search

BrownDwarf / gollum

A microservice for programmatic access to precomputed synthetic spectral model grids in astronomy
https://gollum-astro.readthedocs.io/
MIT License
20 stars 5 forks source link

Compare data fetching and caching strategies to expecto #87

Open gully opened 1 year ago

gully commented 1 year ago

I just noticed a code by @bmorris3 called expecto:

https://github.com/bmorris3/expecto

It's a really lightweight package to fetch PHOENIX models. We should check out the code to compare data fetching and caching strategies. Maybe make a docs page called "related software" or something.

bmorris3 commented 1 year ago

Happy to chat if that's helpful! I suspect it won't be super helpful because gollum has a much broader scope.

We should check out the code to compare data fetching and caching strategies.

The fetching and caching is all handled by astropy.utils.data. The related expecto source code is really brief, see here.

In short, I think expecto is useful if you need a synthetic spectrum that's "better" than a blackbody, but don't care about interpolation. That's not useful for everyone (but it is for me). gollum does all the good/hard work that I dodged!

gully commented 1 year ago

Neat, thanks for this guidance @bmorris3, I think the caching with astropy.utils.data could be useful, since we're hardcoding this ourselves at the moment. May be worth looking into. Thanks!