MARCA: Markov Chain Analyzer
William J. Stewart · 2021
MARkov Chain Analyzer (MARCA) is a software package designed to facilitate the generation of large Markov chains; to analyze the structure of the chain, in particular to determine the extent to which the chain may be decomposed; and to compute both the transient distribution at any time, and the stationary probability distribution. In MARCA, each component of this state descriptor vector is referred to as a bucket. Each bucket contains a number of balls that is equal to the numeric value of the corresponding vector component. The evolution of the Markov chain is represented by the movement of balls among the buckets. As the states of the Markov chain are generated, they are stored in a one-dimensional array called LIST . The different types of algorithms for determining the stationary probability vector in MARCA are direct methods, fixed-point iteration methods, subspace iteration methods, and combination of these.