change_labels for sum species structure does not actually change the labels

[mwhansen]

sage: X = species.SingletonSpecies()
sage: X2 = X+X
sage: X2.structures([1]).list()
[1, 1]
sage: s = X2.structures([1]).list()[0]
sage: s.labels()
[1]
sage: s.change_labels(["a"])
1

CC: @sagetrac-sage-combinat @mantepse

Component: combinatorics

Keywords: species

Author: Mike Hansen

Branch/Commit: 5fda98d

Reviewer: Florent Hivert, Martin Rubey

Issue created by migration from https://trac.sagemath.org/ticket/10227

[mwhansen]

Author: Mike Hansen

[mwhansen]

comment:1

This patch does a couple of things:

1. Fix the change_labels bug above.

2. Add a left/right summand/factor method to sum and product species.

3. Make the labels method return all of the labels rather than just the ones that appear. This is a backwards incompatible change, but I am pretty sure that there isn't really any code out there that depends on this.

4. Add a labels_subset method to do what the old labels methods did for subsets.

5. Add additional documentation to some of the classes.

[hivert]

Reviewer: Florent Hivert

[hivert]

comment:3

I'm assigning the ticket to myself to remember to review it.

[mwhansen]

comment:4

Florent, if you have time, could you take a quick look at this. #5512 is waiting on it.5512 Thanks!

[mwhansen]

Dependencies: #5512

[mwhansen]

comment:5

I rebased trac_10227-species_fixes-mh.patch on #5512

Apply trac_10227-species_fixes-mh.patch

[mwhansen]

Changed dependencies from #5512 to #5512, #13317

[mwhansen]

comment:7

Attachment: trac_10227-species_fixes-mh.patch.gz

[mwhansen]

Changed dependencies from #5512, #13317 to #5457, #5512, #13317

[fchapoton]

Changed keywords from none to species

[fchapoton]

Commit: 33131ae

[fchapoton]

Branch: u/chapoton/10227

[fchapoton]

New commits:

33131ae

#10227: change_labels for sum species structure does not actually change the labels

[sagetrac-git]

Changed commit from 33131ae to 9382d6f

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

0581575	Merge branch 'u/chapoton/10227' of ssh://trac.sagemath.org:22/sage into 10227
9382d6f	trac #10277 a few minor doc formatting

[sagetrac-git]

Changed commit from 9382d6f to d1d5103

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

d1d5103

Merge branch 'u/chapoton/10227' into 6.10.b2

[mantepse]

comment:20

I'm now building this.

[mantepse]

comment:21

Looks OK. It might make sense to add a docstring to change_labels in

composition_species.py:69:    def change_labels(self, labels):
partition_species.py:112:    def change_labels(self, labels):
product_species.py:119:    def change_labels(self, labels):
structure.py:125:    def change_labels(self, labels):
structure.py:267:    def change_labels(self, labels):

I suggest:

Returns a relabelled structure.

INPUT:

- ``labels``, a list of labels.

OUTPUT:

A structure with the i-th label of self replaced with the i-th label of the list.

EXAMPLES::

    sage: C = species.CycleSpecies(3)
    sage: ls1 = C.structures([1,2,3]); ls2 = C.structures(["a", "b", "c"])
    sage: s = s1[0].change_labels(["b", "c", "a"]); t = s2[0]; (s,t,s==t)
    (('b', 'c', 'a'), ('a', 'b', 'c'), False)
    sage: s = s1[0].change_labels(["a", "b", "c"]); t = s2[0]; (s,t,s==t)
    (('a', 'b', 'c'), ('a', 'b', 'c'), True)

[sagetrac-git]

Changed commit from d1d5103 to 5fda98d

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

8a9b25b	Merge branch 'u/chapoton/10227' into 6.7.b3
5fda98d	adding doc to change_labels methods

[mantepse]

comment:23

(You didn't put my example into the doc for change_labels, but that's OK. It's purpose would have been to clarify equality of structures: one might think that (b, c, a) and (a, b, c) are the same cycle, but they are not. So it should actually be in the doc for equality)

Can I give this positive review? (Florent is in the reviewer list, not me)

[fchapoton]

comment:24

Replying to @mantepse:

(You didn't put my example into the doc for change_labels, but that's OK. It's purpose would have been to clarify equality of structures: one might think that (b, c, a) and (a, b, c) are the same cycle, but they are not. So it should actually be in the doc for equality)

Where precisely do you want this example to be added ?

Can I give this positive review? (Florent is in the reviewer list, not me)

Yes, you can ; just add yourself to the reviewer list.

[mantepse]

Changed reviewer from Florent Hivert to Florent Hivert, Martin Rubey

[vbraun]

Changed dependencies from #5457, #5512, #13317 to none

[vbraun]

Changed branch from u/chapoton/10227 to 5fda98d