Am 2012-03-26 00:16, schrieb Nathann Cohen: > Hellooooooooooooooo !!! > > Sage has no Module/Class for things like Finite State Machines, Automata > and Transducers. That should be changed. ;) > > Perhaps the "Word guys" would be interested in the adventure, by the way :-) :)
For clearification, we need that here in our research group, so we will implement it soon. The connection to the word classes should definitely be taken into accout when doing so. > At the moment I think about the following: In which way could this be > done "good"? One idea is to derive it directly from DiGraph and use a > class "State" as label for each vertex and a class Transition as label > for an edge. If you have any comments on that, let me know. (I mean, > tell me, if that is a bad idea, how it can be done better, what I should > be aware of, and so on.) > > > Hmmm... Using a DiGraph to store it sounds rather sensible , but you may > want to "hide" it somehow, a bit like a Poset has a _hasse_diagram > element that is a DiGraph, while a Poset does not extend the DiGraph class. > > The main problem with DiGraph is that... They do a lot of stuff, and you > probably do not want your objects to inherit all of that Usually, when working with transducers, you need a lot of the graph algorithms already implemented. So from that point of view, it is a good thing to use DiGraph extensively. I would guess, that there are two options: - Either to inherit Finite State Machines from DiGraph, - or to use Digraph as a storage for the data only (i.e. in FinitStateMachine, there is a variabel _data_ that is a DiGraph object). Daniel -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
