Currently, solposx has three dependencies:
All functionality depends on numpy. The timestamp handling requires pandas. pvlib is the package with the largest download footprint, and is not required for most functions.
I would like to make pvlib an optional package. My motivation is to make loading solposx faster, which is important for lightweight applications, such as interactive examples using marimo.
I imagine the process would be something like:
- Step 1: add sind, cosd, tand, asind, and acosd to tools.py instead of importing from pvlib
- Step 2: consider copying
pvlib.spa.calculate_deltat to solposx (used by noaa and usno)
- Step 3: add optional dependency warning to
solposx.solarposition.pvlib
Currently, solposx has three dependencies:
All functionality depends on numpy. The timestamp handling requires pandas. pvlib is the package with the largest download footprint, and is not required for most functions.
I would like to make pvlib an optional package. My motivation is to make loading solposx faster, which is important for lightweight applications, such as interactive examples using marimo.
I imagine the process would be something like:
pvlib.spa.calculate_deltatto solposx (used by noaa and usno)solposx.solarposition.pvlib