\n
![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F11026443%2F203d6b0853942ec2dee98cc5a01e852e%2F1eb42e190da8d75926177426030536b.png?generation=1674705813862160&alt=media\")
\nI dont know what dose it mean? Should I use the mean value of these pulse record by one sensor?or sum up the values? or there is any way to understant this situation？i dont know how convert such data to a dataframe for training
My initial thought was single sensor record ONLY one pulse for single event. However, when checking my dataset i find that in single event ,one sensor will record more than one pulse.
\n
\nI dont know what dose it mean? Should I use the mean value of these pulse record by one sensor?or sum up the values? or there is any way to understant this situation？i dont know how convert such data to a dataframe for training
My initial thought was single sensor record ONLY one pulse for single event. However, when checking my dataset i find that in single event ,one sensor will record more than one pulse.
\n
\nI dont know what dose it mean? Should I use the mean value of these pulse record by one sensor?or sum up the values? or there is any way to understant this situation？i dont know how convert such data to a dataframe for training
why is the time becoming longer
", "rawMarkdown": "why is the time becoming longer", "votes": null }, { "id": "2116003", "postDate": "01/26/2023 06:00:31", "content": "This evening reading lots of the documentation, etc it was noted that multiple pulses do occur in events. I think the reported value used was the sum. IF your using some type of weighted model than you probably should do a sum for the sensor per event.
", "rawMarkdown": "This evening reading lots of the documentation, etc it was noted that multiple pulses do occur in events. I think the reported value used was the sum. IF your using some type of weighted model than you probably should do a sum for the sensor per event.", "votes": null }, { "id": "2116049", "postDate": "01/26/2023 06:45:10", "content": "thanks a lot
", "rawMarkdown": "thanks a lot", "votes": null }, { "id": "2116326", "postDate": "01/26/2023 12:22:50", "content": "What reported value? If you mean the charge, then yes, it is summed if the pulses are very close in space and time. But we get it already summed; we do not need to do it ourselves. If we see several pulses on the same sensor at different times (in the same event), there is no reason to sum them. They can be (and probably are?) the results of several different leptons coming from different angles, especially considering the time difference. The neutrino travels close to the speed of light, so it goes from one end of the detector to the other in what, ~3500[ns]? So in the table above, the last pulse occurs long after the neutrino has passed.
", "rawMarkdown": "What reported value? If you mean the charge, then yes, it is summed if the pulses are very close in space and time. But we get it already summed; we do not need to do it ourselves. If we see several pulses on the same sensor at different times (in the same event), there is no reason to sum them. They can be (and probably are?) the results of several different leptons coming from different angles, especially considering the time difference. The neutrino travels close to the speed of light, so it goes from one end of the detector to the other in what, ~3500[ns]? So in the table above, the last pulse occurs long after the neutrino has passed.", "votes": null }, { "id": "2116570", "postDate": "01/26/2023 15:10:21", "content": "As the neutrino travels, it interacts with the ice in multiple places to produce Cherenkov radiation, which is detected by the sensors as an event. Those multiple places have different distances from each sensor, giving rise to multiple detection pulses.
", "rawMarkdown": "As the neutrino travels, it interacts with the ice in multiple places to produce Cherenkov radiation, which is detected by the sensors as an event. Those multiple places have different distances from each sensor, giving rise to multiple detection pulses.", "votes": null }, { "id": "2116975", "postDate": "01/26/2023 21:04:34", "content": "I should have said the neutrino interacts with the ice to produce a charged particle, which generally follows the same path, and that charged particle produces Cherenkov radiation. So I am not sure why there are multiple pulses 😬
", "rawMarkdown": "I should have said the neutrino interacts with the ice to produce a charged particle, which *generally* follows the same path, and that charged particle produces Cherenkov radiation. So I am not sure why there are multiple pulses 😬", "votes": null }, { "id": "2118461", "postDate": "01/28/2023 03:13:56", "content": "Guess it depends on how your using the sensor data - if you use 12 values (in the posted example) or chose to use one value for the sensor/event. Bubble plot by sensor location going to look like crap if you try to use 12 values :)
\nSum them - count them - mean …. the list goes on - intended use drives the stat.
