![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F761268%2F72636bbb3d3795616b1980e133b1c5aa%2FScreenshot%202024-08-09%20at%2010.37.55PM.png?generation=1723223289993867&alt=media\")
\n\nSensor FGS1 => Every 0.1 sec, it capture 32x32 image => (135,00 secs x 10) x ( 32x32 )
\nSensor CH0 => Every time step in axis_info => it capture (11,250 sub steps ) x (32 x 356)
\nTotal Time = 13500 secs => (13500x10)/(11250) = ~ approx every 1.2secs we have sensor data from CH0
\n
\n", "messages": [ { "id": 2954371, "postDate": "2024-08-09T16:26:23.707Z", "content": "wavelengths.csv => w_1 ( 0.70µm ) & w_2 - w_283 ( 1.95µm - 3.9µm )
\naxis_info => AIRS-CH0-axis0-h vs FGS1-axis0-h => correlate 100% => from axis 0 to 7.5 scale
\n
\n\nSensor FGS1 => Every 0.1 sec, it capture 32x32 image => (135,00 secs x 10) x ( 32x32 )
\nSensor CH0 => Every time step in axis_info => it capture (11,250 sub steps ) x (32 x 356)
\nTotal Time = 13500 secs => (13500x10)/(11250) = ~ approx every 1.2secs we have sensor data from CH0
\n
\n", "rawMarkdown": "\n\n---\n\n> ## Sensor FGS1 => Every 0.1 sec, it capture 32x32 image => (135,00 secs x 10) x ( 32x32 )\n> ## Sensor CH0 => Every time step in axis_info => it capture (11,250 sub steps ) x (32 x 356)\n> ## Total Time = 13500 secs => (13500x10)/(11250) = ~ approx every 1.2secs we have sensor data from CH0\n\n---\n\n> ### **wavelengths.csv** => w_1 ( 0.70µm ) & w_2 - w_283 ( 1.95µm - 3.9µm )\n> ### axis_info => AIRS-CH0-axis0-h vs FGS1-axis0-h => correlate 100% => from axis 0 to 7.5 scale\n", "votes": 68 }, { "id": 2954534, "postDate": "2024-08-09T19:57:26.657Z", "content": "wavelengths.csv => w_1 ( 0.70µm ) & w_2 - w_283 ( 1.95µm - 3.9µm )
\naxis_info => AIRS-CH0-axis0-h vs FGS1-axis0-h => correlate 100% => from axis 0 to 7.5 scale
\n
Thank you for the very nice diagram!
", "rawMarkdown": "Thank you for the very nice diagram!", "votes": 3, "replies": [ { "id": 2955164, "postDate": "2024-08-10T17:04:09.307Z", "content": "Can you explain what is 32 in CH0 sensors? is it some kind of grid?
", "rawMarkdown": "Can you explain what is 32 in CH0 sensors? is it some kind of grid?", "votes": 2, "replies": [ { "id": 2955182, "postDate": "2024-08-10T17:40:07.700Z", "content": "\n\n@drhabib, I have consistently followed the solutions from your team in previous competitions, particularly those shared on GitHub codes and as journal code (mainly). Your work has been incredibly motivating, and I have learned the best practices from your solutions, and mainly from journal (how to design the architecture with mini-dataset, then improvements one by one) => Github => icecube-journal -> one of my fav solution journey shared in Kaggle, as everyone shared final solution but your team shared journal -> from starting to final solution journey.
\n
=> it will be not relevant this topic post, but journal solutions are so motivate and great learning. could we expect one more journal for this competition too (if participated).
", "rawMarkdown": "> @drhabib, I have consistently followed the solutions from your team in previous competitions, particularly those shared on GitHub codes and as journal code (mainly). Your work has been incredibly motivating, and I have learned the best practices from your solutions, and mainly from journal (how to design the architecture with mini-dataset, then improvements one by one) => [Github => icecube-journal -> one of my fav solution journey shared in Kaggle](https://github.com/DrHB/icecube-journal), as everyone shared final solution but your team shared journal -> from starting to final solution journey.\n\n=> **it will be not relevant this topic post, but journal solutions are so motivate and great learning. could we expect one more journal for this competition too (if participated).** " }, { "id": 2956173, "postDate": "2024-08-11T20:44:23.797Z", "content": "Sure. AIRS-CH0 has 32 pixels along the detector's spatial direction. We discussed this in the following thread: https://www.kaggle.com/competitions/ariel-data-challenge-2024/discussion/524109#2950885
\nIn particular, I provided a more detailed explanation in this response: https://www.kaggle.com/competitions/ariel-data-challenge-2024/discussion/524109#2954425
\nI hope this helps, but please let me know if it doesn't fully answer your question.
