Sage 10.1.beta3 released

[Volker Braun]

As always, you can get the latest beta version from the "develop" git branch. Alternatively, the self-contained source tarball is at http://www.sagemath.org/download-latest.html

443b7549ad3 (tag: 10.1.beta3, github/develop) Updated SageMath version to 10.1.beta3
464d3d20aaa gh-35738: configure: Remove stray `stdin.info` file created in test
d27b19a4efc gh-35737: build/pkgs/maxima: Fix parallel build
202ccd7ea08 gh-35733: more uses of yield from
aba52d4c06a gh-35731: build/pkgs/pyflakes: Remove
5b446283d86 gh-35723: some pep8 cleanup in rigged-configurations
e90d6be7a5c gh-35722: cython-lint for imports in libs/
9c83f604a46 gh-35721: cython-lint for imports in rings (mostly in padics)
5c9ee290848 gh-35720: fix the linter and add more checks
6d1cb7e3323 gh-35718: `sage.graphs`: More modularization
a2e4cab86c8 gh-35714: provide matrix_symbolic_sparse class
41aa6c290de gh-35713: update symengine(_py) to versions 0.10.0 and 0.10.1
acac1e60e63 gh-35712: fix pep8 E251 in rings/
34b4234573c gh-35711: fix all pep8 E251 in combinat
0a929ff9da9 gh-35710: fix pep8 E251 in categories, coding, crypto, logic
6a142317110 gh-35709: some cython-linting in quadratic forms
d7bb52abf83 gh-35702: GAP: switch some code to use official libgap APIs
afc9232bff4 gh-35692: ⬆️ Bump myrotvorets/set-commit-status-action from 1.1.6 to 1.1.7
6ca634ae1c0 gh-35671: Upgrade networkx to 3.1, igraph/python_igraph to 0.10.4
7e0eb1c9219 gh-35666: remove obsolete .zenodo.json* files, update CITATION.cff to fix zenodo/DOI
86d2be056bf gh-35626: Update  the rank function of elliptic curves to use ellrank in pari
fe931a790a8 gh-35614: allow non-minimal equation for padic heights
5343f2e7ea9 gh-35494: avoid computation of Conway polynomials when comparing elliptic-curve morphisms
2f426a11f4c (tag: 10.1.beta2) Updated SageMath version to 10.1.beta2

[G. M.-S.]

This is on Apple Silicon Macs ("M" series) with Homebrew up to date, on macOS 11.7.7 (Xcode 13.2.1), macOS 12.6.6 (Xcode 14.2) and macOS 13.4 (Xcode 14.3.1).

With Homebrew's Singular version 4.3.2p2, ./configure does a strange thing:  the URL

https://www.singular.uni-kl.de/Manual/4-3-2/

is opened in one or more browsers.
See also Singular failed doctests below.

(Note:  With ./configure --without-system-singular ("SPKG version 4.3.1p3 will be installed") no Singular issues appear.)

After make and make pytest (which succeed) make ptestlong gives the following.

On macOS 11.7.7, macOS 12.6.6, macOS 13.4:

sage -t --long --warn-long … --random-seed=… src/sage/rings/asymptotic/asymptotics_multivariate_generating_functions.py  # 2 doctests failed
(see attached output)

sage -t --long --warn-long … --random-seed=… src/sage/interfaces/singular.py  # 5 doctests failed

sage -t --long --warn-long … --random-seed=… src/sage/libs/singular/function.pyx  # 1 doctest failed
sage -t --long --warn-long … --random-seed=… src/sage/libs/singular/singular.pyx  # 1 doctest failed

(see attached output)

sage -t --long --warn-long … --random-seed=… src/sage/modular/modform_hecketriangle/abstract_space.py  # 3 doctests failed
sage -t --long --warn-long … --random-seed=… src/sage/modular/modform_hecketriangle/readme.py  # 1 doctest failed
sage -t --long --warn-long … --random-seed=… src/sage/modular/modform_hecketriangle/space.py  # 2 doctests failed

(see https://github.com/sagemath/sage/issues/35273)

sage -t --long --warn-long 26.5 --random-seed=164104516405977008420373837524113607070 src/sage/quadratic_forms/binary_qf.py  # 1 doctest failed

(see https://github.com/sagemath/sage/issues/35292 and https://github.com/sagemath/sage/pull/35262)

