coflandscaper.MaceOpt

class coflandscaper.MaceOpt(fmax=0.01, dtype='float64', head='spice_wB97M', model=None, device='cpu', fix_z=False, max_steps=2000, verbose=True)

Optimize CIF structures with MACE and optional z-axis constraints.

This class wraps ASE optimization with a MACE calculator and supports per-mode batch optimization plus optional energy-summary CSV generation.

Returns:

None.

Parameters:

• fmax (float)

• dtype (str)

• head (str)

• model (str | None)

• device (str)

• fix_z (bool)

• max_steps (int)

• verbose (bool)

Methods

optimize_cof	Optimize one CIF structure and write the optimized CIF.
process_cifs	Optimize all CIF files in one folder and return convergence flags.
run_mode	Optimize selected structures for requested mode(s).
run_preopt	Run one-file pre-optimization with COF-specific default paths.

optimize_cof(input_path, output_path)

Optimize one CIF structure and write the optimized CIF.

Parameters:

• input_path (str) – Input CIF file path.

• output_path (str) – Output CIF file path.

Returns:

True if optimization converged within max_steps, else False.

Return type:

bool

process_cifs(input_folder, output_folder)

Optimize all CIF files in one folder and return convergence flags.

Parameters:

• input_folder (str) – Folder containing input CIF files.

• output_folder (str) – Folder where optimized CIF files are written.

Returns:

Mapping from structure stem to convergence status.

Return type:

dict[str, bool]

run_mode(cof_name, mode, output_base=None, input_base=None, save_opt_energies_csv=True)

Optimize selected structures for requested mode(s).

Default folder behavior: - input_base: {cof_name}/3_{cof_name}_landscape/selection - output_base: {cof_name}/4_{cof_name}_optimization

The method routes mode folders (serr, incl, or both), optimizes all CIF files in each folder, and optionally writes a consolidated {cof_name}_opt_energies_per_layer.csv file.

Parameters:

• cof_name (str) – COF identifier used for default path construction.

• mode (str) – Mode selector (“serr”, “incl”, or “both”).

• output_base (str | None) – Optional optimization output base folder. Defaults to None (uses {cof_name}/4_{cof_name}_optimization).

• input_base (str | None) – Optional selection input base folder. Defaults to None (uses {cof_name}/3_{cof_name}_landscape/selection).

• save_opt_energies_csv (bool) – Whether to write merged per-layer energy CSV. Defaults to True.

Return type:

None

run_preopt(cof_name, input_path=None, output_path=None, fix_z=True)

Run one-file pre-optimization with COF-specific default paths.

Parameters:

• cof_name (str) – COF name used for default path construction.

• input_path (str | None) – Optional input CIF path. Defaults to {cof_name}/1_{cof_name}_single_layer/{cof_name}_unopt.cif.

• output_path (str | None) – Optional output CIF path. Defaults to {cof_name}/1_{cof_name}_single_layer/{cof_name}_preopt.cif.

• fix_z (bool) – Whether to fix atomic z coordinates during this preopt run. Defaults to True.

Returns:

True if the optimization converged, else False.

Return type:

bool