The Model class encapsulates a Mesh and a Record object. More...

#include <Model.hpp>

Inheritance diagram for Model:

Classes
struct	ThermalFluctuationRecord

Public Member Functions
	Model (Mesh &mesh_, Record &record_)
	Constructs a new Model object.

bool	should_continue_optimization ()
	Determines whether the optimization algorithm should continue running.

void	determine_trial_step_size ()
	This code determines the step size used in an optimization algorithm. The step size is used to calculate how far to move in each iteration of the optimization algorithm to minimize an energy function.

std::string	to_string_current_step ()
	Returns a string representation of the current step's iteration number, trial step size, and optimal step size. The function concatenates the values of model.iteration, model.oa.trialStepSize, and model.stepSize to the following format: "step: , trial StepSize = , StepSize = ".

void	update_ncg_direction ()
	Updates the direction for nonlinear conjugate gradient (NCG) optimization.

void	reset_ncg_direction ()
	Reset the direction for nonlinear conjugate gradient (NCG) optimization.

double	linear_search_for_stepsize_to_minimize_energy ()
	Performs a linear search to find the optimal step size that minimizes energy using either nonlinear conjugate gradient method or simple line search.

bool	simulated_annealing_next_step (bool forceAttempt=false)
	Use a Metropolis thermal trial move to sample beyond local gradient descent.

void	enforce_boundary_conditions_after_coordinate_update ()
	Synchronize boundary/ghost coordinates after direct coordinate updates.

void	update_vertex_using_NCG ()
	Update vertex coordinates using a non-linear conjugate gradient method.

Public Attributes
Mesh &	mesh

Record &	record

int	iteration

double	stepSize

OptimizationAlgorithm	oa

std::vector< Force >	ncgDirection0
	vector to store the initial direction for nonlinear conjugate gradient optimization

bool	isHeating = true

int	currentCoolingStep = 0

int	coolingStep = 35

int	currentHeatingStep = 0

int	heatingStep = 15

double	highTemperature = 0.0

std::mt19937	thermalRng
	Reproducible random generator for thermal fluctuation trial moves.

int	thermalFluctuationAttemptCount = 0
	Number of attempted thermal trial moves.

std::vector< ThermalFluctuationRecord >	thermalFluctuationRecords
	Diagnostics for thermal trial moves.

Detailed Description

The Model class encapsulates a Mesh and a Record object.

Constructor & Destructor Documentation

◆ Model()

Model::Model	(	Mesh &	mesh_,
		Record &	record_
	)

Constructs a new Model object.

Parameters

mesh_	The mesh object to be encapsulated.
record_	The record object to be encapsulated.

Member Function Documentation

◆ determine_trial_step_size()

void Model::determine_trial_step_size ( )

This code determines the step size used in an optimization algorithm. The step size is used to calculate how far to move in each iteration of the optimization algorithm to minimize an energy function.

Returns: double trialStepSize

◆ enforce_boundary_conditions_after_coordinate_update()

void Model::enforce_boundary_conditions_after_coordinate_update ( )

Synchronize boundary/ghost coordinates after direct coordinate updates.

◆ linear_search_for_stepsize_to_minimize_energy()

double Model::linear_search_for_stepsize_to_minimize_energy ( )

Performs a linear search to find the optimal step size that minimizes energy using either nonlinear conjugate gradient method or simple line search.

Returns: double The optimal step size. Returns -1 if no efficient step size is found with the simple line search method.

◆ reset_ncg_direction()

void Model::reset_ncg_direction ( )

Reset the direction for nonlinear conjugate gradient (NCG) optimization.

◆ should_continue_optimization()

bool Model::should_continue_optimization ( )

Determines whether the optimization algorithm should continue running.

The function returns true if the model has not yet been optimized and the maximum number of iterations has not been reached. Otherwise, the function returns false and the optimization algorithm stops.

Note: The implementation of this function can be changed to customize the optimization.

Returns: true if the optimization algorithm should continue running, false otherwise.

◆ simulated_annealing_next_step()

bool Model::simulated_annealing_next_step ( bool forceAttempt = false )

Use a Metropolis thermal trial move to sample beyond local gradient descent.

Returns: true if a thermal trial was attempted on this iteration.

◆ to_string_current_step()

std::string Model::to_string_current_step ( )

Returns a string representation of the current step's iteration number, trial step size, and optimal step size. The function concatenates the values of model.iteration, model.oa.trialStepSize, and model.stepSize to the following format: "step: , trial StepSize = , StepSize = ".

Returns: std::string A string representation of the current step's information.

◆ update_ncg_direction()

void Model::update_ncg_direction ( )

Updates the direction for nonlinear conjugate gradient (NCG) optimization.

This function calculates the dot product of the previous and current forces to determine the NCG factor. It then updates the NCG direction either as a combination of the current force and the previous NCG direction, or just the current force depending on whether NCG optimization is being used. The NCG direction is updated in the input vector ncgDirections.

◆ update_vertex_using_NCG()

void Model::update_vertex_using_NCG ( )

Update vertex coordinates using a non-linear conjugate gradient method.

Member Data Documentation

◆ coolingStep

int Model::coolingStep = 35

◆ currentCoolingStep

int Model::currentCoolingStep = 0

◆ currentHeatingStep

int Model::currentHeatingStep = 0

◆ heatingStep

int Model::heatingStep = 15

◆ highTemperature

double Model::highTemperature = 0.0

◆ isHeating

bool Model::isHeating = true

◆ iteration

int Model::iteration

The current iteration count of the optimization algorithm.

◆ mesh

Mesh& Model::mesh

The mesh object.

◆ ncgDirection0

std::vector<Force> Model::ncgDirection0

vector to store the initial direction for nonlinear conjugate gradient optimization

◆ oa

OptimizationAlgorithm Model::oa

The state for the optimization algorithm.

◆ record

Record& Model::record

The record object.

◆ stepSize

double Model::stepSize

The current step size

◆ thermalFluctuationAttemptCount

int Model::thermalFluctuationAttemptCount = 0

Number of attempted thermal trial moves.

◆ thermalFluctuationRecords

std::vector<ThermalFluctuationRecord> Model::thermalFluctuationRecords

Diagnostics for thermal trial moves.

◆ thermalRng

std::mt19937 Model::thermalRng

Reproducible random generator for thermal fluctuation trial moves.

The documentation for this class was generated from the following file:

include/model/Model.hpp

Classes

Public Member Functions

Public Attributes