Additionally, on macOS 11.7.7:

sage -t --long --warn-long … --random-seed=… src/sage/matrix/matrix_gf2e_dense.pyx  # 1 doctest failed

(see https://github.com/sagemath/sage/issues/33212)

Additionaly, on macOS 13.4:

sage -t --long --warn-long 30.6 --random-seed=… src/sage/libs/giac/__init__.py  # Killed due to kill signal

(see https://github.com/sagemath/sage/issues/35646)

Guillermo

[Emmanuel Charpentier]

On Debian testing running on core i7 + 16 GB RAM, upgrading 10.1.beta2 to 10.1.beta3 and rinning ptestlong gives 3 permanent failures :

---------------------------------------------------------------------- sage -t --long --warn-long 207.1 --random-seed=291812591553963182024849035945523427319 src/sage/schemes/elliptic_curves/ell_rational_field.py # 1 doctest failed sage -t --long --warn-long 207.1 --random-seed=291812591553963182024849035945523427319 src/sage/coding/linear_code.py # 2 doctests failed sage -t --long --warn-long 207.1 --random-seed=291812591553963182024849035945523427319 src/sage/coding/code_constructions.py # 1 doctest failed ----------------------------------------------------------------------

The last two have been already reported a few times, and seem cosmetic. The first one is new and seemingly not cosmetic :

charpent@zen-book-flip:/usr/local/sage-10$ sage -t --long --warn-long 207.1 --random-seed=291812591553963182024849035945523427319 src/sage/schemes/elliptic_curves/ell_rational_field.py # 1 doctest failed Running doctests with ID 2023-06-13-17-49-08-8dac2a23. Git branch: develop Git ref: 10.1.beta0-453-g443b7549ad Running with SAGE_LOCAL='/usr/local/sage-10/local' and SAGE_VENV='/usr/local/sage-10/local/var/lib/sage/venv-python3.11' Using --optional=debian,dot2tex,fricas,gap_jupyter,gap_packages,libsemigroups,msolve,pip,pysingular,sage,sage_spkg,singular_jupyter Features to be detected: 4ti2,benzene,bliss,buckygen,conway_polynomials,csdp,cvxopt,cvxopt,database_cremona_ellcurve,database_cremona_mini_ellcurve,database_cubic_hecke,database_jones_numfield,database_knotinfo,dvipng,fpylll,gfan,graphviz,imagemagick,ipython,jupymake,kenzo,latte_int,lrcalc_python,lrslib,mcqd,meataxe,mpmath,msolve,nauty,networkx,numpy,palp,pandoc,pdf2svg,pdftocairo,pexpect,phitigra,pillow,plantri,polytopes_db,polytopes_db_4d,pplpy,primecountpy,ptyprocess,pynormaliz,pyparsing,python_igraph,requests,rubiks,sage.combinat,sage.geometry.polyhedron,sage.graphs,sage.groups,sage.libs.gap,sage.libs.pari,sage.libs.singular,sage.misc.cython,sage.modules,sage.plot,sage.rings.finite_rings,sage.rings.function_field,sage.rings.number_field,sage.rings.padics,sage.rings.real_double,sage.rings.real_mpfr,sage.symbolic,sage_numerical_backends_coin,sagemath_doc_html,scipy,singular,sphinx,sympy,tdlib Doctesting 1 file. sage -t --long --warn-long 207.1 --random-seed=291812591553963182024849035945523427319 src/sage/schemes/elliptic_curves/ell_rational_field.py ********************************************************************** File "src/sage/schemes/elliptic_curves/ell_rational_field.py", line 2380, in sage.schemes.elliptic_curves.ell_rational_field.EllipticCurve_rational_field.? Failed example: E.gens(use_database=False, algorithm="pari",pari_effort=4) Expected: [(611429153205013185025/9492121848205441 : 15118836457596902442737698070880/924793900700594415341761 : 1)] Got: [(-38749202011873484470143/306317326339867638016 : 678721624672968530804232808604865/5361142413550167706041194328064 : 1)] ********************************************************************** 1 item had failures: 1 of 100 in sage.schemes.elliptic_curves.ell_rational_field.EllipticCurve_rational_field.? [896 tests, 1 failure, 241.41 s] ---------------------------------------------------------------------- sage -t --long --warn-long 207.1 --random-seed=291812591553963182024849035945523427319 src/sage/schemes/elliptic_curves/ell_rational_field.py # 1 doctest failed ---------------------------------------------------------------------- Total time for all tests: 251.7 seconds cpu time: 227.0 seconds cumulative wall time: 241.4 seconds Features detected for doctesting: sage.rings.finite_rings,sage.rings.number_field

HTH,

​

[John Cremona]

The elliptic curve failure might be a consequence of the recently merged PR #35626 about using libpari to compute ranks and generators.

