{
  "id": 3328,
  "title": "flighthistory.csv question",
  "url": "/competitions/flight/discussion/3328",
  "author_name": "",
  "post_date": "2012-12-14T13:47:57.343Z",
  "votes": null,
  "comment_count": 3,
  "views": 3170,
  "content": "<p>On this page (http://www.gequest.com/wiki/FlightQuest.Data.FlightHistory.FlightHistoryCsv) it says that flight_number is:</p>\r\n<p><span>&quot;A carrier-specific number identifying the flight (unique when combined with date and carrier).&quot;</span></p>\r\n<p>But for example I see 9 rows for&nbsp;flight_number 792 and airline_code 'LBQ'. &nbsp;(I'm looking at the&nbsp;SingleTrainingDay_2012_11_20.zip file.) &nbsp;So it doesn't seem unique to me.</p>\r\n<p>Would I better interpret flighthistory as each record being a &quot;leg&quot; of a full flight? &nbsp;</p>",
  "messages": [
    {
      "id": "17823",
      "postDate": "12/14/2012 13:47:57",
      "content": "<p>On this page (http://www.gequest.com/wiki/FlightQuest.Data.FlightHistory.FlightHistoryCsv) it says that flight_number is:</p>\r\n<p><span>&quot;A carrier-specific number identifying the flight (unique when combined with date and carrier).&quot;</span></p>\r\n<p>But for example I see 9 rows for&nbsp;flight_number 792 and airline_code 'LBQ'. &nbsp;(I'm looking at the&nbsp;SingleTrainingDay_2012_11_20.zip file.) &nbsp;So it doesn't seem unique to me.</p>\r\n<p>Would I better interpret flighthistory as each record being a &quot;leg&quot; of a full flight? &nbsp;</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "17825",
      "postDate": "12/14/2012 14:27:43",
      "content": "<p>You didn't take into account the date.</p>\r\n<p>As you quoted &quot;<span>A carrier-specific number identifying the flight (unique when combined with\r\n<span style=\"text-decoration:underline\"><strong>date and carrier</strong></span>).</span>&quot;</p>\r\n<p>When combined with both date (including timestamp, not just dd.mm.yyyy) and carrier code, then those lines are unique.</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "17846",
      "postDate": "12/14/2012 20:58:46",
      "content": "<p>Is there any problem of confidentiality by giving the flight number and the airline code?</p>",
      "rawMarkdown": "",
      "votes": null
    },
    {
      "id": "17854",
      "postDate": "12/14/2012 23:58:41",
      "content": "<p>DDB, I think that your &quot;leg interpretation&quot; is correct. Each flight<em>number/airline</em>code combination will only occur once per day, but that may involve multiple entries (one for each leg of the journey). What unique is that flight AA333 will run at\r\n most once per day. If another flight runs along the same route later in the day, it will have a different flight_number.</p>\r\n<p>Predrag, You're correct although I don't think that's what they meant. They don't mean that the combination is unique in the database sense of row uniqueness. They mean that Flight AA333 will fly it's route at most once per day. That route may however involve\r\n multiple entries (i.e. legs) in the table. Presumably they won't include duplicate entries for each leg, so like you said, arrival/departure times should be unique.</p>\r\n<p>Toulouse, in what way? Flight numbers are public information. That's how you know which flight to get on when you're at the airport. The airlines publish their schedules... The only thing that might be considered proprietary knowledge would be the actual\r\n tail number of the plane which I think shows up in the events file, but I'm pretty sure that's publicly available as well.</p>",
      "rawMarkdown": "",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 17825,
      "author_name": "predragbradaric",
      "author_url": "",
      "post_date": "12/14/2012 14:27:43",
      "content": "<p>You didn't take into account the date.</p>\r\n<p>As you quoted &quot;<span>A carrier-specific number identifying the flight (unique when combined with\r\n<span style=\"text-decoration:underline\"><strong>date and carrier</strong></span>).</span>&quot;</p>\r\n<p>When combined with both date (including timestamp, not just dd.mm.yyyy) and carrier code, then those lines are unique.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 17846,
      "author_name": "toulouse",
      "author_url": "",
      "post_date": "12/14/2012 20:58:46",
      "content": "<p>Is there any problem of confidentiality by giving the flight number and the airline code?</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 17854,
      "author_name": "cephalopod",
      "author_url": "",
      "post_date": "12/14/2012 23:58:41",
      "content": "<p>DDB, I think that your &quot;leg interpretation&quot; is correct. Each flight<em>number/airline</em>code combination will only occur once per day, but that may involve multiple entries (one for each leg of the journey). What unique is that flight AA333 will run at\r\n most once per day. If another flight runs along the same route later in the day, it will have a different flight_number.</p>\r\n<p>Predrag, You're correct although I don't think that's what they meant. They don't mean that the combination is unique in the database sense of row uniqueness. They mean that Flight AA333 will fly it's route at most once per day. That route may however involve\r\n multiple entries (i.e. legs) in the table. Presumably they won't include duplicate entries for each leg, so like you said, arrival/departure times should be unique.</p>\r\n<p>Toulouse, in what way? Flight numbers are public information. That's how you know which flight to get on when you're at the airport. The airlines publish their schedules... The only thing that might be considered proprietary knowledge would be the actual\r\n tail number of the plane which I think shows up in the events file, but I'm pretty sure that's publicly available as well.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "17823": "",
    "17825": "",
    "17846": "",
    "17854": ""
  },
  "source": "meta"
}