", "rawMarkdown": "Guess it depends on how your using the sensor data - if you use 12 values (in the posted example) or chose to use one value for the sensor/event. Bubble plot by sensor location going to look like crap if you try to use 12 values :)\n\nSum them - count them - mean .... the list goes on - intended use drives the stat.", "votes": null }, { "id": "2118685", "postDate": "01/28/2023 07:31:57", "content": "It’s worth pointing out that the charged particle will on average initially follow the same path, but there will be a probability cone associated with this.
\nAlso since the particle is charged, it will now interact with the ice and will change directions. If it’s a really high energy particle, it can produce more particles in secondary and tertiary collisions (similar to the hadronic showers in the LHC)
\nMy point being, the neutrino event will rarely be a laser beam of blue Cherenkov light. It’s more like a fireworks display
", "rawMarkdown": "It’s worth pointing out that the charged particle will *on average* initially follow the same path, but there will be a probability cone associated with this.\n\nAlso since the particle is charged, it will now interact with the ice and will change directions. If it’s a really high energy particle, it can produce more particles in secondary and tertiary collisions (similar to the hadronic showers in the LHC)\n\nMy point being, the neutrino event will rarely be a laser beam of blue Cherenkov light. It’s more like a fireworks display", "votes": null }, { "id": "2119436", "postDate": "01/28/2023 19:04:39", "content": "This paper helped me further understand what's going on, though I admit it's pretty hard to understand itself without a little high level understanding: https://arxiv.org/pdf/1307.3795.pdf
\nBasically, I view it as a continuous and/or repeated cloud of photons being released omni-directionally over time. Not to mention noise from other high-energy events besides the neutrino caused event. With light absorption, energy decay and many other effects making it sparser and less likely to reach at longer distances.
\nAside: Even Cherenkov radiation, if I understand it correctly, is still a special case of constant omnidirectional emission. From wikipedia: normally \"The corresponding emitted wavefronts may be bunched up but they do not coincide or cross and there are no interference effects to worry about.\", but in the Cherenkov radiation case \"This results in overlapping waveforms (as in the animation) and constructive interference leads to an observed cone-like light signal at a characteristic angle: Cherenkov light.\" In other words, the Cherenkov difference is that the clouds overlap and thus travel in a conic shape rather than purely spherical.
\nAnyways, if point A is near the travel path V that is repeatedly sending out omni-directional photons, then it is logical that it MIGHT receive numerous photons arriving at separate times, mostly clustered, with the possibility of even getting some stragglers sent 'backwards' some time after the neutrino event has passed.
", "rawMarkdown": "This paper helped me further understand what's going on, though I admit it's pretty hard to understand itself without a little high level understanding: https://arxiv.org/pdf/1307.3795.pdf\n\nBasically, I view it as a continuous and/or repeated cloud of photons being released omni-directionally over time. Not to mention noise from other high-energy events besides the neutrino caused event. With light absorption, energy decay and many other effects making it sparser and less likely to reach at longer distances.\n\nAside: Even Cherenkov radiation, if I understand it correctly, is still a special case of constant omnidirectional emission. [From wikipedia](https://en.wikipedia.org/wiki/Cherenkov_radiation): normally \"The corresponding emitted wavefronts may be bunched up but they do not coincide or cross and there are no interference effects to worry about.\", but in the Cherenkov radiation case \"This results in overlapping waveforms (as in the animation) and constructive interference leads to an observed cone-like light signal at a characteristic angle: Cherenkov light.\" In other words, the Cherenkov difference is that the clouds overlap and thus travel in a conic shape rather than purely spherical.\n\nAnyways, if point A is near the travel path V that is repeatedly sending out omni-directional photons, then it is logical that it MIGHT receive numerous photons arriving at separate times, mostly clustered, with the possibility of even getting some stragglers sent 'backwards' some time after the neutrino event has passed.", "votes": null } ], "comments": [ { "id": 2115913, "author_name": "enochhe", "author_url": "", "post_date": "01/26/2023 04:20:51", "content": "why is the time becoming longer
", "votes": null, "replies": [] }, { "id": 2116003, "author_name": "pcjimmmy", "author_url": "", "post_date": "01/26/2023 06:00:31", "content": "This evening reading lots of the documentation, etc it was noted that multiple pulses do occur in events. I think the reported value used was the sum. IF your using some type of weighted model than you probably should do a sum for the sensor per event.
