{
  "id": 368620,
  "title": "STFT Time domain length question",
  "url": "/competitions/g2net-detecting-continuous-gravitational-waves/discussion/368620",
  "author_name": "Husam Elfadil",
  "post_date": "2022-11-26T16:20:50.741000",
  "votes": 1,
  "comment_count": 3,
  "views": 0,
  "content": "<p>We know we have 360 frequency pins. For time domain pins (the second dimension of STFT), I noticed the length of H1 sensor is different than L1 sensor within the same sample, which is fine. My question is the H1 sensor number of time pins is the same across all samples? also similar question for L1 sensor?</p>",
  "messages": [
    {
      "id": 2044627,
      "postDate": "2022-11-26T17:09:47.473Z",
      "content": "<p>I think they are all longer than 4000, but I'm not sure. According to Jun Koda's notebook, he compressed the first 4096 timestamps of all data in train and test dataset, and seems to meet no trouble.</p>",
      "rawMarkdown": "I think they are all longer than 4000, but I'm not sure. According to Jun Koda's notebook, he compressed the first 4096 timestamps of all data in train and test dataset, and seems to meet no trouble."
    },
    {
      "id": 2044617,
      "postDate": "2022-11-26T16:58:40.060Z",
      "content": "<p>Thanks for your response. Is the a minimum number of time pins (or lower limit) for H1 or L1 sensor, or it can be any number of pins?</p>",
      "rawMarkdown": "Thanks for your response. Is the a minimum number of time pins (or lower limit) for H1 or L1 sensor, or it can be any number of pins?"
    },
    {
      "id": 2044606,
      "postDate": "2022-11-26T16:53:10.373Z",
      "content": "<p>No, they are different. Figure below demontrates the distributions of the 'number of time pins' for both H1 and L1.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7065518%2Fc5097e9c1358be5c431557cdf1cdc5a5%2FScreenshot_20221127_005058.png?generation=1669481577781865&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "No, they are different. Figure below demontrates the distributions of the 'number of time pins' for both H1 and L1.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7065518%2Fc5097e9c1358be5c431557cdf1cdc5a5%2FScreenshot_20221127_005058.png?generation=1669481577781865&alt=media)"
    },
    {
      "id": 2044574,
      "postDate": "2022-11-26T16:20:50.743Z",
      "content": "<p>We know we have 360 frequency pins. For time domain pins (the second dimension of STFT), I noticed the length of H1 sensor is different than L1 sensor within the same sample, which is fine. My question is the H1 sensor number of time pins is the same across all samples? also similar question for L1 sensor?</p>",
      "rawMarkdown": "We know we have 360 frequency pins. For time domain pins (the second dimension of STFT), I noticed the length of H1 sensor is different than L1 sensor within the same sample, which is fine. My question is the H1 sensor number of time pins is the same across all samples? also similar question for L1 sensor?"
    }
  ],
  "comments": [
    {
      "id": 2044627,
      "author_name": "zzzc18",
      "author_url": "",
      "post_date": "2022-11-26T17:09:47.473000",
      "content": "<p>I think they are all longer than 4000, but I'm not sure. According to Jun Koda's notebook, he compressed the first 4096 timestamps of all data in train and test dataset, and seems to meet no trouble.</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2044617,
      "author_name": "Husam Elfadil",
      "author_url": "",
      "post_date": "2022-11-26T16:58:40.060000",
      "content": "<p>Thanks for your response. Is the a minimum number of time pins (or lower limit) for H1 or L1 sensor, or it can be any number of pins?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 2044606,
      "author_name": "zzzc18",
      "author_url": "",
      "post_date": "2022-11-26T16:53:10.373000",
      "content": "<p>No, they are different. Figure below demontrates the distributions of the 'number of time pins' for both H1 and L1.</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7065518%2Fc5097e9c1358be5c431557cdf1cdc5a5%2FScreenshot_20221127_005058.png?generation=1669481577781865&amp;alt=media\" alt=\"\"></p>",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "2044627": "I think they are all longer than 4000, but I'm not sure. According to Jun Koda's notebook, he compressed the first 4096 timestamps of all data in train and test dataset, and seems to meet no trouble.",
    "2044617": "Thanks for your response. Is the a minimum number of time pins (or lower limit) for H1 or L1 sensor, or it can be any number of pins?",
    "2044606": "No, they are different. Figure below demontrates the distributions of the 'number of time pins' for both H1 and L1.\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F7065518%2Fc5097e9c1358be5c431557cdf1cdc5a5%2FScreenshot_20221127_005058.png?generation=1669481577781865&alt=media)",
    "2044574": "We know we have 360 frequency pins. For time domain pins (the second dimension of STFT), I noticed the length of H1 sensor is different than L1 sensor within the same sample, which is fine. My question is the H1 sensor number of time pins is the same across all samples? also similar question for L1 sensor?"
  }
}