TransitionMatrix Struct Reference

Represents the transition matrix for TMMC simulations. More...

Collaboration diagram for TransitionMatrix:

Public Member Functions
bool	operator== (TransitionMatrix const &) const =default
void	initialize ()
	Initializes the transition matrix and related data structures.
void	updateMatrix (double3 Pacc, std::size_t oldN)
	Updates the collection matrix with acceptance probabilities.
void	updateHistogram (std::size_t N)
	Updates the histogram of macrostate visits.
double	biasFactor (std::size_t newN, std::size_t oldN)
	Calculates the bias factor between two macrostates.
void	adjustBias ()
	Adjusts the bias factors based on collected statistics.
void	clearCMatrix ()
	Clears the collection matrix and resets counters.
void	writeStatistics ()
	Writes the transition matrix statistics to a file.

Public Attributes
std::uint64_t	versionNumber {1}
	Version number for serialization compatibility.
std::vector< double3 >	cmatrix
	Collection matrix: x=deletion, y=no change, z=insertion.
std::vector< double >	bias
	Bias factors for each macrostate.
std::vector< double >	lnpi
	Natural logarithm of the probability distribution.
std::vector< double >	forward_lnpi
	Forward ln(pi) for debugging purposes.
std::vector< double >	reverse_lnpi
	Reverse ln(pi) for debugging purposes.
std::vector< std::size_t >	histogram
	Histogram of macrostate visits.
std::size_t	numberOfSteps = {0}
	Number of steps performed in the TMMC simulation.
std::size_t	minMacrostate = {0}
	Minimum macrostate value.
std::size_t	maxMacrostate = {100}
	Maximum macrostate value.
std::size_t	updateTMEvery = {1000000}
	Number of steps between bias updates.
std::size_t	numberOfUpdates = {0}
	Number of times the bias has been updated.
bool	doTMMC = {false}
	Flag indicating whether to perform TMMC simulation.
bool	useBias = {false}
	Flag indicating whether to use bias for changing macrostates.
bool	useTMBias = {true}
	Flag indicating whether to use Transition Matrix bias.
bool	rejectOutofBound = {true}
	Flag indicating whether to reject moves outside macrostate bounds.
bool	rezeroAfterInitialization = {false}
	Flag indicating whether to reset statistics after initialization.

Friends
Archive< std::ofstream > &	operator<< (Archive< std::ofstream > &archive, const TransitionMatrix &m)
Archive< std::ifstream > &	operator>> (Archive< std::ifstream > &archive, TransitionMatrix &m)

Detailed Description

Represents the transition matrix for TMMC simulations.

The TransitionMatrix struct encapsulates the data and methods required for performing Transition Matrix Monte Carlo (TMMC) simulations. It maintains the collection matrix, bias factors, natural logarithm of probability distributions, and histograms to compute macrostates and adjust biases during the simulation. It provides methods to initialize data structures, update matrices and histograms, calculate bias factors, adjust biases, clear statistics, and write statistics to files.

Member Function Documentation

adjustBias()

void TransitionMatrix::adjustBias

(

)

Adjusts the bias factors based on collected statistics.

Updates the bias vector and the natural logarithm of the probability distribution (lnpi) using the current state of the collection matrix and histogram. This method recalculates biases to improve sampling efficiency.

biasFactor()

double TransitionMatrix::biasFactor	(	std::size_t	newN,
		std::size_t	oldN
	)

Calculates the bias factor between two macrostates.

Computes the bias factor used to adjust the acceptance probability of moves between macrostates. The bias factor is calculated as the exponential of the difference in bias between the new and old macrostate.

Parameters

newN	The macrostate value after the move.
oldN	The macrostate value before the move.

Returns

The bias factor for the transition from oldN to newN.

clearCMatrix()

void TransitionMatrix::clearCMatrix

(

)

Clears the collection matrix and resets counters.

Resets the collection matrix, histogram, bias factors, and related counters to their initial values. Typically used after initialization cycles to reset statistics.

initialize()

void TransitionMatrix::initialize

(

)

Initializes the transition matrix and related data structures.

Sets up the collection matrix, bias vectors, and histograms based on the specified macrostate range. Initializes bias factors and probability distributions to default values. Should be called before starting the TMMC simulation.

updateHistogram()

void TransitionMatrix::updateHistogram

(

std::size_t

N

)

Updates the histogram of macrostate visits.

Increments the histogram count for the given macrostate N, tracking how often each macrostate is visited during the simulation.

Parameters

N	The current macrostate value.

updateMatrix()

void TransitionMatrix::updateMatrix	(	double3	Pacc,
		std::size_t	oldN
	)

Updates the collection matrix with acceptance probabilities.

Records the probabilities of transitions between macrostates in the collection matrix. The acceptance probabilities vector Pacc contains probabilities for deletion (x), no change (y), and insertion (z) moves.

Parameters

Pacc	A double3 vector containing acceptance probabilities for deletion (x), no change (y), and insertion (z) moves.
oldN	The macrostate value before the move.

writeStatistics()

void TransitionMatrix::writeStatistics

(

)

Writes the transition matrix statistics to a file.

Outputs the current state of the collection matrix, bias factors, probability distributions, and histogram data to a text file for analysis and debugging.

---

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

• src/raspakit/properties/transition_matrix.ixx