Orbital decomposition for ill-behaved event sequences: transients and superordinate structures

Time series analysis is often challenged by the presence of transient functions. We examined some types of transients found in time series of events that lend themselves to symbolic dynamics analysis through the method of orbital decomposition, which is based on the principle that chaotic series arise from coupled oscillators. Synthetic data sets were constructed to study the impact of intrusive events, intrusive series, merged functions, non-coupled oscillators, and driving oscillations on the patterns of final statistics obtained from orbital decomposition analysis. Two real-world data sets - a logbook of the ritual behaviors of a patient with obsessive compulsive disorder and a time series of kill dates from an infamous serial murderer - were examined for non-ergodic properties similar to those found in the synthetic data.