algorithm_functions.py

This file contains helper functions for the SM-SPSA algorithm, such as methods to calculate a stationary distribution or perform the transformation.

generate_centred_mass_initial_matrix(M, C)	Applies the centred mass heuristic to a matrix.
hyperlink_cost_function(x)	Calculates the hyperlink cost function for a transition probability or matrix.
calculate_objective_value(stat_dist, ...)	Calculates the objective function value.
stationary_distribution(Matrix)	Calculates the stationary distribution of a transition matrix.
inverse_logistic_transformation_matrix(...)	Calculates the inverse logistic transformation of a matrix.
logistic_transformation(matrix)	Applies the logistic function to a matrix.
normalisation(matrix)	Normalises a matrix so that the row sums are equal to 1.
scale_matrix(matrix, initial_trans_matrix, C)	Scales a matrix.
logistic_simplex_transformation(matrix, ...)	Performs the complete transformation yielding a transition matrix.