From: "S.J.Cooksey" <sjc205@kent.ac.uk>
Hello,
I think this is a question of taste. I'm doing some research with
Isabelle, and require defining my own transitive closure function. For
the termination proof of this function it seems desirable to pull out a
finite domain of the relation being closed.
Is it better to make the type of transitive closure something like
trans_clos:: "'a fset => ('a => 'a => bool) => 'a => 'a => bool"
or to use non-finite sets and prove finiteness for the given concrete
inputs.
Many thanks,
Simon Cooksey
Last updated: Nov 21 2024 at 12:39 UTC