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "(Minor issue, but i'll ask about it since i came across it)\n\nI'm trying to make sense of some weird, time-travelling, itineraries.\n\nTake this round trip, its outbound departing on \"2024-08-21T16:00:00 UTC+5\" and its inbound arriving on \"2024-08-21T15:10:00 UTC+5\".\n\nWhile segment-wise it makes sense since both flight times are ok, itinerary-wise it does not.\n\nShould i regard these itineraries as product-building errors and simply remove them, or should i be cautious these departure & arrival dates might not be ok and avoid using them?\n\nThx!!\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F951212%2Fa78ce8547573b3b097669f565b615d59%2Fflightrank_example.png?generation=1751574089758286&alt=media)",
      "votes": null
    },
    {
      "id": "3241137",
      "postDate": "07/04/2025 15:22:19",
      "content": "<p>Could you please share ranker_id for that case?</p>",
      "rawMarkdown": "Could you please share ranker_id for that case?",
      "votes": null
    },
    {
      "id": "3241556",
      "postDate": "07/04/2025 21:49:54",
      "content": "<p>sure, these cases belong:<br>\n<code>ranker_id = \"05435e89b0b441edb26caad88e3027a7\"</code><br>\n<code>Id in [6900543, 6900544, 6900545]</code></p>",
      "rawMarkdown": "sure, these cases belong:\n`ranker_id = \"05435e89b0b441edb26caad88e3027a7\"`\n`Id in [6900543, 6900544, 6900545]`",
      "votes": null
    },
    {
      "id": "3242084",
      "postDate": "07/05/2025 14:35:34",
      "content": "<p>Thanks for pointing this out. These are interesting data anomalies. Flight aggregators and airline APIs sometimes return logically impossible itineraries like this. While uncommon, it's exactly the kind of raw data quality issue we encounter in production.</p>",
      "rawMarkdown": "Thanks for pointing this out. These are interesting data anomalies. Flight aggregators and airline APIs sometimes return logically impossible itineraries like this. While uncommon, it's exactly the kind of raw data quality issue we encounter in production.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3241137,
      "author_name": "samvelkoch",
      "author_url": "",
      "post_date": "07/04/2025 15:22:19",
      "content": "<p>Could you please share ranker_id for that case?</p>",
      "votes": null,
      "replies": [
        {
          "id": 3241556,
          "author_name": "lquiroz",
          "author_url": "",
          "post_date": "07/04/2025 21:49:54",
          "content": "<p>sure, these cases belong:<br>\n<code>ranker_id = \"05435e89b0b441edb26caad88e3027a7\"</code><br>\n<code>Id in [6900543, 6900544, 6900545]</code></p>",
          "votes": null,
          "replies": [
            {
              "id": 3242084,
              "author_name": "samvelkoch",
              "author_url": "",
              "post_date": "07/05/2025 14:35:34",
              "content": "<p>Thanks for pointing this out. These are interesting data anomalies. Flight aggregators and airline APIs sometimes return logically impossible itineraries like this. While uncommon, it's exactly the kind of raw data quality issue we encounter in production.</p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3240454": "(Minor issue, but i'll ask about it since i came across it)\n\nI'm trying to make sense of some weird, time-travelling, itineraries.\n\nTake this round trip, its outbound departing on \"2024-08-21T16:00:00 UTC+5\" and its inbound arriving on \"2024-08-21T15:10:00 UTC+5\".\n\nWhile segment-wise it makes sense since both flight times are ok, itinerary-wise it does not.\n\nShould i regard these itineraries as product-building errors and simply remove them, or should i be cautious these departure & arrival dates might not be ok and avoid using them?\n\nThx!!\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F951212%2Fa78ce8547573b3b097669f565b615d59%2Fflightrank_example.png?generation=1751574089758286&alt=media)",
    "3241137": "Could you please share ranker_id for that case?",
    "3241556": "sure, these cases belong:\n`ranker_id = \"05435e89b0b441edb26caad88e3027a7\"`\n`Id in [6900543, 6900544, 6900545]`",
    "3242084": "Thanks for pointing this out. These are interesting data anomalies. Flight aggregators and airline APIs sometimes return logically impossible itineraries like this. While uncommon, it's exactly the kind of raw data quality issue we encounter in production."
  },
  "source": "meta"
}