ANALYSIS BY STATE: AN ALTERNATIVE VIEW ON DISCRETE-EVENT TIME SERIES

摘要

In discrete event simulation experiments, state variables' values are recorded and further processed to explore the dynamics of the modelled system. This paper introduces a family of so-called Analysis by State methods for exploration of relationships between two discrete event simulation output time series. Here, state intervals of a primary time series are visually augmented with information gained by processing corresponding time intervals of a secondary time series, e.g. by displaying interval-wise correlation, distribution, sample aggregates or sample parameters in form of background histograms or heat maps. The desired benefit is to further support and comfortably enhance identification of characteristics and relationships in pairs of discrete-event time series.