{
  "id": 23312,
  "title": "timestamp for when each measurement was taken",
  "url": "/competitions/bosch-production-line-performance/discussion/23312",
  "author_name": "",
  "post_date": "2016-09-03T06:53:50.123Z",
  "votes": null,
  "comment_count": 5,
  "views": 1071,
  "content": "<p>The description <a href=\"https://www.kaggle.com/c/bosch-production-line-performance/data\">https://www.kaggle.com/c/bosch-production-line-performance/data</a> says &quot;The date features provide a timestamp for when each measurement was taken&quot;. However, I found that there are 3109 features (2140 categorical features + 969 numeric features) and the number of date fields is only 1156. That means almost 2000 features have no timestamp.</p>\n\n<p>I am wondering how come a big company like Bosch can &quot;forget&quot; to add a timestamp for those features. Or the Kaggle team can explain us more about those missing timestamp fields.</p>",
  "messages": [
    {
      "id": "133661",
      "postDate": "09/03/2016 06:53:50",
      "content": "<p>The description <a href=\"https://www.kaggle.com/c/bosch-production-line-performance/data\">https://www.kaggle.com/c/bosch-production-line-performance/data</a> says &quot;The date features provide a timestamp for when each measurement was taken&quot;. However, I found that there are 3109 features (2140 categorical features + 969 numeric features) and the number of date fields is only 1156. That means almost 2000 features have no timestamp.</p>\n\n<p>I am wondering how come a big company like Bosch can &quot;forget&quot; to add a timestamp for those features. Or the Kaggle team can explain us more about those missing timestamp fields.</p>",
      "rawMarkdown": "The description [https://www.kaggle.com/c/bosch-production-line-performance/data][1] says \"The date features provide a timestamp for when each measurement was taken\". However, I found that there are 3109 features (2140 categorical features + 969 numeric features) and the number of date fields is only 1156. That means almost 2000 features have no timestamp.\r\n\r\nI am wondering how come a big company like Bosch can \"forget\" to add a timestamp for those features. Or the Kaggle team can explain us more about those missing timestamp fields.\r\n\r\n\r\n  [1]: https://www.kaggle.com/c/bosch-production-line-performance/data",
      "votes": null
    },
    {
      "id": "133668",
      "postDate": "09/03/2016 07:54:18",
      "content": "<p>There will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.</p>",
      "rawMarkdown": "There will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.",
      "votes": null
    },
    {
      "id": "134150",
      "postDate": "09/05/2016 12:41:48",
      "content": "<p>In my factory, we often take multiple measurements at a station. For example, we may measure thickness at several locations. I suppose this is what Prasanth means by &quot;literally impossible to register timestamp for every measurement that is taken&quot;.  That said, every measurement event has a timestamp associated with the collection of data. It seems that there is one timestamp for all data associated with a particular station (i.e. L0_S0_% has the same timestamp for items in each row)</p>",
      "rawMarkdown": "In my factory, we often take multiple measurements at a station. For example, we may measure thickness at several locations. I suppose this is what Prasanth means by \"literally impossible to register timestamp for every measurement that is taken\".  That said, every measurement event has a timestamp associated with the collection of data. It seems that there is one timestamp for all data associated with a particular station (i.e. L0_S0_% has the same timestamp for items in each row)",
      "votes": null
    },
    {
      "id": "134364",
      "postDate": "09/06/2016 10:22:19",
      "content": "<p>Another question i have on timestamp:</p>\n\n<p>Data Description states : &quot;The date features provide a timestamp for when each measurement was taken. Each date column ends in a number that corresponds to the previous feature number. E.g. the value of L0_S0_D1 is the time at which L0_S0_F0 was taken.&quot;</p>\n\n<p>So my question is : Which observation field corresponds to Date Field : L0_S7_D139, L0_S7_D140 and L0_S7_D141?\nOne has L0_S7_F138 (related to L0_S7_D139 [Good]) as a Numeric measurement. What about measurements L0_S7_F139 (for L0_S7_D140 [?]) and L0_S7_F140  (for L0_S7_D41 [?])? There is none of these in the categorical / numerical file.</p>\n\n<p>Pardon please, in case I am wrong in my observations.</p>",
      "rawMarkdown": "Another question i have on timestamp:\r\n\r\nData Description states : \"The date features provide a timestamp for when each measurement was taken. Each date column ends in a number that corresponds to the previous feature number. E.g. the value of L0_S0_D1 is the time at which L0_S0_F0 was taken.\"\r\n\r\nSo my question is : Which observation field corresponds to Date Field : L0_S7_D139, L0_S7_D140 and L0_S7_D141?\r\nOne has L0_S7_F138 (related to L0_S7_D139 [Good]) as a Numeric measurement. What about measurements L0_S7_F139 (for L0_S7_D140 [?]) and L0_S7_F140  (for L0_S7_D41 [?])? There is none of these in the categorical / numerical file.\r\n\r\nPardon please, in case I am wrong in my observations.",
      "votes": null
    },
    {
      "id": "134509",
      "postDate": "09/07/2016 01:17:56",
      "content": "<p>Also when I thought of timestamp, I was thinking of an actual date and time.  Here it looks like a Float (Possibly Days with decimal fraction of day).  Maybe I am missing something.</p>",
      "rawMarkdown": "Also when I thought of timestamp, I was thinking of an actual date and time.  Here it looks like a Float (Possibly Days with decimal fraction of day).  Maybe I am missing something.",
      "votes": null
    },
    {
      "id": "134531",
      "postDate": "09/07/2016 04:50:08",
