Opacity complexity of automatic sequences. The general case
J. -P. Allouche, J. -Y. Yao
TL;DR
The paper proposes opacity complexity as a measure for $k$-automatic sequences, formalized via finite $k$-automata with output and the intrinsic automaton concept. It shows that opacity can be computed via the intrinsic automaton and that for the standard prefix metric the opacity constant is $M_k=1/2$, enabling concrete calculations. The authors compute opacity properties for classical sequences (e.g., Thue-Morse, period-doubling, Golay-Shapiro-Rudin, paperfolding, Baum-Sweet, Tower of Hanoi) and show a spectrum of opacity outcomes. This framework provides a tool to classify automatic sequences by external visibility and lays groundwork for further study under alternative comparison metrics.
Abstract
In this work we introduce a new notion called opacity complexity to measure the complexity of automatic sequences. We study basic properties of this notion, and exhibit an algorithm to compute it. As applications, we compute the opacity complexity of some well-known automatic sequences, including in particular constant sequences, purely periodic sequences, the Thue-Morse sequence, the period-doubling sequence, the Golay-Shapiro(-Rudin) sequence, the paperfolding sequence, the Baum-Sweet sequence, the Tower of Hanoi sequence, and so on.
