Refactor photometry.py - Githubissues

SNflows / flows

Flows Pipeline

https://flows.phys.au.dk

GNU General Public License v3.0

3 stars 4 forks source link

Refactor photometry.py #61

Closed emirkmo closed 2 years ago

emirkmo commented 2 years ago

Photometry.py should be refactored to take in the new FlowsImage dataclass, and be re-written in an OOP format.

Key things to initially abstract out:

[x] Zero point and field calibration
[x] Output I/O
[x] File I/O,
[ ] Template subtraction
[x] WCS checking/recalculation
[x] Reference star instantiation and cleaning
[x] (Including Star detection and Catalog matching)

Issues to tackle:

[ ] Add in WCS calculation via astrometry.net api.
[x] Handle proper motion
[x] Dealing with local archive.
[x] Add basic try catch for sep errors (or try to move to photutils for this).
[ ] uncertainty estimation from background rms.
[ ] Initial guesses for PSF should be saved in FlowsImage and be overriden by Instrument class instance. (Alt. both Image and Instrument instances should be used by Photometry class).

emirkmo commented 2 years ago

Doing these will ensure a basic refactor of photometry.py, allowing us to abstract certain key parts that can be easily substituted by alternatives or skipped, make the code extensible and maintainable.

Ideally we also try to at least keep to 3/5 of SOLID to achieve this.

Bonus idea is to create a generic enough Photometry base class that can use different photometry implementations, so we can test our pipeline and use the best one.