![\"GW170814\"](\"https://www.ligo.org/science/Publication-GW170814/images/figure2.png\")
Not sure if I missed something, but the magnitude of the wave from third detector (virgo i assume) looks noticeably lower than LIGO ones, is it because Virgo has lower noise level (but the true gw signal would still share same magnitude across all three detectors) or its due to some other reasons?
\nalso, what exactly does the y-axis represents? from what i can gather from gw wiki page, seems like it's the stretch caused by gw measured from inteferometer?
", "messages": [ { "id": "1378989", "postDate": "07/07/2021 02:51:05", "content": "Not sure if I missed something, but the magnitude of the wave from third detector (virgo i assume) looks noticeably lower than LIGO ones, is it because Virgo has lower noise level (but the true gw signal would still share same magnitude across all three detectors) or its due to some other reasons?
\nalso, what exactly does the y-axis represents? from what i can gather from gw wiki page, seems like it's the stretch caused by gw measured from inteferometer?
", "rawMarkdown": "Not sure if I missed something, but the magnitude of the wave from third detector (virgo i assume) looks noticeably lower than LIGO ones, is it because Virgo has lower noise level (but the true gw signal would still share same magnitude across all three detectors) or its due to some other reasons?\n\nalso, what exactly does the y-axis represents? from what i can gather from gw wiki page, seems like it's the stretch caused by gw measured from inteferometer?", "votes": null }, { "id": "1379169", "postDate": "07/07/2021 06:32:29", "content": "According to your idea, I will do two experiments, cancel or leave only the third detector to see the difference and do a comparative experiment.
", "rawMarkdown": "According to your idea, I will do two experiments, cancel or leave only the third detector to see the difference and do a comparative experiment.", "votes": null }, { "id": "1379275", "postDate": "07/07/2021 08:17:43", "content": "You're correct about the differences in the data for the third detector. Virgo has a slightly different design and, therefore, sensitivity to the two LIGO detectors. We accounted for this when simulating the data and it results in the differences you're seeing.
\nFor the data with signals, the gravitational wave will be present in all three detectors, but because of the different sensitivities and other factors, like the position in the sky, its amplitude in each detector will be different. Here's an example from one of our papers that shows the difference in amplitude (see the bottom three panels).
\n
As for the y-axis, it represents a dimensionless quantity we call the strain. As you mentioned, it is a measure of the stretching and squeezing caused in the detectors by a gravitational wave as it passes through. Specifically, it quantifies how much the length of the interferometer has changed compared to its original length.
\nHope this helps.
", "rawMarkdown": "You're correct about the differences in the data for the third detector. Virgo has a slightly different design and, therefore, sensitivity to the two LIGO detectors. We accounted for this when simulating the data and it results in the differences you're seeing. \n\nFor the data with signals, the gravitational wave will be present in all three detectors, but because of the different sensitivities and other factors, like the position in the sky, its amplitude in each detector will be different. Here's an example from one of our papers that shows the difference in amplitude (see the bottom three panels).\n\n\n\nAs for the y-axis, it represents a dimensionless quantity we call the strain. As you mentioned, it is a measure of the stretching and squeezing caused in the detectors by a gravitational wave as it passes through. Specifically, it quantifies how much the length of the interferometer has changed compared to its original length.\n\nHope this helps.", "votes": null }, { "id": "1379306", "postDate": "07/07/2021 08:46:15", "content": "I had the same question, thanks for asking it.
", "rawMarkdown": "I had the same question, thanks for asking it.", "votes": null }, { "id": "1379462", "postDate": "07/07/2021 11:51:53", "content": "Hey, have you tried that yet? I got 0.82 AUC by using only the first 2 detectors in Mel spectrogram images and the third detector got about 0.67 AUC but using all three together gave a 0.83 AUC so for me using all three detectors seem to give a better AUC score
", "rawMarkdown": "Hey, have you tried that yet? I got 0.82 AUC by using only the first 2 detectors in Mel spectrogram images and the third detector got about 0.67 AUC but using all three together gave a 0.83 AUC so for me using all three detectors seem to give a better AUC score", "votes": null }, { "id": "1381225", "postDate": "07/08/2021 18:27:55", "content": "Using the same model, I move 5 folds CV from 0.866 using all data to 0.844 using only the first 2 detectors. I will not submit the latter to check LB…
", "rawMarkdown": "Using the same model, I move 5 folds CV from 0.866 using all data to 0.844 using only the first 2 detectors. I will not submit the latter to check LB...", "votes": null }, { "id": "1381740", "postDate": "07/09/2021 07:40:12", "content": "How are you training the 3 waves together?
