Translation of SWAP to AIA

This example shows how to enhance PROBA2/SWAP observations to SDO/AIA observations.

# from itipy.evaluation.util import *
import glob
# from itipy.download.download_proba2 import PROBA2Downloader
# from itipy.download.download_sdo import SDODownloader
# from itipy.data.editor import proba2_norm
# from itipy.translate import *
# from datetime import timedelta, datetime

#base_path = os.getcwd()
#

We provide a publicly available dataset which allows the users to play around with a subset of the data available without downloading the entire database.

This dataset contains .fits files from PROBA2/SWAP, SDO/AIA and Solar Orbiter/EUI (FSI and HRI).

In addition 3 trained models are stored with: 1. PROBA2/SWAP to SDO/AIA 2. Solar Orbiter/EUI FSI to SDO/AIA and 3. SDO/AIA to Solar Orbiter/EUI HRI

to perform the translation.

#download_gcp_bucket('iti-dataset', base_path+'/iti-testset/')
#
# # If you wish to translate different time periods that are not included in the test dataset, we provide download routines for the instruments used for ITI.
# # In order to download data from JSOC (SDO) you need to register your email at `JSOC <http://jsoc.stanford.edu/ajax/register_email.html>`__. If you are registered you can set the environment variable ``JSOC_EMAIL`` to your email address.
# ############################################################################################################################################################################
# # Downloading SWAP data
# swap_downloader = PROBA2Downloader(base_path=base_path+'/swap')
# swap_downloader.downloadDate(date=datetime(2024, 5, 8, 15))
#
# ############################################################################################################################################################################
# # Downloading AIA data
# jsoc_email = os.environ["JSOC_EMAIL"]
#
# sdo_downloader = SDODownloader(base_path=base_path+'/sdo', email=jsoc_email)
# sdo_downloader.downloadDate(date=datetime(2024, 5, 8, 15))
#
# ############################################################################################################################################################################
# # Glob the downloaded files and sort them by date. For SDO we use observations only the 171 Ångström channel.
# swap_files = sorted(glob.glob('swap/*/*.fits', recursive=True))
# sdo_files = sorted(glob.glob('sdo/171/*.fits', recursive=True))
#
# ############################################################################################################################################################################
# # In the next step we load the `.fits`files as SunPy maps. Here we crop the observations to 1.1 solar radii to cover the same Field-of-View (FOV).
# # For SDO/AIA this additionally includes a degradation correction of the instrument.
# swap_data = [getSWAPdata(f) for f in tqdm(swap_files)]
# aia_data = [getAIAdata(f) for f in tqdm(sdo_files)]
#
#
# ############################################################################################################################################################################
# # The translator classes are the core element of the ITI translation. They follow the notation: `InstrumentAToInstrumentB`. We initialize the translation class by giving it the path where the model is stored. We use a patch factor of 2 to save memory.
#
# translator = SWAPToAIA(model_name=base_path+'/iti-testset/models/swap_to_aia_v0_4.pt', patch_factor=2)
#
# ############################################################################################################################################################################
#
# # The result of the ITI translation is a SunPy map that stores all necessary coordinate information.
#
# iti_maps = list(translator.translate(swap_files))
#
# ############################################################################################################################################################################
# # Now that we have the translated maps, we can plot them side by side with the original data and the ground truth.
#
# fig, axs = plt.subplots(1, 3, subplot_kw={'projection': aia_data[0]}, figsize=(40, 20), dpi=100)
# swap_data[0].plot(axes=axs[0], norm=proba2_norm[174])
# aia_data[0].plot(axes=axs[1], norm=sdo_norms[171])
# iti_maps[0].plot(axes=axs[2], norm=sdo_norms[171])
# axs[0].set_title('Original-SWAP', fontsize=30)
# axs[1].set_title('Ground Truth-AIA', fontsize=30)
# axs[2].set_title('ITI', fontsize=30)
# plt.show()