Opacity of an automaton. Application to inhomogeneous Ising chain (Q1180980)

Using a finite automaton to transform (to transduce) an infinite sequence into another one, the author is interested in a certain ``distance'' between the input sequence and the output sequence. This distance is called opacity: an opaque automaton modifies dramatically the input sequences, while a transparent automaton does not really affect them. He then applies this notion to the ``Ising transducer'': the 1-D Ising model with impurities can be seen as an automaton which maps a sequence of impurities to a sequence of induced magnetic fields. The author describes effectively these automata, and computes their opacity (in terms of the ratio \(H/J\) where \(H\) is the external field and \(J\) the coupling constant). Note that the forthcoming paper of \textit{Broglio} and \textit{P. Liardet} announced by the author has been published: Predictions with automata, Proc. Conf. Symbolic dynamics and its applications, New Haven 1991, Contemp. Math. 135 (1992).