{
  "id": 23201,
  "title": "Understanding of train_date",
  "url": "/competitions/bosch-production-line-performance/discussion/23201",
  "author_name": "",
  "post_date": "2016-08-31T03:43:37.480Z",
  "votes": null,
  "comment_count": 3,
  "views": 1331,
  "content": "<p>I am struggling to understand  the meaning of train_date means? It says &quot;  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>But how do we know that L0_S0_D1 corresponds to L0_S0_F0? </p>\n\n<p>I am exploring the possibility of using survival analysis for dealing with this competition but stucked in fully understanding the dates. </p>",
  "messages": [
    {
      "id": "133054",
      "postDate": "08/31/2016 03:43:37",
      "content": "<p>I am struggling to understand  the meaning of train_date means? It says &quot;  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>But how do we know that L0_S0_D1 corresponds to L0_S0_F0? </p>\n\n<p>I am exploring the possibility of using survival analysis for dealing with this competition but stucked in fully understanding the dates. </p>",
      "rawMarkdown": "I am struggling to understand  the meaning of train_date means? It says \"  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\nBut how do we know that L0_S0_D1 corresponds to L0_S0_F0? \r\n\r\nI am exploring the possibility of using survival analysis for dealing with this competition but stucked in fully understanding the dates.",
      "votes": null
    },
    {
      "id": "133115",
      "postDate": "08/31/2016 11:03:00",
      "content": "<p>You know it from the last number in the feature name.  </p>\n\n<p>A date name ending with n is the date at which the feature ending with n-1 was measured, id per id.</p>",
      "rawMarkdown": "You know it from the last number in the feature name.  \r\n\r\nA date name ending with n is the date at which the feature ending with n-1 was measured, id per id.",
      "votes": null
    },
    {
      "id": "133354",
      "postDate": "09/01/2016 16:36:25",
      "content": "<p>I have also the trouble to understand the  data. \nRegarding the values in the date file, for example the first line in the train file, does it mean the time is 82.24 while the feature L0_S0_F0 was measured?</p>\n\n<p>how do we interpret 82.24 as time?   I think it would be interesting to know, the issue may appear periodically along with the time... </p>",
      "rawMarkdown": "I have also the trouble to understand the  data. \r\nRegarding the values in the date file, for example the first line in the train file, does it mean the time is 82.24 while the feature L0_S0_F0 was measured?\r\n\r\nhow do we interpret 82.24 as time?   I think it would be interesting to know, the issue may appear periodically along with the time...",
      "votes": null
    },
    {
      "id": "135287",
      "postDate": "09/12/2016 20:43:46",
      "content": "<p>Not sure how we can interpret it exactly, we have no information about that. I guess we can use it as it is.</p>\n\n<ul>\n<li>One line corresponds to all the measurement timestamps for a given part (going from 82.24 to 82.27 for the first file for example).</li>\n<li>One column is the timestamps for a given feature (CPMP explained how to match feature with timestamp in the previous comment).</li>\n</ul>",
      "rawMarkdown": "Not sure how we can interpret it exactly, we have no information about that. I guess we can use it as it is.\r\n\r\n- One line corresponds to all the measurement timestamps for a given part (going from 82.24 to 82.27 for the first file for example).\r\n- One column is the timestamps for a given feature (CPMP explained how to match feature with timestamp in the previous comment).",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 133115,
      "author_name": "cpmpml",
      "author_url": "",
      "post_date": "08/31/2016 11:03:00",
      "content": "<p>You know it from the last number in the feature name.  </p>\n\n<p>A date name ending with n is the date at which the feature ending with n-1 was measured, id per id.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 133354,
      "author_name": "victorgu",
      "author_url": "",
      "post_date": "09/01/2016 16:36:25",
      "content": "<p>I have also the trouble to understand the  data. \nRegarding the values in the date file, for example the first line in the train file, does it mean the time is 82.24 while the feature L0_S0_F0 was measured?</p>\n\n<p>how do we interpret 82.24 as time?   I think it would be interesting to know, the issue may appear periodically along with the time... </p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 135287,
      "author_name": "phansoks",
      "author_url": "",
      "post_date": "09/12/2016 20:43:46",
      "content": "<p>Not sure how we can interpret it exactly, we have no information about that. I guess we can use it as it is.</p>\n\n<ul>\n<li>One line corresponds to all the measurement timestamps for a given part (going from 82.24 to 82.27 for the first file for example).</li>\n<li>One column is the timestamps for a given feature (CPMP explained how to match feature with timestamp in the previous comment).</li>\n</ul>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "133054": "I am struggling to understand  the meaning of train_date means? It says \"  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\nBut how do we know that L0_S0_D1 corresponds to L0_S0_F0? \r\n\r\nI am exploring the possibility of using survival analysis for dealing with this competition but stucked in fully understanding the dates.",
    "133115": "You know it from the last number in the feature name.  \r\n\r\nA date name ending with n is the date at which the feature ending with n-1 was measured, id per id.",
    "133354": "I have also the trouble to understand the  data. \r\nRegarding the values in the date file, for example the first line in the train file, does it mean the time is 82.24 while the feature L0_S0_F0 was measured?\r\n\r\nhow do we interpret 82.24 as time?   I think it would be interesting to know, the issue may appear periodically along with the time...",
    "135287": "Not sure how we can interpret it exactly, we have no information about that. I guess we can use it as it is.\r\n\r\n- One line corresponds to all the measurement timestamps for a given part (going from 82.24 to 82.27 for the first file for example).\r\n- One column is the timestamps for a given feature (CPMP explained how to match feature with timestamp in the previous comment)."
  },
  "source": "meta"
}