\nAre you stacking the 3 spectrograms to make a 3 channel image and passing them to a pretrained model?
I plotted box plot for mean of 3 Detectors signals for each signal id and the data looks really quite different
\n
There is only one way to answer questions like that: try your ideas and see which ones work.
", "rawMarkdown": "There is only one way to answer questions like that: try your ideas and see which ones work.", "votes": null }, { "id": "1382551", "postDate": "07/10/2021 03:23:11", "content": "@mrinath I've tired stacking a few ways. First I normalised each time series:
\nHey,
\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.
\nThank you all,
\nRegards,
\nChris
According to your idea, I will do two experiments, cancel or leave only the third detector to see the difference and do a comparative experiment.
", "votes": null, "replies": [ { "id": 1379462, "author_name": "mrigendraagrawal", "author_url": "", "post_date": "07/07/2021 11:51:53", "content": "Hey, have you tried that yet? I got 0.82 AUC by using only the first 2 detectors in Mel spectrogram images and the third detector got about 0.67 AUC but using all three together gave a 0.83 AUC so for me using all three detectors seem to give a better AUC score
", "votes": null, "replies": [] }, { "id": 1381225, "author_name": "cpmpml", "author_url": "", "post_date": "07/08/2021 18:27:55", "content": "Using the same model, I move 5 folds CV from 0.866 using all data to 0.844 using only the first 2 detectors. I will not submit the latter to check LB…
", "votes": null, "replies": [] }, { "id": 1381740, "author_name": "mrinath", "author_url": "", "post_date": "07/09/2021 07:40:12", "content": "How are you training the 3 waves together?
\nAre you stacking the 3 spectrograms to make a 3 channel image and passing them to a pretrained model?
There is only one way to answer questions like that: try your ideas and see which ones work.
", "votes": null, "replies": [] }, { "id": 1382551, "author_name": "kevinmcisaac", "author_url": "", "post_date": "07/10/2021 03:23:11", "content": "@mrinath I've tired stacking a few ways. First I normalised each time series:
\nYou're correct about the differences in the data for the third detector. Virgo has a slightly different design and, therefore, sensitivity to the two LIGO detectors. We accounted for this when simulating the data and it results in the differences you're seeing.
\nFor the data with signals, the gravitational wave will be present in all three detectors, but because of the different sensitivities and other factors, like the position in the sky, its amplitude in each detector will be different. Here's an example from one of our papers that shows the difference in amplitude (see the bottom three panels).
\n
As for the y-axis, it represents a dimensionless quantity we call the strain. As you mentioned, it is a measure of the stretching and squeezing caused in the detectors by a gravitational wave as it passes through. Specifically, it quantifies how much the length of the interferometer has changed compared to its original length.
\nHope this helps.
", "votes": null, "replies": [] }, { "id": 1379306, "author_name": "cpmpml", "author_url": "", "post_date": "07/07/2021 08:46:15", "content": "I had the same question, thanks for asking it.
", "votes": null, "replies": [] }, { "id": 1381791, "author_name": "sayedathar11", "author_url": "", "post_date": "07/09/2021 08:14:02", "content": "I plotted box plot for mean of 3 Detectors signals for each signal id and the data looks really quite different
\n
Hey,
\nThank you all for taking part in our competition. The participation has been overwhelmingly positive. We are currently conducting a survey to gauge the demographic and outreach achieved. Kindly spare 2min and fill in this survey https://forms.gle/QP9L16niPexozyhu5.
\nThank you all,
\nRegards,
\nChris