test basic properties of polyhedra

[kliem]

We add a method that tests basic properties, when the TestSuite is run.

Depends on #29903

CC: @jplab @LaisRast @dimpase @orlitzky

Component: geometry

Keywords: polyhedra, test suite

Author: Jonathan Kliem

Branch/Commit: 699d4e8

Reviewer: Matthias Koeppe

Issue created by migration from https://trac.sagemath.org/ticket/29905

[kliem]

Commit: 6eaf649

[kliem]

Branch: public/29905

[kliem]

New commits:

0bdbe49	test suites for the library
5c7e562	fix double description of hypercube
206dbb7	merged in public/29904
c6ea1ec	added long time flags
5d79b2b	some more testsuites
6eaf649	test some basic properties

[mkoeppe]

Reviewer: Matthias Koeppe

[kliem]

comment:3

Thank you.

[mkoeppe]

comment:4

In light of the discussion on sage-devel (https://groups.google.com/d/msg/sage-devel/c4UbKSdt3Aw/UQAo1iYoAAAJ), this needs more work

[kliem]

comment:5

Why? What's wrong with this ticket?

Note that this ticket really has only one commit. If there is something wrong with #29903 or #29904 it should be fixed there. Not that I would know what the problem with those tickets would be.

One thing I could do is to use this ticket to improve the doctest for the hypercubes. Is that what you mean? Somethings as

sage: P = polytopes.hypercube(intervals=random_intervals, backend='field')
sage: P._test_basic_properties()

[mkoeppe]

comment:6

The set_random_seed added in the doctest.

[kliem]

comment:7

If that is the problem, then this ticket here is still green and #29903 should be "needs work". To avoid confusion I can of course quickly push one commit to all three tickets.

So we don't put set_random_seed anymore because we will go through with #29935 soon?

[mkoeppe]

comment:8

Please feel free to move the "needs work" to the correct ticket. Thanks

[mkoeppe]

comment:9

Yes, if you are concerned about the random seeds, this needs a proper solution, not one that is localized to one doctest.

[mkoeppe]

comment:10

By the way, if you want to fuzz this particular doctest, why not just execute it in a loop so it picks up a new set of deterministic pseudo random numbers in each iteration? That should be good enough.

[sagetrac-git]

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

699d4e8

remove set_random seed

[sagetrac-git]

Changed commit from 6eaf649 to 699d4e8

[kliem]

comment:12

Replying to @mkoeppe:

By the way, if you want to fuzz this particular doctest, why not just execute it in a loop so it picks up a new set of deterministic pseudo random numbers in each iteration? That should be good enough.

I already added a new doctest for #29904 that manually tests the only interesting case besides the static case that is always being tested.

[kliem]

comment:15

Thank you.

[vbraun]

Changed branch from public/29905 to 699d4e8