{
  "id": 191564,
  "title": "a few questions",
  "url": "/competitions/predict-volcanic-eruptions-ingv-oe/discussion/191564",
  "author_name": "",
  "post_date": "2020-10-17T07:34:59.959660700Z",
  "votes": 10,
  "comment_count": 3,
  "views": 0,
  "content": "<ol>\n<li>Is zero (0) a valid sensor reading? For example, in file 1000213997.csv sensor #10 has only zeros. Is that a valid reading?</li>\n<li>I see that the time to erupt is an integer. Do you also expect the prediction to be an integer?</li>\n<li>Just to make sure, given a sequence of sensor readings we need to predict when is the next eruption? So, the eruption cloud not have happened sometime in the middle of these readings?</li>\n</ol>",
  "messages": [
    {
      "id": "1052016",
      "postDate": "10/17/2020 07:34:59",
      "content": "<ol>\n<li>Is zero (0) a valid sensor reading? For example, in file 1000213997.csv sensor #10 has only zeros. Is that a valid reading?</li>\n<li>I see that the time to erupt is an integer. Do you also expect the prediction to be an integer?</li>\n<li>Just to make sure, given a sequence of sensor readings we need to predict when is the next eruption? So, the eruption cloud not have happened sometime in the middle of these readings?</li>\n</ol>",
      "rawMarkdown": "1. Is zero (0) a valid sensor reading? For example, in file 1000213997.csv sensor #10 has only zeros. Is that a valid reading?\n2. I see that the time to erupt is an integer. Do you also expect the prediction to be an integer?\n3. Just to make sure, given a sequence of sensor readings we need to predict when is the next eruption? So, the eruption cloud not have happened sometime in the middle of these readings?",
      "votes": null
    },
    {
      "id": "1052116",
      "postDate": "10/17/2020 10:45:18",
      "content": "<p>1) No, sequences of 0s indicate malfunctions in the sensor.<br>\n2) Considering the sampling rate (100Hz) and the uncertainty associated to the eruption start times the decimals lose their meaning.<br>\n3) Right.</p>",
      "rawMarkdown": "1) No, sequences of 0s indicate malfunctions in the sensor.\n2) Considering the sampling rate (100Hz) and the uncertainty associated to the eruption start times the decimals lose their meaning.\n3) Right.",
      "votes": null
    },
    {
      "id": "1052680",
      "postDate": "10/18/2020 05:32:18",
      "content": "<p>Only sequences of 0s indicate malfunctions in the sensor or any 0 value? I'm asking becasue if 0 (or another value) are not valid sensor reading, then we need to change it to NULL. Otherwise, 0s will be considered in the analysis.</p>",
      "rawMarkdown": "Only sequences of 0s indicate malfunctions in the sensor or any 0 value? I'm asking becasue if 0 (or another value) are not valid sensor reading, then we need to change it to NULL. Otherwise, 0s will be considered in the analysis.",
      "votes": null
    },
    {
      "id": "1053041",
      "postDate": "10/18/2020 14:38:50",
      "content": "<p>Isolated 0s could be true readings, while this is extremely unlikely for sequences of 0s. </p>",
      "rawMarkdown": "Isolated 0s could be true readings, while this is extremely unlikely for sequences of 0s.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 1052116,
      "author_name": "flaxio",
      "author_url": "",
      "post_date": "10/17/2020 10:45:18",
      "content": "<p>1) No, sequences of 0s indicate malfunctions in the sensor.<br>\n2) Considering the sampling rate (100Hz) and the uncertainty associated to the eruption start times the decimals lose their meaning.<br>\n3) Right.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1052680,
      "author_name": "ranpelta",
      "author_url": "",
      "post_date": "10/18/2020 05:32:18",
      "content": "<p>Only sequences of 0s indicate malfunctions in the sensor or any 0 value? I'm asking becasue if 0 (or another value) are not valid sensor reading, then we need to change it to NULL. Otherwise, 0s will be considered in the analysis.</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 1053041,
      "author_name": "flaxio",
      "author_url": "",
      "post_date": "10/18/2020 14:38:50",
      "content": "<p>Isolated 0s could be true readings, while this is extremely unlikely for sequences of 0s. </p>",
      "votes": null,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "1052016": "1. Is zero (0) a valid sensor reading? For example, in file 1000213997.csv sensor #10 has only zeros. Is that a valid reading?\n2. I see that the time to erupt is an integer. Do you also expect the prediction to be an integer?\n3. Just to make sure, given a sequence of sensor readings we need to predict when is the next eruption? So, the eruption cloud not have happened sometime in the middle of these readings?",
    "1052116": "1) No, sequences of 0s indicate malfunctions in the sensor.\n2) Considering the sampling rate (100Hz) and the uncertainty associated to the eruption start times the decimals lose their meaning.\n3) Right.",
    "1052680": "Only sequences of 0s indicate malfunctions in the sensor or any 0 value? I'm asking becasue if 0 (or another value) are not valid sensor reading, then we need to change it to NULL. Otherwise, 0s will be considered in the analysis.",
    "1053041": "Isolated 0s could be true readings, while this is extremely unlikely for sequences of 0s."
  },
  "source": "meta"
}