NEB Simulation Engine

class mlip.simulation.ts_search.neb_simulation_engine.NEBSimulationEngine(atoms: list[Atoms], force_field: ForceField, config: NEBSimulationConfig)

Simulation engine handling Nudged Elastic Band (NEB) transition state searches using the ASE backend.

__init__(atoms: list[Atoms], force_field: ForceField, config: NEBSimulationConfig) → None

Constructor.

atoms

List of Atoms objects. If the list has two elements, these will be considered the initial and final images of the NEB. If the list has three elements, the image in the middle will be considered the transition state guess. If the list has more elements, interpolation will be skipped. In case of two or three provided images, the initial images in between will be interpolated using the IDPP method to yield the total number of images specified in the config.

force_field

Force field model used to compute the predictions.

config

Configuration for the NEB simulation.

run() → None

Run the NEB simulation.