![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1455075%2Fd31435e07a93dbd32a658f3f2df3afde%2FVerticalSignals.png?generation=1667017760886819&alt=media\")
Observing the test data, I found what It looks \"signals\" from top center to bottom center, leaving \"dark sides\" in the spectrograms. How we must consider these cases? I have though that \"dark side\" is the gap when the detector was turned off, but still there are \"signals\" or artifacts in that areas. I prefer don't do any assumption, and ask it better. Thanks in advance.
\nThis there are some graphic examples of those cases:
\n
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F1455075%2F7fceeee6e39f9f89df5c76a4f6301873%2FVerticalSignals2.png?generation=1667019158810680&alt=media\")
Cheers :)
", "messages": [ { "id": 2008445, "postDate": "2022-10-29T04:32:40.413Z", "content": "Observing the test data, I found what It looks \"signals\" from top center to bottom center, leaving \"dark sides\" in the spectrograms. How we must consider these cases? I have though that \"dark side\" is the gap when the detector was turned off, but still there are \"signals\" or artifacts in that areas. I prefer don't do any assumption, and ask it better. Thanks in advance.
\nThis there are some graphic examples of those cases:
\n
Cheers :)
", "rawMarkdown": "Observing the test data, I found what It looks \"signals\" from top center to bottom center, leaving \"dark sides\" in the spectrograms. How we must consider these cases? I have though that \"dark side\" is the gap when the detector was turned off, but still there are \"signals\" or artifacts in that areas. I prefer don't do any assumption, and ask it better. Thanks in advance.\n\nThis there are some graphic examples of those cases:\n\n\n\n\n\nCheers :)", "votes": 6 }, { "id": 2015807, "postDate": "2022-11-03T14:58:29.073Z", "content": "The vertical stripes indicates that the sqrt(SN) is not stationary, i.e., it varies over time. That applies for the top graph.
\nThe bottom graph is a mistery. My first assumption is that it is an square signal. The frequency profile (for each timestep) looks like a sinc (in the FFT domain), this usually corresponds to an square signal in time domain.
\nBut this is just an assumption.
Thanks for your answer. I didn't know the way it looks a not stationary signal in the frequency domain, it's really amazing. I'm only worked with invariant in time signals befote with QFT. All these variations are new for me. I'm agreed with you with the bottom-right graph. It's an absolute mistery. But I don't think that It corresponds to a star signal.
", "rawMarkdown": "Thanks for your answer. I didn't know the way it looks a not stationary signal in the frequency domain, it's really amazing. I'm only worked with invariant in time signals befote with QFT. All these variations are new for me. I'm agreed with you with the bottom-right graph. It's an absolute mistery. But I don't think that It corresponds to a star signal." } ] } ], "comments": [ { "id": 2015807, "author_name": "Cristo JV", "author_url": "", "post_date": "2022-11-03T14:58:29.073000", "content": "The vertical stripes indicates that the sqrt(SN) is not stationary, i.e., it varies over time. That applies for the top graph.
\nThe bottom graph is a mistery. My first assumption is that it is an square signal. The frequency profile (for each timestep) looks like a sinc (in the FFT domain), this usually corresponds to an square signal in time domain.
\nBut this is just an assumption.
Thanks for your answer. I didn't know the way it looks a not stationary signal in the frequency domain, it's really amazing. I'm only worked with invariant in time signals befote with QFT. All these variations are new for me. I'm agreed with you with the bottom-right graph. It's an absolute mistery. But I don't think that It corresponds to a star signal.
", "votes": 0, "replies": [] } ] } ], "raw_markdown_by_id": { "2008445": "Observing the test data, I found what It looks \"signals\" from top center to bottom center, leaving \"dark sides\" in the spectrograms. How we must consider these cases? I have though that \"dark side\" is the gap when the detector was turned off, but still there are \"signals\" or artifacts in that areas. I prefer don't do any assumption, and ask it better. Thanks in advance.\n\nThis there are some graphic examples of those cases:\n\n\n\n\n\nCheers :)", "2015807": "The vertical stripes indicates that the sqrt(SN) is not stationary, i.e., it varies over time. That applies for the top graph.\n\nThe bottom graph is a mistery. My first assumption is that it is an square signal. The frequency profile (for each timestep) looks like a sinc (in the FFT domain), this usually corresponds to an square signal in time domain.\nBut this is just an assumption." } }