{
  "id": 57650,
  "title": " Garbage or artifacts in CELLS data",
  "url": "/competitions/trackml-particle-identification/discussion/57650",
  "author_name": "",
  "post_date": "2018-05-26T19:20:51.040844900Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>While doing data exploration I have found that large quantity of the signal values are exactly equel to 1. </p>\n\n<p>From physical or modelling perspective it seems non-realistic. Has anybody  figured out if this is a garbage, model limitations or something else?</p>\n\n<p><img src=\"https://s5.wampi.ru/2018/05/26/cells.jpg\" alt=\"values\"></p>",
  "messages": [
    {
      "id": "334251",
      "postDate": "05/26/2018 19:20:51",
      "content": "<p>While doing data exploration I have found that large quantity of the signal values are exactly equel to 1. </p>\n\n<p>From physical or modelling perspective it seems non-realistic. Has anybody  figured out if this is a garbage, model limitations or something else?</p>\n\n<p><img src=\"https://s5.wampi.ru/2018/05/26/cells.jpg\" alt=\"values\"></p>",
      "rawMarkdown": "While doing data exploration I have found that large quantity of the signal values are exactly equel to 1. \n\nFrom physical or modelling perspective it seems non-realistic. Has anybody  figured out if this is a garbage, model limitations or something else?\n\n![values][1]\n\n\n  [1]: https://s5.wampi.ru/2018/05/26/cells.jpg",
      "votes": null
    },
    {
      "id": "334278",
      "postDate": "05/26/2018 21:27:39",
      "content": "<p>If you split by volume_id, you'll see that this comes from the strip detectors (the outermost ones). This is because in the strip detector a digital readout (no information on the charge, just 0 or 1) is simulated. This is also the case for real experiments because the outermost  detectors being larger in square meters, it would be quite expensive to have an analogic measurement there.</p>",
      "rawMarkdown": "If you split by volume_id, you'll see that this comes from the strip detectors (the outermost ones). This is because in the strip detector a digital readout (no information on the charge, just 0 or 1) is simulated. This is also the case for real experiments because the outermost  detectors being larger in square meters, it would be quite expensive to have an analogic measurement there.",
      "votes": null
    },
    {
      "id": "334432",
      "postDate": "05/27/2018 11:27:13",
      "content": "<p>Interesting, thanks for clarification.</p>",
      "rawMarkdown": "Interesting, thanks for clarification.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 334278,
      "author_name": "droussea",
      "author_url": "",
      "post_date": "05/26/2018 21:27:39",
      "content": "<p>If you split by volume_id, you'll see that this comes from the strip detectors (the outermost ones). This is because in the strip detector a digital readout (no information on the charge, just 0 or 1) is simulated. This is also the case for real experiments because the outermost  detectors being larger in square meters, it would be quite expensive to have an analogic measurement there.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 334432,
      "author_name": "denisdmitriev",
      "author_url": "",
      "post_date": "05/27/2018 11:27:13",
      "content": "<p>Interesting, thanks for clarification.</p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "334251": "While doing data exploration I have found that large quantity of the signal values are exactly equel to 1. \n\nFrom physical or modelling perspective it seems non-realistic. Has anybody  figured out if this is a garbage, model limitations or something else?\n\n![values][1]\n\n\n  [1]: https://s5.wampi.ru/2018/05/26/cells.jpg",
    "334278": "If you split by volume_id, you'll see that this comes from the strip detectors (the outermost ones). This is because in the strip detector a digital readout (no information on the charge, just 0 or 1) is simulated. This is also the case for real experiments because the outermost  detectors being larger in square meters, it would be quite expensive to have an analogic measurement there.",
    "334432": "Interesting, thanks for clarification."
  },
  "source": "meta"
}