", "rawMarkdown": "Sure. AIRS-CH0 has 32 pixels along the detector's spatial direction. We discussed this in the following thread: https://www.kaggle.com/competitions/ariel-data-challenge-2024/discussion/524109#2950885\n\nIn particular, I provided a more detailed explanation in this response: https://www.kaggle.com/competitions/ariel-data-challenge-2024/discussion/524109#2954425\n\nI hope this helps, but please let me know if it doesn't fully answer your question.", "votes": 5 } ] } ] }, { "id": 2976654, "postDate": "2024-09-02T07:43:28.907Z", "content": "Thank you for making things clearer.
\nSo each planet has 11250x32x356 CH0 measurements. Which means each 32x356 image is scanned 11250 times at some interval (or time resolution if you will).
\nWithin those images, the sensor array of 356 elements is scanned 32 times, at some interval. Why are the scans in the middle brighter?
\n/Thanks
", "rawMarkdown": "Thank you for making things clearer.\n\nSo each planet has 11250x32x356 CH0 measurements. Which means each 32x356 image is scanned 11250 times at some interval (or time resolution if you will).\n\nWithin those images, the sensor array of 356 elements is scanned 32 times, at some interval. Why are the scans in the middle brighter?\n\n/Thanks", "votes": 1, "replies": [ { "id": 3019082, "postDate": "2024-10-16T08:44:11.027Z", "content": "it is not scanned 32 times, thats the spatial dimension of the image, it is like taking a picture from your camera. The middle bright bit or strip are the flux coming from the star, or what we called, a PSF, spread across 356 wavelengths
", "rawMarkdown": "it is not scanned 32 times, thats the spatial dimension of the image, it is like taking a picture from your camera. The middle bright bit or strip are the flux coming from the star, or what we called, a PSF, spread across 356 wavelengths", "votes": 1 } ] }, { "id": 2955082, "postDate": "2024-08-10T15:17:20.647Z", "content": "I'm extremely grateful for your discussion, which has been very helpful for me to understand the data.
", "rawMarkdown": "I'm extremely grateful for your discussion, which has been very helpful for me to understand the data." }, { "id": 2963855, "postDate": "2024-08-19T09:44:45.150Z", "content": "@gordonyip @seshurajup I would greatly appreciate if you guys could point me to how the 39th index of CH0 sensor maps to 1.95µm. I am trying to understand the wavelengths measured by AIRS CH0.
\n\n\n\"AIRS-CH0 is the first channel (CH0) of the Ariel InfraRed Spectrometer (AIRS). It is an infrared spectrometer with a sensitivity between 1.95 and 3.90 µm\"
\n
According to the quoted Data section of the Kaggle challenge, I would assume that the 0th index maps to 1.95µm and the last index maps to 3.90 µm but this does not seem to be the case presented in the image above.
\nThanks in advance!
", "rawMarkdown": "@gordonyip @seshurajup I would greatly appreciate if you guys could point me to how the 39th index of CH0 sensor maps to 1.95µm. I am trying to understand the wavelengths measured by AIRS CH0.\n \n>\"AIRS-CH0 is the first channel (CH0) of the Ariel InfraRed Spectrometer (AIRS). It is an infrared spectrometer with a sensitivity between 1.95 and 3.90 µm\"\n\nAccording to the quoted Data section of the Kaggle challenge, I would assume that the 0th index maps to 1.95µm and the last index maps to 3.90 µm but this does not seem to be the case presented in the image above.\n\nThanks in advance! ", "votes": 2, "isDeleted": true, "replies": [ { "id": 2963882, "postDate": "2024-08-19T10:52:41.050Z", "content": "you should be able to find the proper order of the wavelengths in the Axis-info session. Inside you can see how the wavelengths are arranged.
", "rawMarkdown": "you should be able to find the proper order of the wavelengths in the Axis-info session. Inside you can see how the wavelengths are arranged. ", "votes": 2, "replies": [ { "id": 2963903, "postDate": "2024-08-19T11:35:28.037Z", "content": "@gordonyip
\nIn the axis_info.parquet file, could you please explain what the columns are?
\nIs AIRS-CH0-axis2-um the wavelength axis?
\nWhat are FGS1-axis0-h and AIRS-CH0-axis0-h?
np.diff(a['AIRS-CH0-axis0-h'].dropna().values)*3600
alternates between 0.1 and 4.7 while
\nnp.diff(a['FGS1-axis0-h'].dropna().values)*3600
alternates between 0.1 and 0.3.
\nAre these two columns related to the exposure times?
Yes AIRS-CH0-axis2-um this is the wavelength axis,
\nWhat are FGS1-axis0-h and AIRS-CH0-axis0-h?
