Adding source for Nançay Decameter Array (NDA) observations

[Amityush-lgtm]

PR Description

Fixes #226
This PR adds support for reading and interacting with solar radio burst data from the Nançay Decameter Array (NDA) in radiospectra.

Changes:

• Added an NDASpectrogram class for handling NDA observations.
• Added support for recognizing and loading NDA FITS files.
• Extracted and attached relevant metadata such as frequency information and observatory details.
• Registered the NDA spectrogram source so it works with the existing spectrogram factory.

AI Assistance Disclosure

AI tools were used for:

• Code generation (e.g., when writing an implementation or fixing a bug)
• Test/benchmark generation
• Documentation (including examples)
• Research and understanding
• No AI tools were used

Regardless of AI use, the human contributor remains fully responsible for correctness, design choices, licensing compatibility, and long-term maintainability.

Working on the test, will push the tests in a while.

[samaloney]

Does this run for you locally?

Looks good just need to make sure the generic fits reader work with the NDA file and if not add a new branch

radiospectra/radiospectra/spectrogram/spectrogram_factory.py

Line 593 in a31141d

def _read_fits(file):

[Amityush-lgtm]

Does this run for you locally?

Looks good just need to make sure the generic fits reader work with the NDA file and if not add a new branch

radiospectra/radiospectra/spectrogram/spectrogram_factory.py

Line 593 in a31141d

def _read_fits(file):

No, it's not working, I'm totally confused with how to fix this, that why i decided to open a PR first and get some inputs

[samaloney]

So you can run the test with the python debugger pdb which should give you a console at the point of the error and you can step up/down the stack and evaluate variables etc or set a break point in your IDE

pytest --remote-data=any --pdb radiospectra/spectrogram/sources/nda.py

[Amityush-lgtm]

Thanks for the tip! I'll try debugging it and see what i find.

[Amityush-lgtm]

The reader failed because the filename has a .1 in v1.1.fits and NDA stores time and frequency in different columns than normal. I've tried to update the factory to handle both.

[samaloney]

Ok nearly there only one small fix left

[hayesla]

this is looking good! one quick question, passing the file returns two spectrograms (RR, and LL) - do we have other classes that do the same thing?

thing to think about for the future on how we deal with this?

[Amityush-lgtm]

Yes, WAVES does the same thing returning one spectrogram per receiver (RAD1 and RAD2) from a single file and i think SoloRPW l2 does too.

[samaloney]

Yea this was something I just did as it was easy - for this specific case the data array could be 3d, time, freq, polarisation as both the time and freq axis are identical.

More generally think like map we could have a keyword that would return either and SpectrogramSeqeunce (WAVES, WIND, RPW etc where the time axis is the same but freq axis is different) or SpectrogramCollection (if both axis are different).

Need to think about how to expose this to users and etc.

[hayesla]

yeah not a blocker on this PR just soemthing to think about. I think that these would be a NDCubecollection rather than a sequence

[Amityush-lgtm]

is it normal for the py313-online CI job to take this long or does it have an issue??

[hayesla]

I'm not sure whats going on with the tests

[Amityush-lgtm]

some of the tests are rerunning quite a few times and still failing.

[hayesla]

yes, i dont think any of the fails are related to this PR

[Amityush-lgtm]

@samaloney might know about this test issue!

[samaloney]

Yea I added retry and timeouts so the test can take bit longer but should pass more often. Do they run locally buy ruing tox in the root of repo (you might have to pip install tox)

[samaloney]

Looks good to go