I'll take a look to at least see if the output you get is mathematically correct -- generators are not unique.

John

--
You received this message because you are subscribed to the Google Groups "sage-release" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-release...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/774ed98e-96e3-44bf-b1ad-fe5259d90cb8n%40googlegroups.com.

[John Cremona]

On Tue, 13 Jun 2023 at 17:06, John Cremona <john.c...@gmail.com> wrote:

The elliptic curve failure might be a consequence of the recently merged PR #35626 about using libpari to compute ranks and generators.

I'll take a look to at least see if the output you get is mathematically correct -- generators are not unique.

I was right. This curve has rank 1 and torsion (Z/2Z)^2 so if P is one generator (of infinte order) then there are 3 other, P+T for T a point of order 2.  The one your are getting is one of these.

So we conclude that we cannot reply on pari always returning the same generator despite random seeds etc being fixed in doctests.  I don't know why that is, but if it is the case then this (and other similar) doctests will have to be written accordingly.

[chris wuthrich]

I agree with John. Lines 2380 and 2388 need a "random output" for that reason, like the other calls of that function have already.

Is this a new pull request or is the old #35626 https://github.com/sagemath/sage/pull/35626 opened again for corrections?

Chris

[tsc...@ucdavis.edu]

I don't think it should be random; that has the chance of hiding actual bugs. I think we should check that it belongs to a particular set (specifically, the one from John's comment).

Best,

Travis

[chris wuthrich]

I replaced old tests with "random" by new ones with "random" (and forgot these two). Some other tests do check consistency. 

All that the function really does is converting a pari output back into a point on the elliptic curve. Therefore the test checks if pari runs without error and if the returned is a point on the curve. Checking if it is a generator modulo torsion would be checking if pari gets this right. Is that what we should do?

But this discussing should happen on a pull request. I repeat my question: Do I open a new one or is the faulty initial one reopened? (Trac tickets used to be reopened, that is why I ask).

Chris

[Samuel Lelièvre]

The original pull request was merged, so open a new one.

Trac tickets were occasionally reopened but not after
they were merged in a development release.

[Dima Pasechnik]

You can't reopen PRs, I think, you can certainly reopen issues.

No need to fill in the longish PR boilerplate, just remove it and write  what it does, mentioning #35626

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/93736586-2c19-436d-b4c4-e37987123b50n%40googlegroups.com.

[John Cremona]

I'll review the new PR, though I probably should have seen this problem arising with the original one. 

John

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/CAAWYfq0XsR0ivtRE95-56bRSLsr8CidNMJjruQzstndeSa3uyA%40mail.gmail.com.

[Dave Morris]

Maybe I'm being stupid, but I don't think the given list of generators is complete. The inverse of a generator is also a generator, so aren't there 8 generators, not 4?  (Sorry I couldn't find the PR, to put this comment there.)

[John Cremona]

On Sat, 17 Jun 2023, 04:34 Dave Morris, <dmo...@deductivepress.ca> wrote:

Maybe I'm being stupid, but I don't think the given list of generators is complete. The inverse of a generator is also a generator, so aren't there 8 generators, not 4?  (Sorry I couldn't find the PR, to put this comment there.)

You are right.

That's for a rank 1 curve -- number of generators is twice the number of torsion points. For rank at least 2 there are infinitely many set of generators so putting a complete list into the doctest is impossible anyway, and a different approach is needed.

Back in about 2007 William and I tried to come up with a deterministic and unique generating set for any elliptic curve but came to conclusion that it was not possible, though we came quite close 

John

To view this discussion on the web visit https://groups.google.com/d/msgid/sage-release/a8234925-4f54-42c6-a521-5da346f0ae8cn%40googlegroups.com.

[chris wuthrich]

It is https://github.com/sagemath/sage/pull/35787 .