Fitsio issues

[astrobel]

Hi! I had trouble installing lkprf due to fitsio - I run Linux, and the issue is in how my Anaconda set-up speaks to my C backend (this is a known issue). I was able to use one of the workarounds suggested in the linked thread and get everything installed, but I have a student who is currently unable to install fitsio and lkprfon a Windows machine, where Conda issues are harder to diagnose.

To this end, I was wondering what the reason is for using fitsio instead of Astropy FITS handling? Regular Lightkurve uses Astropy for FITS and seems fine - there are parts of Astropy that interface with C, but I've never had any trouble installing or using it, and it seems like they're actively developing to keep up with gcc updates, etc.

Thanks!

[Nschanche]

Hi @astrobel, we went with fitsio to keep the requirements for this package as slim as possible. Adding an astropy dependency can add extra complications we were trying to avoid. However, we were unaware of the install issues for Windows systems. We will look into whether we can find a solution to make the install smoother.

In the meantime, I suggest your student may be interested in checking out TIKE. This is MAST's new virtual jupyter platform. This platform has pre-made TESS-friendly environments, relieving a lot of the install and setup burdens for students. We have tested this with lkprf successfully. Using this service is free, it just requires an StSci account.

[astrobel]

Thanks @Nschanche, I appreciate the explanation!