Stepsize#

class Stepsize(initial_value=0.0, stepsize_change_ratio=None, t_stop=None, stepsize_stop_max=None, max_stepsize=None, dt_min=None, milestones=None)[source]#

Bases: object

Description of Stepsize

Parameters:

  • initial_value (float, optional) – initial stepsize. Defaults to 0.0.

  • stepsize_change_ratio (float, optional) – stepsize change ratio. Defaults to None.

  • t_stop (float, optional) – time at which the adaptive stepsize stops. Defaults to None.

  • stepsize_stop_max (float, optional) – Maximum stepsize after t_stop. Defaults to None.

  • max_stepsize (float or callable, optional) – Maximum stepsize. Can be a function of festim.t. Defaults to None.

  • dt_min (float, optional) – Minimum stepsize below which an error is raised. Defaults to None.

  • milestones (list, optional) – list of times by which the simulation must pass. Defaults to None.

Variables:

  • adaptive_stepsize (dict) – contains the parameters for adaptive stepsize

  • value (fenics.Constant) – value of dt

  • milestones (list) – list of times by which the simulation must pass.

Example:

my_stepsize = Stepsize(
    initial_value=0.5,
    stepsize_change_ratio=1.1,
    max_stepsize=lambda t: None if t < 1 else 2,
    dt_min=1e-05
)
adapt(t, nb_it, converged)[source]#

Changes the stepsize based on convergence.

Parameters:

  • t (float) – current time.

  • nb_it (int) – number of iterations the solver required to converge.

  • converged (bool) – True if the solver converged, else False.

initialise_value()[source]#

Creates a fenics.Constant object initialised with self.initial_value and stores it in self.value

next_milestone(current_time: float)[source]#

Returns the next milestone that the simulation must pass. Returns None if there are no more milestones.

Parameters:

current_time (float) – current time.

Returns:

next milestone.

Return type:

float