\nthese are exposure time - am working to update the axis-info file. (which also means the notebook needs updating! thanks for flagging it up)
I'm struggling to understand how the 39th index of the CH0 sensor maps to 1.95µm. According to the axis_info file, the 39th row in the AIRS-CH0-axis2-um column should correspond to a value of 1.95µm, which I would expect to align with w_2 (wavelength 2). However, this doesn't seem to be the case. Could there be a misalignment, or am I misunderstanding the mapping logic here?
Please Help!
", "rawMarkdown": "I'm struggling to understand how the 39th index of the CH0 sensor maps to 1.95µm. According to the `axis_info` file, the 39th row in the `AIRS-CH0-axis2-um` column should correspond to a value of 1.95µm, which I would expect to align with `w_2` (wavelength 2). However, this doesn't seem to be the case. Could there be a misalignment, or am I misunderstanding the mapping logic here?\n\nPlease Help!" } ] }, { "id": 2963907, "postDate": "2024-08-19T11:39:11.220Z", "content": "You can see from the calibration notebook from the hosts that they calibrate the data starting from the column 39 and ending in 321 (python indexing will not include the 321th column).
\nsearch in the notebook:
cut_inf, cut_sup = , \nl = cut_sup - cut_inf\nThank you! great job
", "rawMarkdown": "Thank you! great job", "votes": 1 }, { "id": 2955605, "postDate": "2024-08-11T11:02:09.480Z", "content": "This is really good, thank you.
", "rawMarkdown": "This is really good, thank you.", "votes": 1 } ], "comments": [ { "id": 2954534, "author_name": "Gordon Yip", "author_url": "", "post_date": "2024-08-09T19:57:26.657000", "content": "Thank you for the very nice diagram!
", "votes": 3, "replies": [ { "id": 2955164, "author_name": "DrHB", "author_url": "", "post_date": "2024-08-10T17:04:09.307000", "content": "Can you explain what is 32 in CH0 sensors? is it some kind of grid?
", "votes": 2, "replies": [ { "id": 2955182, "author_name": "SeshuRaju 🧘♂️", "author_url": "", "post_date": "2024-08-10T17:40:07.700000", "content": "\n\n@drhabib, I have consistently followed the solutions from your team in previous competitions, particularly those shared on GitHub codes and as journal code (mainly). Your work has been incredibly motivating, and I have learned the best practices from your solutions, and mainly from journal (how to design the architecture with mini-dataset, then improvements one by one) => Github => icecube-journal -> one of my fav solution journey shared in Kaggle, as everyone shared final solution but your team shared journal -> from starting to final solution journey.
\n
=> it will be not relevant this topic post, but journal solutions are so motivate and great learning. could we expect one more journal for this competition too (if participated).
", "votes": 0, "replies": [] }, { "id": 2956173, "author_name": "Lorenzo Mugnai", "author_url": "", "post_date": "2024-08-11T20:44:23.797000", "content": "Sure. AIRS-CH0 has 32 pixels along the detector's spatial direction. We discussed this in the following thread: https://www.kaggle.com/competitions/ariel-data-challenge-2024/discussion/524109#2950885
\nIn particular, I provided a more detailed explanation in this response: https://www.kaggle.com/competitions/ariel-data-challenge-2024/discussion/524109#2954425
\nI hope this helps, but please let me know if it doesn't fully answer your question.
", "votes": 5, "replies": [] } ] } ] }, { "id": 2976654, "author_name": "CJ ROCKBALL", "author_url": "", "post_date": "2024-09-02T07:43:28.907000", "content": "Thank you for making things clearer.
\nSo each planet has 11250x32x356 CH0 measurements. Which means each 32x356 image is scanned 11250 times at some interval (or time resolution if you will).
\nWithin those images, the sensor array of 356 elements is scanned 32 times, at some interval. Why are the scans in the middle brighter?
\n/Thanks
", "votes": 1, "replies": [ { "id": 3019082, "author_name": "Gordon Yip", "author_url": "", "post_date": "2024-10-16T08:44:11.027000", "content": "it is not scanned 32 times, thats the spatial dimension of the image, it is like taking a picture from your camera. The middle bright bit or strip are the flux coming from the star, or what we called, a PSF, spread across 356 wavelengths
", "votes": 1, "replies": [] } ] }, { "id": 2955082, "author_name": "Xiaolei Lian", "author_url": "", "post_date": "2024-08-10T15:17:20.647000", "content": "I'm extremely grateful for your discussion, which has been very helpful for me to understand the data.
", "votes": 0, "replies": [] }, { "id": 2963855, "author_name": "", "author_url": "", "post_date": "2024-08-19T09:44:45.150000", "content": "@gordonyip @seshurajup I would greatly appreciate if you guys could point me to how the 39th index of CH0 sensor maps to 1.95µm. I am trying to understand the wavelengths measured by AIRS CH0.
