#16598: Useless new classes and a replacement for _check_pbd
-------------------------------------+-------------------------------------
Reporter: ncohen | Owner:
Type: enhancement | Status: needs_work
Priority: major | Milestone: sage-6.3
Component: combinatorial | Resolution:
designs | Merged in:
Keywords: | Reviewers:
Authors: Nathann Cohen | Work issues:
Report Upstream: N/A | Commit:
Branch: u/ncohen/16598 | cdaf1e2e834f3895d1fb096a139446bf8620d319
Dependencies: #16553 | Stopgaps:
-------------------------------------+-------------------------------------
Comment (by ncohen):
Yo !
> In group divisible design, there must be more parameters:
> - `v` the size of the ground set
> - `K` the admissible size of the blocks
> - `G` the admissible size of the groups
> The definition seems to be standard (see the big paper of Hanani and the
definition IV.1.3 of the Handbook)
Seems more than we need at the moment, but okay ...
> Anyway, GDD is just a strict generalization of PBD and it is very clear
from Handbook IV.1.3.
> That way, your `is_pbd` would just be a lambda function.
lambda functions have no documentation. But yes, the code just takes one
line.
> Actually, it would be nicer that `IncidenceStructure` would actually be
a `GroupDivisibleDesign` (groups being the partition into points). The
points can be ordered in such way that the groups are
> {{{
> 0 1 ... g1-1|g1 g1+1 ... g1+g2-1|...
> }}}
> and then the new class would just need an extra `._group_sizes`
attribute that would be an integer partition.
Until we want GDD to be defined on something different than `0,...,n-1`
just like you did for `IncidenceStructure`
> Some methods are not well adapted to group divisible designs
> - `dual`: is the dual well defined for GDD?
Given that the GDD is a set of blocks, why not ? Plus you wanted two lines
above that GDD be equal to an `IncidenceStructure`.
> - `incidence_matrix`: must be the incidence matrix group/block (or put
a flag allowing to have points/blocks)
Why ?
> - `automorphis_group` is wrong since it must preserver the groups
Then the implementation is correct.
> Why do you use tuples where everywhere in `IncidenceStructure` there are
lists?
>
> Would you mind having a "stupid" `.group_sizes()` method?
Yes, I hate those methods. Write it if you want one.
Nathann
--
Ticket URL: <http://trac.sagemath.org/ticket/16598#comment:6>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" 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 http://groups.google.com/group/sage-trac.
For more options, visit https://groups.google.com/d/optout.
