verlet

verlet#

Generic Verlet integrator

Classes#

`TemperatureStateItem`	Base class for dataclasses that should act like a JAX pytree node.
`ConsErrTracker`	A class that tracks the errors on the conserved quantity.
`VerletHook`	Specialized Verlet hook.
`ThermostatHook`	Specialized Verlet hook.
`BarostatHook`	Specialized Verlet hook.
`NVE`	Specialized Verlet hook.
`VerletIntegrator`	Base class for dataclasses that should act like a JAX pytree node.

Module Contents#

class verlet.TemperatureStateItem(*args, **kwargs)#

Bases: IMLCV.new_yaff.iterative.StateItem

Base class for dataclasses that should act like a JAX pytree node.

get_value(iterative)#

class verlet.ConsErrTracker(*args, **kwargs)#

Bases: IMLCV.base.datastructures.MyPyTreeNode

A class that tracks the errors on the conserved quantity. Given its superior numerical accuracy, the algorithm below is used to calculate the running average. Its properties are discussed in Donald Knuth’s Art of Computer Programming, vol. 2, p. 232, 3rd edition.

counter: jax.Array#

ekin_m: jax.Array#

econs_m: jax.Array#

ekin_s: jax.Array#

econs_s: jax.Array#

update(ekin, econs)#

get()#

class verlet.VerletHook(*args, **kwargs)#

Bases: IMLCV.new_yaff.iterative.Hook

Specialized Verlet hook.

This is mainly used for the implementation of thermostats and barostats.

econs_correction: jax.Array#

__call__(iterative: VerletIntegrator)#

abstract init(iterative: VerletIntegrator) → tuple[Self, VerletIntegrator]#

abstract pre(iterative: VerletIntegrator, **kwargs) → tuple[Self, VerletIntegrator]#

abstract post(iterative: VerletIntegrator, **kwargs) → tuple[Self, VerletIntegrator]#

class verlet.ThermostatHook(*args, **kwargs)#

Bases: VerletHook

Specialized Verlet hook.

This is mainly used for the implementation of thermostats and barostats.

temp: jax.Array#

abstract pre(iterative: VerletIntegrator, G1_add: jax.Array | None = None) → tuple[Self, VerletIntegrator]#

abstract post(iterative: VerletIntegrator, G1_add: jax.Array | None = None) → tuple[Self, VerletIntegrator]#

class verlet.BarostatHook(*args, **kwargs)#

Bases: VerletHook

Specialized Verlet hook.

This is mainly used for the implementation of thermostats and barostats.

temp: jax.Array#

press: jax.Array#

baro_ndof: int | None = None#

dim: int | None = None#

abstract pre(iterative: VerletIntegrator, chainvel0: jax.Array | None = None) → tuple[Self, VerletIntegrator]#

abstract post(iterative: VerletIntegrator, chainvel0: jax.Array | None = None) → tuple[Self, VerletIntegrator]#

class verlet.NVE(*args, **kwargs)#

Bases: VerletHook

Specialized Verlet hook.

This is mainly used for the implementation of thermostats and barostats.

_: dataclasses.KW_ONLY#

temp: float = 0.0#

Specialized Verlet hook.

This is mainly used for the implementation of thermostats and barostats.

econs_correction: jax.Array#

__call__(iterative)#

init(iterative: VerletIntegrator) → tuple[VerletHook, VerletIntegrator]#

pre(iterative: VerletIntegrator) → tuple[VerletHook, VerletIntegrator]#

post(iterative: VerletIntegrator) → tuple[VerletHook, VerletIntegrator]#

class verlet.VerletIntegrator(*args, **kwargs)#

Bases: IMLCV.base.datastructures.MyPyTreeNode

Base class for dataclasses that should act like a JAX pytree node.

ff: IMLCV.new_yaff.ff.YaffFF#

counter: jax.Array#

time: jax.Array#

ndof: int#

pos: jax.Array#

vel: jax.Array#

acc: jax.Array#

masses: jax.Array#

timestep: jax.Array#

ptens: jax.Array#

rvecs: jax.Array#

vtens: jax.Array#

verlet_hook: VerletHook#

delta: jax.Array#

pos_old: jax.Array#

gpos: jax.Array#

gpos_bias: jax.Array#

temp: jax.Array#

press: jax.Array#

ekin: jax.Array#

ekin_new: jax.Array#

epot: jax.Array#

etot: jax.Array#

e_bias: jax.Array#

_cons_err_tracker: ConsErrTracker | None = None#

econs: jax.Array#

cons_err: jax.Array#

rmsd_delta: jax.Array#

rmsd_gpos: jax.Array#

rmsd_gpos_bias: jax.Array#

cv: jax.Array#

other_hooks: list[IMLCV.new_yaff.iterative.Hook]#

static create(ff: IMLCV.new_yaff.ff.YaffFF, other_hooks: list[IMLCV.new_yaff.iterative.Hook], timestep: float, thermostat: ThermostatHook | None = None, barostat: BarostatHook | None = None, vel0: jax.Array | None = None, temp0: float = 300.0, scalevel0: bool = True, time0: float | None = None, ndof: int | None = None, counter0: int | None = None, key=42)#

Arguments:

ff: A ForceField instance

Optional arguments:

timestep: The integration time step (in atomic units)
state: A list with state items. State items are simple objects that take or derive a property from the current state of the iterative algorithm.
hooks: A function (or a list of functions) that is called after every iterative.
vel0: An array with initial velocities. If not given, random velocities are sampled from the Maxwell-Boltzmann distribution corresponding to the optional arguments temp0 and scalevel0
temp0: The (initial) temperature for the random initial velocities
scalevel0: If True (the default), the random velocities are rescaled such that the instantaneous temperature coincides with temp0.
time0: The time associated with the initial state.
ndof: When given, this option overrides the number of degrees of freedom determined from internal heuristics. When ndof is not given, its default value depends on the thermostat used. In most cases it is 3*natom, except for the NHC thermostat where the number if internal degrees of freedom is counted. The ndof attribute is used to derive the temperature from the kinetic energy.
counter0: The counter value associated with the initial state.

propagate()#

_compute_ekin()#

Auxiliary routine to compute the kinetic energy

This is used internally and often also by the Verlet hooks.

compute_properties()#

finalize()#

call_hooks()#

run(nstep=None)#