\n\n\n\"AIRS-CH0 is the first channel (CH0) of the Ariel InfraRed Spectrometer (AIRS). It is an infrared spectrometer with a sensitivity between 1.95 and 3.90 µm\"
\n
According to the quoted Data section of the Kaggle challenge, I would assume that the 0th index maps to 1.95µm and the last index maps to 3.90 µm but this does not seem to be the case presented in the image above.
\nThanks in advance!
", "votes": 2, "replies": [ { "id": 2963882, "author_name": "Gordon Yip", "author_url": "", "post_date": "2024-08-19T10:52:41.050000", "content": "you should be able to find the proper order of the wavelengths in the Axis-info session. Inside you can see how the wavelengths are arranged.
", "votes": 2, "replies": [ { "id": 2963903, "author_name": "Reza R. Choubeh", "author_url": "", "post_date": "2024-08-19T11:35:28.037000", "content": "@gordonyip
\nIn the axis_info.parquet file, could you please explain what the columns are?
\nIs AIRS-CH0-axis2-um the wavelength axis?
\nWhat are FGS1-axis0-h and AIRS-CH0-axis0-h?
np.diff(a['AIRS-CH0-axis0-h'].dropna().values)*3600
alternates between 0.1 and 4.7 while
\nnp.diff(a['FGS1-axis0-h'].dropna().values)*3600
alternates between 0.1 and 0.3.
\nAre these two columns related to the exposure times?
Yes AIRS-CH0-axis2-um this is the wavelength axis,
\nWhat are FGS1-axis0-h and AIRS-CH0-axis0-h?
\nthese are exposure time - am working to update the axis-info file. (which also means the notebook needs updating! thanks for flagging it up)
I'm struggling to understand how the 39th index of the CH0 sensor maps to 1.95µm. According to the axis_info file, the 39th row in the AIRS-CH0-axis2-um column should correspond to a value of 1.95µm, which I would expect to align with w_2 (wavelength 2). However, this doesn't seem to be the case. Could there be a misalignment, or am I misunderstanding the mapping logic here?
Please Help!
", "votes": 0, "replies": [] } ] }, { "id": 2963907, "author_name": "Reza R. Choubeh", "author_url": "", "post_date": "2024-08-19T11:39:11.220000", "content": "You can see from the calibration notebook from the hosts that they calibrate the data starting from the column 39 and ending in 321 (python indexing will not include the 321th column).
\nsearch in the notebook:
cut_inf, cut_sup = , \nl = cut_sup - cut_inf\nThank you! great job
", "votes": 1, "replies": [] }, { "id": 2955605, "author_name": "Naqi Naqvi", "author_url": "", "post_date": "2024-08-11T11:02:09.480000", "content": "This is really good, thank you.
", "votes": 1, "replies": [] } ], "raw_markdown_by_id": { "2954371": "\n\n---\n\n> ## Sensor FGS1 => Every 0.1 sec, it capture 32x32 image => (135,00 secs x 10) x ( 32x32 )\n> ## Sensor CH0 => Every time step in axis_info => it capture (11,250 sub steps ) x (32 x 356)\n> ## Total Time = 13500 secs => (13500x10)/(11250) = ~ approx every 1.2secs we have sensor data from CH0\n\n---\n\n> ### **wavelengths.csv** => w_1 ( 0.70µm ) & w_2 - w_283 ( 1.95µm - 3.9µm )\n> ### axis_info => AIRS-CH0-axis0-h vs FGS1-axis0-h => correlate 100% => from axis 0 to 7.5 scale\n", "2954534": "Thank you for the very nice diagram!", "2976654": "Thank you for making things clearer.\n\nSo each planet has 11250x32x356 CH0 measurements. Which means each 32x356 image is scanned 11250 times at some interval (or time resolution if you will).\n\nWithin those images, the sensor array of 356 elements is scanned 32 times, at some interval. Why are the scans in the middle brighter?\n\n/Thanks", "2955082": "I'm extremely grateful for your discussion, which has been very helpful for me to understand the data.", "2963855": "@gordonyip @seshurajup I would greatly appreciate if you guys could point me to how the 39th index of CH0 sensor maps to 1.95µm. I am trying to understand the wavelengths measured by AIRS CH0.\n \n>\"AIRS-CH0 is the first channel (CH0) of the Ariel InfraRed Spectrometer (AIRS). It is an infrared spectrometer with a sensitivity between 1.95 and 3.90 µm\"\n\nAccording to the quoted Data section of the Kaggle challenge, I would assume that the 0th index maps to 1.95µm and the last index maps to 3.90 µm but this does not seem to be the case presented in the image above.\n\nThanks in advance! ", "2960215": "", "2973898": "Thank you! great job", "2955605": "This is really good, thank you." } }