      "content": "<p>[quote=Prasanth;133668]</p>\n\n<p>There will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.</p>\n\n<p>[/quote]</p>\n\n<p>What I am guessing is, each product need not pass through all the stages. Different products pass through different levels and stages. For the same reason, we see its very sparse!</p>",
      "rawMarkdown": "[quote=Prasanth;133668]\r\n\r\nThere will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.\r\n\r\n[/quote]\r\n\r\nWhat I am guessing is, each product need not pass through all the stages. Different products pass through different levels and stages. For the same reason, we see its very sparse!",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 133668,
      "author_name": "datasciencerocks",
      "author_url": "",
      "post_date": "09/03/2016 07:54:18",
      "content": "<p>There will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 134150,
      "author_name": "jeffhebert",
      "author_url": "",
      "post_date": "09/05/2016 12:41:48",
      "content": "<p>In my factory, we often take multiple measurements at a station. For example, we may measure thickness at several locations. I suppose this is what Prasanth means by &quot;literally impossible to register timestamp for every measurement that is taken&quot;.  That said, every measurement event has a timestamp associated with the collection of data. It seems that there is one timestamp for all data associated with a particular station (i.e. L0_S0_% has the same timestamp for items in each row)</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 134364,
      "author_name": "sudhirghiya",
      "author_url": "",
      "post_date": "09/06/2016 10:22:19",
      "content": "<p>Another question i have on timestamp:</p>\n\n<p>Data Description states : &quot;The date features provide a timestamp for when each measurement was taken. Each date column ends in a number that corresponds to the previous feature number. E.g. the value of L0_S0_D1 is the time at which L0_S0_F0 was taken.&quot;</p>\n\n<p>So my question is : Which observation field corresponds to Date Field : L0_S7_D139, L0_S7_D140 and L0_S7_D141?\nOne has L0_S7_F138 (related to L0_S7_D139 [Good]) as a Numeric measurement. What about measurements L0_S7_F139 (for L0_S7_D140 [?]) and L0_S7_F140  (for L0_S7_D41 [?])? There is none of these in the categorical / numerical file.</p>\n\n<p>Pardon please, in case I am wrong in my observations.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 134509,
      "author_name": "rdslater",
      "author_url": "",
      "post_date": "09/07/2016 01:17:56",
      "content": "<p>Also when I thought of timestamp, I was thinking of an actual date and time.  Here it looks like a Float (Possibly Days with decimal fraction of day).  Maybe I am missing something.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 134531,
      "author_name": "saikumarallaka",
      "author_url": "",
      "post_date": "09/07/2016 04:50:08",
      "content": "<p>[quote=Prasanth;133668]</p>\n\n<p>There will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.</p>\n\n<p>[/quote]</p>\n\n<p>What I am guessing is, each product need not pass through all the stages. Different products pass through different levels and stages. For the same reason, we see its very sparse!</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "133661": "The description [https://www.kaggle.com/c/bosch-production-line-performance/data][1] says \"The date features provide a timestamp for when each measurement was taken\". However, I found that there are 3109 features (2140 categorical features + 969 numeric features) and the number of date fields is only 1156. That means almost 2000 features have no timestamp.\r\n\r\nI am wondering how come a big company like Bosch can \"forget\" to add a timestamp for those features. Or the Kaggle team can explain us more about those missing timestamp fields.\r\n\r\n\r\n  [1]: https://www.kaggle.com/c/bosch-production-line-performance/data",
    "133668": "There will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.",
    "134150": "In my factory, we often take multiple measurements at a station. For example, we may measure thickness at several locations. I suppose this is what Prasanth means by \"literally impossible to register timestamp for every measurement that is taken\".  That said, every measurement event has a timestamp associated with the collection of data. It seems that there is one timestamp for all data associated with a particular station (i.e. L0_S0_% has the same timestamp for items in each row)",
    "134364": "Another question i have on timestamp:\r\n\r\nData Description states : \"The date features provide a timestamp for when each measurement was taken. Each date column ends in a number that corresponds to the previous feature number. E.g. the value of L0_S0_D1 is the time at which L0_S0_F0 was taken.\"\r\n\r\nSo my question is : Which observation field corresponds to Date Field : L0_S7_D139, L0_S7_D140 and L0_S7_D141?\r\nOne has L0_S7_F138 (related to L0_S7_D139 [Good]) as a Numeric measurement. What about measurements L0_S7_F139 (for L0_S7_D140 [?]) and L0_S7_F140  (for L0_S7_D41 [?])? There is none of these in the categorical / numerical file.\r\n\r\nPardon please, in case I am wrong in my observations.",
    "134509": "Also when I thought of timestamp, I was thinking of an actual date and time.  Here it looks like a Float (Possibly Days with decimal fraction of day).  Maybe I am missing something.",
    "134531": "[quote=Prasanth;133668]\r\n\r\nThere will always be some missing data. And also it is literally impossible to register timestamp for every measurement that is taken.\r\n\r\n[/quote]\r\n\r\nWhat I am guessing is, each product need not pass through all the stages. Different products pass through different levels and stages. For the same reason, we see its very sparse!"
  },
  "source": "meta"
}