{
  "id": 483072,
  "title": "understanding spectrogram plots",
  "url": "/competitions/hms-harmful-brain-activity-classification/discussion/483072",
  "author_name": "",
  "post_date": "2024-03-10T19:34:28.220471700Z",
  "votes": 1,
  "comment_count": 2,
  "views": 0,
  "content": "<p>Hi guys, <br>\nin my jupyter notebook I did plot eeg signals from one parquet file from eeg_signals folder. an example is Fp1 signal:<br>\n <img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F0ea2a93957e14bb84de490c1fc0deb4c%2Feeg_signal_fp1.png?generation=1710098594827899&amp;alt=media\"></p>\n<p>I did perform moving average to smooth the signal. The sampling frequency from my undertstanding of data explanation is 200 Hz</p>\n<p>I ploted its spectrogram using scipy.sginals.spectrogram:<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F7bd5b9731575404474f9816facb9e52f%2Feeg_signal_spectrogram_fp1.png?generation=1710098817394458&amp;alt=media\"></p>\n<p>Then i plotted the realted spectrogram using parquet file from folder eeg_spectrograms:<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F37a984dcdb133d9ec7972dd6284ca900%2Feeg_spectrograms_parquet.png?generation=1710098957022133&amp;alt=media\"></p>\n<p>Maybe my question directed more to signal processing people and engineers. Could you please explain to me the difference between these two spectrograms. Do plots look approperiate? Maybe if it is possible you can explain to me the train csv columns from the plots because I don't understand the 50 seconds eeg subsample id column explanation and 10 mintues in spectrogram_sub_id. </p>\n<p>Thank you in advance</p>",
  "messages": [
    {
      "id": "2690784",
      "postDate": "03/10/2024 19:34:28",
      "content": "<p>Hi guys, <br>\nin my jupyter notebook I did plot eeg signals from one parquet file from eeg_signals folder. an example is Fp1 signal:<br>\n <img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F0ea2a93957e14bb84de490c1fc0deb4c%2Feeg_signal_fp1.png?generation=1710098594827899&amp;alt=media\"></p>\n<p>I did perform moving average to smooth the signal. The sampling frequency from my undertstanding of data explanation is 200 Hz</p>\n<p>I ploted its spectrogram using scipy.sginals.spectrogram:<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F7bd5b9731575404474f9816facb9e52f%2Feeg_signal_spectrogram_fp1.png?generation=1710098817394458&amp;alt=media\"></p>\n<p>Then i plotted the realted spectrogram using parquet file from folder eeg_spectrograms:<br>\n<img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F37a984dcdb133d9ec7972dd6284ca900%2Feeg_spectrograms_parquet.png?generation=1710098957022133&amp;alt=media\"></p>\n<p>Maybe my question directed more to signal processing people and engineers. Could you please explain to me the difference between these two spectrograms. Do plots look approperiate? Maybe if it is possible you can explain to me the train csv columns from the plots because I don't understand the 50 seconds eeg subsample id column explanation and 10 mintues in spectrogram_sub_id. </p>\n<p>Thank you in advance</p>",
      "rawMarkdown": "Hi guys, \nin my jupyter notebook I did plot eeg signals from one parquet file from eeg_signals folder. an example is Fp1 signal:\n ![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F0ea2a93957e14bb84de490c1fc0deb4c%2Feeg_signal_fp1.png?generation=1710098594827899&alt=media)\n\nI did perform moving average to smooth the signal. The sampling frequency from my undertstanding of data explanation is 200 Hz\n\nI ploted its spectrogram using scipy.sginals.spectrogram:\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F7bd5b9731575404474f9816facb9e52f%2Feeg_signal_spectrogram_fp1.png?generation=1710098817394458&alt=media)\n\nThen i plotted the realted spectrogram using parquet file from folder eeg_spectrograms:\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F37a984dcdb133d9ec7972dd6284ca900%2Feeg_spectrograms_parquet.png?generation=1710098957022133&alt=media)\n\nMaybe my question directed more to signal processing people and engineers. Could you please explain to me the difference between these two spectrograms. Do plots look approperiate? Maybe if it is possible you can explain to me the train csv columns from the plots because I don't understand the 50 seconds eeg subsample id column explanation and 10 mintues in spectrogram_sub_id. \n\nThank you in advance",
      "votes": null
    },
    {
      "id": "2690856",
      "postDate": "03/10/2024 20:48:39",
      "content": "<p><a href=\"https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification/discussion/468010\" target=\"_blank\">https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification/discussion/468010</a><br>\ngive this a look, it will explain everything!<br>\nAs far as how the kaggle spectrogram were generated, i dont think we have any information on how they preprocessed. </p>",
      "rawMarkdown": "https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification/discussion/468010\ngive this a look, it will explain everything!\nAs far as how the kaggle spectrogram were generated, i dont think we have any information on how they preprocessed.",
      "votes": null
    },
    {
      "id": "2692312",
      "postDate": "03/11/2024 19:28:24",
      "content": "<p>thank you for your response</p>",
      "rawMarkdown": "thank you for your response",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 2690856,
      "author_name": "tashin47",
      "author_url": "",
      "post_date": "03/10/2024 20:48:39",
      "content": "<p><a href=\"https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification/discussion/468010\" target=\"_blank\">https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification/discussion/468010</a><br>\ngive this a look, it will explain everything!<br>\nAs far as how the kaggle spectrogram were generated, i dont think we have any information on how they preprocessed. </p>",
      "votes": null,
      "replies": [
        {
          "id": 2692312,
          "author_name": "mahmoodomaira",
          "author_url": "",
          "post_date": "03/11/2024 19:28:24",
          "content": "<p>thank you for your response</p>",
          "votes": null,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "2690784": "Hi guys, \nin my jupyter notebook I did plot eeg signals from one parquet file from eeg_signals folder. an example is Fp1 signal:\n ![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F0ea2a93957e14bb84de490c1fc0deb4c%2Feeg_signal_fp1.png?generation=1710098594827899&alt=media)\n\nI did perform moving average to smooth the signal. The sampling frequency from my undertstanding of data explanation is 200 Hz\n\nI ploted its spectrogram using scipy.sginals.spectrogram:\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F7bd5b9731575404474f9816facb9e52f%2Feeg_signal_spectrogram_fp1.png?generation=1710098817394458&alt=media)\n\nThen i plotted the realted spectrogram using parquet file from folder eeg_spectrograms:\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F3709138%2F37a984dcdb133d9ec7972dd6284ca900%2Feeg_spectrograms_parquet.png?generation=1710098957022133&alt=media)\n\nMaybe my question directed more to signal processing people and engineers. Could you please explain to me the difference between these two spectrograms. Do plots look approperiate? Maybe if it is possible you can explain to me the train csv columns from the plots because I don't understand the 50 seconds eeg subsample id column explanation and 10 mintues in spectrogram_sub_id. \n\nThank you in advance",
    "2690856": "https://www.kaggle.com/competitions/hms-harmful-brain-activity-classification/discussion/468010\ngive this a look, it will explain everything!\nAs far as how the kaggle spectrogram were generated, i dont think we have any information on how they preprocessed.",
    "2692312": "thank you for your response"
  },
  "source": "meta"
}