Simple Radynversion Example

The following illustrates configuring and using Radynversion with some small SST/CRISP images. To load this data we use the crispy package.

from crispy import CRISPSequence
from crispy.utils import CRISP_sequence_constructor
from matplotlib import pyplot as plt
from smug.radynversion_adapter import RadynversionAdapter
from smug.radynversion_model import model_params, pretrained_radynversion

Plot the Halpha intensity map (see crispy’s documentation for how this is used.).

files = [
    "../tests/mini_crisp_l2_20140906_152724_6563_r00459.fits",
    "../tests/mini_crisp_l2_20140906_152724_8542_r00459.fits",
]
ims = CRISPSequence(CRISP_sequence_constructor(files))
ims.list[0][7].intensity_map()
plt.show()

/opt/hostedtoolcache/Python/3.9.13/x64/lib/python3.9/site-packages/astropy/wcs/wcs.py:725: FITSFixedWarning: 'datfix' made the change 'Set MJD-AVG to 56906.705521 from DATE-AVG'.
  warnings.warn(
/opt/hostedtoolcache/Python/3.9.13/x64/lib/python3.9/site-packages/astropy/wcs/wcs.py:725: FITSFixedWarning: 'datfix' made the change 'Set MJD-AVG to 56906.705580 from DATE-AVG'.
  warnings.warn(
/opt/hostedtoolcache/Python/3.9.13/x64/lib/python3.9/site-packages/astropy/wcs/wcsapi/fitswcs.py:493: AstropyUserWarning: target cannot be converted to ICRS, so will not be set on SpectralCoord
  warnings.warn('target cannot be converted to ICRS, so will '

Configure model, and Adapter class containing utilities to run the model and correctly transform the parameters.

radynversion_version = "1.1.1"
ad = RadynversionAdapter(
    model=pretrained_radynversion(version=radynversion_version),
    **model_params[radynversion_version]
)

Downloading: "https://www.astro.gla.ac.uk/users/USER-MANAGED/solar_model_weights/Radynversion_1.1.1.pth.tar" to /home/runner/.cache/torch/hub/checkpoints/Radynversion_1.1.1.pth.tar

  0%|          | 0.00/45.2M [00:00<?, ?B/s]
  0%|          | 32.0k/45.2M [00:00<02:48, 280kB/s]
  0%|          | 64.0k/45.2M [00:00<02:49, 279kB/s]
  0%|          | 96.0k/45.2M [00:00<02:49, 279kB/s]
  0%|          | 144k/45.2M [00:00<02:22, 333kB/s]
  0%|          | 208k/45.2M [00:00<01:54, 413kB/s]
  1%|          | 288k/45.2M [00:00<01:32, 508kB/s]
  1%|          | 368k/45.2M [00:00<01:22, 569kB/s]
  1%|1         | 464k/45.2M [00:00<01:11, 652kB/s]
  1%|1         | 576k/45.2M [00:01<01:02, 752kB/s]
  1%|1         | 688k/45.2M [00:01<00:57, 819kB/s]
  2%|1         | 848k/45.2M [00:01<00:46, 993kB/s]
  2%|2         | 0.98M/45.2M [00:01<00:41, 1.11MB/s]
  3%|2         | 1.19M/45.2M [00:01<00:35, 1.32MB/s]
  3%|3         | 1.39M/45.2M [00:01<00:31, 1.46MB/s]
  4%|3         | 1.59M/45.2M [00:01<00:29, 1.57MB/s]
  4%|4         | 1.86M/45.2M [00:01<00:25, 1.80MB/s]
  5%|4         | 2.17M/45.2M [00:02<00:21, 2.10MB/s]
  6%|5         | 2.61M/45.2M [00:02<00:16, 2.63MB/s]
  7%|6         | 3.08M/45.2M [00:02<00:14, 3.09MB/s]
  8%|8         | 3.72M/45.2M [00:02<00:11, 3.87MB/s]
 10%|9         | 4.50M/45.2M [00:02<00:08, 4.79MB/s]
 12%|#2        | 5.56M/45.2M [00:02<00:06, 6.18MB/s]
 15%|#5        | 6.91M/45.2M [00:02<00:05, 7.91MB/s]
 19%|#9        | 8.64M/45.2M [00:02<00:03, 10.2MB/s]
 23%|##2       | 10.4M/45.2M [00:02<00:03, 11.7MB/s]
 27%|##6       | 12.1M/45.2M [00:03<00:02, 12.9MB/s]
 31%|###       | 13.9M/45.2M [00:03<00:02, 13.7MB/s]
 35%|###4      | 15.6M/45.2M [00:03<00:02, 14.2MB/s]
 38%|###8      | 17.4M/45.2M [00:03<00:01, 14.6MB/s]
 42%|####2     | 19.1M/45.2M [00:03<00:01, 15.0MB/s]
 46%|####6     | 20.9M/45.2M [00:03<00:01, 15.2MB/s]
 50%|#####     | 22.7M/45.2M [00:03<00:01, 15.3MB/s]
 54%|#####3    | 24.4M/45.2M [00:03<00:01, 15.4MB/s]
 58%|#####7    | 26.2M/45.2M [00:04<00:01, 15.4MB/s]
 62%|######1   | 27.9M/45.2M [00:04<00:01, 15.4MB/s]
 66%|######5   | 29.7M/45.2M [00:04<00:01, 15.5MB/s]
 70%|######9   | 31.5M/45.2M [00:04<00:00, 15.5MB/s]
 73%|#######3  | 33.2M/45.2M [00:04<00:00, 15.6MB/s]
 77%|#######7  | 35.0M/45.2M [00:04<00:00, 15.5MB/s]
 81%|########1 | 36.7M/45.2M [00:04<00:00, 15.5MB/s]
 85%|########5 | 38.5M/45.2M [00:04<00:00, 15.6MB/s]
 89%|########8 | 40.2M/45.2M [00:04<00:00, 15.6MB/s]
 93%|#########2| 42.0M/45.2M [00:05<00:00, 15.6MB/s]
 97%|#########6| 43.8M/45.2M [00:05<00:00, 15.6MB/s]
100%|##########| 45.2M/45.2M [00:05<00:00, 9.03MB/s]

Run the inversion for a small slice of the data.

inv = ad.invert_dual_cubes(ims[:, :1, :3])

  0%|          | 0/2 [00:00<?, ?it/s]
 50%|#####     | 1/2 [00:04<00:04,  4.78s/it]
100%|##########| 2/2 [00:07<00:00,  3.37s/it]
100%|##########| 2/2 [00:07<00:00,  3.59s/it]

Display the inversion results for a pixel

inv[:, 0, 1].plot_params(eb=True)
plt.show()