", "votes": null, "replies": [ { "id": 2116049, "author_name": "enochhe", "author_url": "", "post_date": "01/26/2023 06:45:10", "content": "thanks a lot
", "votes": null, "replies": [] }, { "id": 2116326, "author_name": "shlomoron", "author_url": "", "post_date": "01/26/2023 12:22:50", "content": "What reported value? If you mean the charge, then yes, it is summed if the pulses are very close in space and time. But we get it already summed; we do not need to do it ourselves. If we see several pulses on the same sensor at different times (in the same event), there is no reason to sum them. They can be (and probably are?) the results of several different leptons coming from different angles, especially considering the time difference. The neutrino travels close to the speed of light, so it goes from one end of the detector to the other in what, ~3500[ns]? So in the table above, the last pulse occurs long after the neutrino has passed.
", "votes": null, "replies": [ { "id": 2118461, "author_name": "pcjimmmy", "author_url": "", "post_date": "01/28/2023 03:13:56", "content": "Guess it depends on how your using the sensor data - if you use 12 values (in the posted example) or chose to use one value for the sensor/event. Bubble plot by sensor location going to look like crap if you try to use 12 values :)
\nSum them - count them - mean …. the list goes on - intended use drives the stat.
", "votes": null, "replies": [] } ] } ] }, { "id": 2116570, "author_name": "solverworld", "author_url": "", "post_date": "01/26/2023 15:10:21", "content": "As the neutrino travels, it interacts with the ice in multiple places to produce Cherenkov radiation, which is detected by the sensors as an event. Those multiple places have different distances from each sensor, giving rise to multiple detection pulses.
", "votes": null, "replies": [ { "id": 2116975, "author_name": "solverworld", "author_url": "", "post_date": "01/26/2023 21:04:34", "content": "I should have said the neutrino interacts with the ice to produce a charged particle, which generally follows the same path, and that charged particle produces Cherenkov radiation. So I am not sure why there are multiple pulses 😬
", "votes": null, "replies": [ { "id": 2118685, "author_name": "anjum48", "author_url": "", "post_date": "01/28/2023 07:31:57", "content": "It’s worth pointing out that the charged particle will on average initially follow the same path, but there will be a probability cone associated with this.
\nAlso since the particle is charged, it will now interact with the ice and will change directions. If it’s a really high energy particle, it can produce more particles in secondary and tertiary collisions (similar to the hadronic showers in the LHC)
\nMy point being, the neutrino event will rarely be a laser beam of blue Cherenkov light. It’s more like a fireworks display
", "votes": null, "replies": [] } ] } ] }, { "id": 2119436, "author_name": "roberthatch", "author_url": "", "post_date": "01/28/2023 19:04:39", "content": "This paper helped me further understand what's going on, though I admit it's pretty hard to understand itself without a little high level understanding: https://arxiv.org/pdf/1307.3795.pdf
\nBasically, I view it as a continuous and/or repeated cloud of photons being released omni-directionally over time. Not to mention noise from other high-energy events besides the neutrino caused event. With light absorption, energy decay and many other effects making it sparser and less likely to reach at longer distances.
\nAside: Even Cherenkov radiation, if I understand it correctly, is still a special case of constant omnidirectional emission. From wikipedia: normally \"The corresponding emitted wavefronts may be bunched up but they do not coincide or cross and there are no interference effects to worry about.\", but in the Cherenkov radiation case \"This results in overlapping waveforms (as in the animation) and constructive interference leads to an observed cone-like light signal at a characteristic angle: Cherenkov light.\" In other words, the Cherenkov difference is that the clouds overlap and thus travel in a conic shape rather than purely spherical.
