Package csb :: Package numeric :: Module integrators :: Class AbstractIntegrator

Class AbstractIntegrator

object --+
         |
        AbstractIntegrator

Known Subclasses:

Abstract integrator class. Subclasses implement different integration schemes for solving deterministic equations of motion.

Nested Classes
	__metaclass__ Metaclass for defining Abstract Base Classes (ABCs).

Instance Methods

__init__(self, timestep, gradient)
x.__init__(...) initializes x; see help(type(x)) for signature

source code

AbstractPropagationResult

integrate(self, init_state, length, mass_matrix=None, return_trajectory=False)
Integrates equations of motion starting from an initial state a certain number of steps.

source code

State

integrate_once(self, state, current_step, mass_matrix=None)
Integrates one step starting from an initial state and an initial time given by the product of the timestep and the current_step parameter.

source code

Inherited from object: __delattr__, __format__, __getattribute__, __hash__, __new__, __reduce__, __reduce_ex__, __repr__, __setattr__, __sizeof__, __str__, __subclasshook__

Class Variables
	__abstractmethods__ = `frozenset(['integrate_once'])`

Properties
Inherited from `object`: `__class__`

Method Details

init(self, timestep, gradient)
(Constructor)

source code

x.__init__(...) initializes x; see help(type(x)) for signature

Parameters:

timestep (float) - Integration timestep
gradient (AbstractGradient) - Gradient of potential energy

Overrides: object.__init__

(inherited documentation)

integrate(self, init_state, length, mass_matrix=None, return_trajectory=False)

source code

Integrates equations of motion starting from an initial state a certain number of steps.

Parameters:

init_state (State) - Initial state from which to start integration
length (int) - Nubmer of integration steps to be performed
mass_matrix (n-dimensional InvertibleMatrix with n being the dimension of the configuration space, that is, the dimension of the position / momentum vectors) - Mass matrix
return_trajectory (boolean) - Return complete Trajectory instead of the initial and final states only (PropagationResult). This reduces performance.

Returns: AbstractPropagationResult

integrate_once(self, state, current_step, mass_matrix=None)

source code

Integrates one step starting from an initial state and an initial time given by the product of the timestep and the current_step parameter. The input state is changed in place.

Parameters:

state (State) - State which to evolve one integration step
current_step (int) - Current integration step
mass_matrix (n-dimensional numpy array with n being the dimension of the configuration space, that is, the dimension of the position / momentum vectors) - mass matrix

Returns: State

the altered state

Class AbstractIntegrator

__init__(self, timestep, gradient) (Constructor)

integrate(self, init_state, length, mass_matrix=None, return_trajectory=False)

integrate_once(self, state, current_step, mass_matrix=None)

init(self, timestep, gradient)
(Constructor)