minimize

Energy minimization helpers for molecular mechanics systems.

import mlx_atomistic.minimize

Classes

MinimizationResult

class MinimizationResult
    def __init__(positions: mx.array, energy: mx.array, energy_history: mx.array, max_force_history: mx.array, steps: int, converged: bool, method: str = 'steepest_descent', convergence_reason: str = '')

Result of a simple force-based energy minimization.

Parameters

Name	Type	Default	Description
positions	mx.array
energy	mx.array
energy_history	mx.array
max_force_history	mx.array
steps	int
converged	bool
method	str	'steepest_descent'
convergence_reason	str	''

Functions

minimize_energy

def minimize_energy(positions, force_terms: ForceTerm | list[ForceTerm] | tuple[ForceTerm, ...], *, cell: Cell | None = None, max_steps: int = 200, step_size: float = 0.001, force_tolerance: float = 0.001, backtracking: float = 0.5, min_step_size: float = 1e-12, neighbor_manager: NeighborListManager | None = None, method: str = 'steepest_descent') -> MinimizationResult

Minimize potential energy with a selectable local optimizer.

Parameters

Name	Type	Default	Description
positions
force_terms	ForceTerm | list[ForceTerm] | tuple[ForceTerm, ...]
cell	Cell | None	None
max_steps	int	200
step_size	float	0.001
force_tolerance	float	0.001
backtracking	float	0.5
min_step_size	float	1e-12
neighbor_manager	NeighborListManager | None	None
method	str	'steepest_descent'

Returns

• MinimizationResult