Mon 2017-02-27 10:02:50 UTC-6, Salvatore Stella:
> Dear all,
> some time ago I reported a bug in the implementation of `ClusterQuiver`
> which yields wrong answers. The relevant ticket is #22381. To give you
> an idea things like this happen:
>
> {{{
> sage: B = Matrix([[0,1,0],[-1,0,1],[0,-1,0],[2,0,0]])
> sage: Q = ClusterQuiver(B)
> sage: Q.is_mutation_finite() # This should return True
> False
> }}}
>
> After some thought I realized that properly fixing all the issues in that
> ticket will require quite a lot of work: they all stem from how the data
> structure is implemented. Unfortunately I am currently unable to set aside
> enough time to handle it. Which is the best way of proceeding here?
> Should I add a warning message and leave it be? Is there any other option
> available?
> Thanks
> S.
Travis Scrimshaw [2017-02-27 09:01:03]:
> Hey Salvatore,
> The typical approach is to put a stopgap, but it is best to do it in
> such a way that is as minimal as possible (i.e., hopefully not every
> ClusterQuiver from a matrix will have this stopgap pop up, or at worst
> only when the input is a matrix).
>
> Best,
> Travis
Mon 2017-02-27 18:58:30 UTC+1, Salvatore Stella:
> Sorry for the mess, I just read the procedure on the developer's manual.
> I'll be creating a new ticket for the stopgap in a second.
> S.
For reference, links to tickets related to this discussion:
- #22381 Bugs in how ClusterQuiver handles coefficients
https://trac.sagemath.org/ticket/22381
- #22464 Add stopgap for #22381
https://trac.sagemath.org/ticket/22464
- #22482 Non-deterministic test failure in cluster_algebra_quiver
https://trac.sagemath.org/ticket/22482
--
You received this message because you are subscribed to the Google Groups
"sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/sage-devel.
For more options, visit https://groups.google.com/d/optout.
