This PR simply refactors the 100-line function which controls a lazy-ptdf pass to call several new functions for its main components to make the existing functionality more modular.
Bigger question: there's now ~300 lines of lazy PTDF logic in unit_commitment.py that should probably be in its own file, and probably be further refactored. There's also the function _lazy_ptdf_dcopf_solve_loop in dcopf.py and the files egret/common/lazy_ptdf_utils.py and the file egret/data/ptdf_utils.py.
Should we organize Egret's PTDF functionality into its own (top-level) directory?
This PR simply refactors the 100-line function which controls a lazy-ptdf pass to call several new functions for its main components to make the existing functionality more modular.
Bigger question: there's now ~300 lines of lazy PTDF logic in unit_commitment.py that should probably be in its own file, and probably be further refactored. There's also the function
_lazy_ptdf_dcopf_solve_loopin dcopf.py and the filesegret/common/lazy_ptdf_utils.pyand the fileegret/data/ptdf_utils.py. Should we organize Egret's PTDF functionality into its own (top-level) directory?Also makes some unrelated changes to the Prescient CI tests because Prescient's dependencies are not currently reliably installed via pip https://github.com/grid-parity-exchange/Prescient/pull/92.