UniversalCyclotomicField is not finite

[soehms]

sage: UCF = UniversalCyclotomicField()
sage: UCF.is_finite()
True

This is explicitly given in the code of universal_cyclotomic_field.py:

    def is_finite(self):
        r"""
        Returns ``True``.

        EXAMPLES::

            sage: UniversalCyclotomicField().is_finite()
            True

        .. TODO::

            this method should be provided by the category.
        """
        return True

This has been correct before a change in April 2015 (Trac #1852). It looks like a mistake in connection with that ticket. The old code from git looked like this:

author  Jean-Philippe Labbé <labbe@math.fu-berlin.de>   2015-04-29 17:56:59 +0300
committer       Jean-Philippe Labbé <labbe@math.fu-berlin.de>   2015-04-29 17:56:59 +0300
commit  dc3de84809792bde5e2cde8c8c579f42496e0647 (patch)
tree    2a5ed5f0daf452364580856388e5717cf47f1b60
parent  Updated Sage version to 6.7.beta3 (diff)
parent  Trac 18152: __neg__ for UCF elements (diff)


-    def is_finite(self):
-        r"""
-        Returns False as ``self`` is not finite.
-
-        EXAMPLES::
-
-            sage: UCF = UniversalCyclotomicField()
-            sage: UCF.is_finite()
-            False
-        """
-        return False

Thus the task might be to restore the old code!

CC: @tscrim

Component: algebra

Keywords: days94

Author: Sebastian Oehms

Branch/Commit: 03055ab

Reviewer: Tomer Bauer, Luis Felipe Tabera

Issue created by migration from https://trac.sagemath.org/ticket/25686

[soehms]

Branch: u/soehms/ucf_not_finite

[soehms]

comment:3

We (Travis and me) tried to move the is_finite method to the category framework. But this lead into difficulties. Thus, we let it fall back to is_finite in rings.pyx replacing the raise of NotImplementedError to a super(Ring, self).is_finite()-call.

---

New commits:

a5673d1

remove meth is_finite from ring.pyx and universal_cyclotomic_field in order to fall back to the categorial framework

[soehms]

Commit: a5673d1

[sagetrac-git]

Changed commit from a5673d1 to 16b48b5

[sagetrac-git]

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

16b48b5

correction

[sagetrac-git]

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

ae0cbb8

correction of comment

[sagetrac-git]

Changed commit from 16b48b5 to ae0cbb8

[soehms]

New commits:

ae0cbb8

correction of comment

[lftabera]

comment:7

You should add a test that shows that the bug is fixed. Since you have deleted the method, this test has to be added elsewhere. In this case, it can be added to the general set of tests at the beggining of the universal_cyclotomic_field.py file

[sagetrac-git]

Changed commit from ae0cbb8 to d3b2c00

[sagetrac-git]

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

d3b2c00

test added

[mathzeta]

Reviewer: Tomer Bauer

[lftabera]

Changed reviewer from Tomer Bauer to Tomer Bauer, Luis Felipe Tabera

[vbraun]

comment:11

See patchbot

[mathzeta]

comment:12

The line return super(Ring, self).is_finite() is probably the culprit, we need to update the error message of that test.

[sagetrac-git]

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

03055ab

doctest error in _list_from_iterator fixed

[sagetrac-git]

Changed commit from d3b2c00 to 03055ab

[mathzeta]

comment:15

Works for me. If the patchbot is happy, then positive review.

[vbraun]

Changed branch from u/soehms/ucf_not_finite to 03055ab