Doubts about detecting the periodicity of psuedo random functions
I'm trying to measure the periodicity of a random generator, but I've run into a theoretical problem. Basically, I was going on the assumption that if a generator repeated a sequence then the period was certain. And so far, this has been the case. But now that I think of it, a sufficiently complex generator could possibly repeat itself but not actually be stuck in a periodic cycle, couldn't it? If so, the only way I can think to handle it is to increase the number of passes and return some sort of probablistic "certificate" of periodicity. Am I way off base here?