\nAnyways, if point A is near the travel path V that is repeatedly sending out omni-directional photons, then it is logical that it MIGHT receive numerous photons arriving at separate times, mostly clustered, with the possibility of even getting some stragglers sent 'backwards' some time after the neutrino event has passed.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "2115899": "My initial thought was single sensor record ONLY one pulse for single event. However, when checking my dataset i find that in single event ,one sensor will record more than one pulse. \n\nI dont know what dose it mean? Should I use the mean value of these pulse record by one sensor?or sum up the values? or there is any way to understant this situation？i dont know how convert such data to a dataframe for training", "2115913": "why is the time becoming longer", "2116003": "This evening reading lots of the documentation, etc it was noted that multiple pulses do occur in events. I think the reported value used was the sum. IF your using some type of weighted model than you probably should do a sum for the sensor per event.", "2116049": "thanks a lot", "2116326": "What reported value? If you mean the charge, then yes, it is summed if the pulses are very close in space and time. But we get it already summed; we do not need to do it ourselves. If we see several pulses on the same sensor at different times (in the same event), there is no reason to sum them. They can be (and probably are?) the results of several different leptons coming from different angles, especially considering the time difference. The neutrino travels close to the speed of light, so it goes from one end of the detector to the other in what, ~3500[ns]? So in the table above, the last pulse occurs long after the neutrino has passed.", "2116570": "As the neutrino travels, it interacts with the ice in multiple places to produce Cherenkov radiation, which is detected by the sensors as an event. Those multiple places have different distances from each sensor, giving rise to multiple detection pulses.", "2116975": "I should have said the neutrino interacts with the ice to produce a charged particle, which *generally* follows the same path, and that charged particle produces Cherenkov radiation. So I am not sure why there are multiple pulses 😬", "2118461": "Guess it depends on how your using the sensor data - if you use 12 values (in the posted example) or chose to use one value for the sensor/event. Bubble plot by sensor location going to look like crap if you try to use 12 values :)\n\nSum them - count them - mean .... the list goes on - intended use drives the stat.", "2118685": "It’s worth pointing out that the charged particle will *on average* initially follow the same path, but there will be a probability cone associated with this.\n\nAlso since the particle is charged, it will now interact with the ice and will change directions. If it’s a really high energy particle, it can produce more particles in secondary and tertiary collisions (similar to the hadronic showers in the LHC)\n\nMy point being, the neutrino event will rarely be a laser beam of blue Cherenkov light. It’s more like a fireworks display", "2119436": "This paper helped me further understand what's going on, though I admit it's pretty hard to understand itself without a little high level understanding: https://arxiv.org/pdf/1307.3795.pdf\n\nBasically, I view it as a continuous and/or repeated cloud of photons being released omni-directionally over time. Not to mention noise from other high-energy events besides the neutrino caused event. With light absorption, energy decay and many other effects making it sparser and less likely to reach at longer distances.\n\nAside: Even Cherenkov radiation, if I understand it correctly, is still a special case of constant omnidirectional emission. [From wikipedia](https://en.wikipedia.org/wiki/Cherenkov_radiation): normally \"The corresponding emitted wavefronts may be bunched up but they do not coincide or cross and there are no interference effects to worry about.\", but in the Cherenkov radiation case \"This results in overlapping waveforms (as in the animation) and constructive interference leads to an observed cone-like light signal at a characteristic angle: Cherenkov light.\" In other words, the Cherenkov difference is that the clouds overlap and thus travel in a conic shape rather than purely spherical.\n\nAnyways, if point A is near the travel path V that is repeatedly sending out omni-directional photons, then it is logical that it MIGHT receive numerous photons arriving at separate times, mostly clustered, with the possibility of even getting some stragglers sent 'backwards' some time after the neutrino event has passed." }, "source": "meta" }