TODO: A bit about Time_Integration...


Time_Integration


Time Integrators’ Base Class

class TimeIntegratorBase(options_dict=None)

Bases: object

Base class defining the time integration process.

Time Integrator constructor.

Parameters:
  • model – Model Object
  • options_dict

    Dictionary object with all time integration objects Support configurations:

    • ‘model’: None,
    • ‘initial_state’: None,
    • ‘checkpoints’: None,
    • ‘initial_time’: None,
    • ‘final_time’: None,
    • ‘step_size’: None,
    • ‘rel_tol’: 1e-3,
    • ‘abs_tol’: 1e-3,
    • ‘int_method’: 0,
    • ‘adaptive_stepping’: False,
    • ‘fixed_stepping’: True

Methods

__init__(options_dict=None)
integrate(initial_state=None, checkpoints=None, step_size=None, rel_tol=None, abs_tol=None)

Integrates forward in time with a variable timestep with relative, rel_tol, and absolute, abs_tol, tolerances between checkpoints(i) and checkpoints(i+1).

Parameters:
  • initial_state – initial state of the model
  • checkpoints – out times for the model states. This list of model states corresponds to the times in checkpoints.
  • step_size – time integration step size.
  • rel_tol – relative tolerance for adaptive integration
  • abs_tol – absolute tolerance for adaptive integration
Returns:

A list of solution vectors corresponding to the times specified in checkpoints or self._checkpoints

Return type:

soln_state

set_coeffs()

set method coefficients

set_options(options_dict)

Set the options of the time integrators. If only a single dictionary is provided, update current dictionary with new values, otherwise reconcile with the default options dictionary. Then update class variables with the provided / reconciled dictionary values.

Parameters:
  • options_dict – a dictionary object with the following keys:
  • model – The model object to be integrated forward in time.
  • initial_state – initial state of the model.
  • initial_time – initial model time.
  • final_time – final model time.
  • step_size – time integration step size.
  • rel_tol – relative tolerance for adaptive integration.
  • abs_tol – absolute tolerance for adaptive integration.
  • int_method – a key corresponding to the integration method chosen.
  • adaptive_stepping – A boolean value, that forces use of adaptive time stepping.
  • fixed_stepping – A boolean value, that forces use of fixed time stepping.
  • def_options_dict (optional) – a dictionary object with same keys as above.
sub_iterate(local_t0, local_tf)

Integrates across a single subinterval of the time integration window.

Parameters:
  • local_t0 – beginning time of local integration window.
  • local_tf – end time of local integration window.
Returns:

solution at local_tf

Return type:

soln_state