Stopgap for IntegerListsLex

[nathanncohen]

As reported on #17548, the class IntegerListsLex returns wrong result. Until this is fixed, we need a stopgap to warn users.

Component: combinatorics

Author: Nathann Cohen

Branch: 5f00624

Reviewer: Jeroen Demeyer

Issue created by migration from https://trac.sagemath.org/ticket/17637

[nathanncohen]

Branch: public/17637

[sagetrac-git]

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

5f00624

trac #17637: Stopgap for IntegerListsLex

[sagetrac-git]

Commit: 5f00624

[jdemeyer]

comment:3

Since #17548 is about min_slope, shouldn't the warning be only if the min_slope argument was given?

[nathanncohen]

comment:4

Since #17548 is about min_slope, shouldn't the warning be only if the min_slope argument was given?

In the code, it is said that the constraints are "assumed to be correct". I have no idea what happens if min_slope is not defined by max_slope is, or max_part or anything. If you believe that there is a situation in which the code will never return anything wrong you can adapt the error message.

Nathann

[nathanncohen]

comment:6

Thanks

Nathann

[vbraun]

comment:7

Reviewer name missing

[jdemeyer]

Reviewer: Jeroen Demeyer

[vbraun]

Changed branch from public/17637 to 5f00624

[anneschilling]

comment:10

As far as I can see from #17548 there are actually no known bugs with regards to IntegerListsLex. So I find it a total overreaction to put a stopgap as the limitations of IntegerListsLex are correctly documented. Now warnings appear in many parts of the code which are completely unrelated to this issue (and as I said, there does not appear to be a bug). For example

sage: K = crystals.KirillovReshetikhin(['D',4,1],1,1)
/Applications/sage/local/lib/python2.7/site-packages/sage/combinat/partition.py:4827:
********************************************************************************
This code contains bugs and may be mathematically unreliable.
This issue is being tracked at https://github.com/sagemath/sage-prod/issues/17548.
********************************************************************************

or

sage: Partitions(3,max_part=2)
/Applications/sage/local/lib/python2.7/site-packages/sage/combinat/partition.py:6622:
********************************************************************************
This code contains bugs and may be mathematically unreliable.
This issue is being tracked at https://github.com/sagemath/sage-prod/issues/17548.
********************************************************************************
Partitions of 3 having parts less than or equal to 2

both of which are totally fine!

Best,

Anne

[anneschilling]

Changed commit from 5f00624 to none

[nathanncohen]

comment:11

Dear Anne,

As far as I can see from #17548 there are actually no known bugs with regards to IntegerListsLex.

What about those examples from #17548 ?

sage: Partitions(5, min_slope=1).list()
ValueError: [2, 4] is not a valid partition
sage: IntegerListsLex(5, length=3, max_slope=0).list()   # 0 is allowed in the partition
[[5, 0, 0], [4, 1, 0], [3, 2, 0], [3, 1, 1], [2, 2, 1]]
sage: IntegerListsLex(5, max_slope=0).list()             # now its not
[[5], [4, 1], [3, 2], [3, 1, 1], [2, 2, 1], [2, 1, 1, 1], [1, 1, 1, 1, 1]]

I will answer other points on #17898.

Nathann

[anneschilling]

comment:12

Replying to @nathanncohen:

Dear Anne,

As far as I can see from #17548 there are actually no known bugs with regards to IntegerListsLex.

What about those examples from #17548 ?

sage: Partitions(5, min_slope=1).list()
ValueError: [2, 4] is not a valid partition
sage: IntegerListsLex(5, length=3, max_slope=0).list()   # 0 is allowed in the partition
[[5, 0, 0], [4, 1, 0], [3, 2, 0], [3, 1, 1], [2, 2, 1]]
sage: IntegerListsLex(5, max_slope=0).list()             # now its not
[[5], [4, 1], [3, 2], [3, 1, 1], [2, 2, 1], [2, 1, 1, 1], [1, 1, 1, 1, 1]]

I will answer other points on #17898.

Nathann

Read the answers that Travis gave on that ticket. They are consistent with the specifications or bad user input.

Anne