Sage 9.6.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

1e8ba0aac4 (trac/develop, tag: 9.6.beta3) Updated SageMath version to 9.6.beta3
0fed9a145e Trac #33392: sage.matrix: Modularization fixes after #33159
5f2efac091 Trac #33391: Remove compile dependency on flint from CombinatorialPolyhedron
4761943a1b Trac #33389: GenericGraph.distance_matrix: accept keyword arguments for matrix constructor
de4f3f05e3 Trac #33388: Graph: accept keyword arguments for matrix constructor in methods returning a matrix
4c814d6649 Trac #33387: BipartiteGraph.reduced_adjacency_matrix: accept keyword arguments for matrix constructor
9b24711f58 Trac #33386: MR55: Make orders unique parents
ef3270f669 Trac #33376: Apply some coding conventions to sage/modular/overconvergent/hecke_series.py
57787eddf1 Trac #33371: Remove sage-location and NEW_SAGE_ROOT code
a7eabd143a Trac #33272: Update sphinx to 4.4
70a841107e Trac #33194: openblas: Update to 0.3.20
714cf9fe3a Trac #33165: faster evaluation of polynomials in R[x] at monomials in R[y] or R[u,v] and variables in R[x]/(f)
5649402da8 Trac #33001: Invoke manifolds with structure more conveniently
3454122b2e Trac #32958: findstat: enforce limit on how many elements to compute
cd78e1bb53 Trac #32953: Sphere: improve handling of default charts
ef15b5575a Trac #32847: Use lazy_import for module-level import from sage.plot (outside of sage.plot)
f2ed106a91 Trac #32688: Matrix_dense API functions set_unsafe_si vs. set_unsafe_int
590cfa95d0 Trac #32644: Typos in documentation of sage/modular/quasimodform
39b19e63df Trac #32121: Generalize MapCombinatorialClass to ImageSubobject/ImageSet, add method Map.image
ff4fc3f93f Trac #31576: Projective points over rings with zero divisors
ebbf07248d Trac #31355: upgrade lrcalc to 2.1
2c35e81943 Trac #30680: Finitely presented modules over the Steenrod algebra
02bde440d3 Trac #29927: shifting issue in padic _set_from_list functions
5f4c2c49b8 Trac #29919: Restore "huge" package type
8c5f78c826 Trac #22081: Add new option to hyperbolic graphics to select the model
72b70cb269 Trac #14147: weyl group orbit of a weight hangs in infinite type
47b7c6cbce Trac #33390: singular spkg-configure.m4: Better test for help
58e2a16d8c Trac #33390: singular spkg-configure.m4: Better test for help
b96b86a4bc Trac #33377: GenericGraph.[weighted]adjacency_matrix, incidence_matrix: Accept keyword arguments for matrix constructor
5705d221fd Trac #33374: adjust error messages in algebras/
f71a59cf4d Trac #33367: Adjust some error messages in `combinat/`
d2b071e65f Trac #33364: remove traces of # py3 and some # py2
518f9fc106 Trac #33358: Fix for rename of conda e-antic package to libeantic
7abda14ee4 Trac #33357: Random set partition with fixed block sizes does not respect the constraint and other improvements
b17a291d87 Trac #33353: increase tolerance of doctests in modular/modform/numerical.py
bfa90bbd18 Trac #33351: raw docs and other pycodestyle details in some pyx files
09c55f3d97 Trac #33349: a few details in designs
ed86487af3 Trac #33342: fix indentation in toy_variety.py
2fd405438b Trac #33341: pycodestyle cleaning in convolution.py
016f864459 Trac #33336: Fix deprecation warning with scipy 1.8.
a64514b5ce Trac #33335: fix E111 (indentation) in geometry, groups, logic, matrix
aa255bf264 Trac #33334: fix E111 (indentation) in rings
b35def6864 Trac #33333: Deduplicate package names in installation guide
df13563584 Trac #33332: fix E111 (indentation) in dynamics, numerical, modules, interfaces
fd2d77c807 Trac #33328: Some improvements for signed permutations
9d77db7871 Trac #33323: Cardinality of partitions of negative values
567da5a2d4 Trac #33309: Make Sage documentation functional
70505410c7 Trac #33247: Restore quiet in `./configure -q`
56c6240253 Trac #33149: use lazy_string() for string formatting when creating exceptions
5bf1f88d7b Trac #32465: Refactor {Matrix,Vector}_double_dense through ..._numpy_dense, add ..._numpy_integer_dense
72f573e606 Trac #30649: Reimplement "sage -p SPKG" as "cd $SAGE_ROOT && make SPKG-no-deps"
9941688328 Trac #33150: more direct conversion from QQbar to real and complex ball fields
d08b099239 Trac #33127: Fix warning about missing sage-site when SAGE_ROOT is removed after installation
f456256a0a Trac #33100: is_positive_definite returns wrong results
ca6ac0bcd0 Trac #33054: conda-forge (linux): primecount fails to install
347ee9349c Trac #32987: Deprecate sage.misc.misc.sage_makedirs
433e4d9eb2 Trac #30252: Make TensorProducts of finite-dimensional modules / vector spaces finite-dimensional
29b0897532 Trac #32965: tox / GH Actions: Add centos-7-devtoolset-gcc_{9,10,11}
15c8011dd0 (tag: 9.6.beta2) Updated SageMath version to 9.6.beta2

[Thierry]

Hi,

thanks for this beta release.

Patchbots are locked because of some numerical noise in a doctest
introduced in this ticket:

> de4f3f05e3 Trac #33388: Graph: accept keyword arguments for matrix
> constructor in methods returning a matrix

See https://patchbot.sagemath.org/ticket/0/

And see https://trac.sagemath.org/ticket/33427 for a fix.

Ciao,
Thierry

> --
> 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/ed451bd9-76ed-4229-8392-914a134aac15n%40googlegroups.com.

[Matthias Köppe]

On Monday, February 28, 2022 at 3:33:58 AM UTC-8 sage-goo...@lma.metelu.net wrote:

Patchbots are locked because

Please help by reviewing https://trac.sagemath.org/ticket/33233 "sage -t --baseline-stats-path"

[Samuel Lelièvre]

2022-02-28 00:37 UTC, Volker Braun:
>
> 1e8ba0aac4 (tag: 9.6.beta3) Updated SageMath version to 9.6.beta3

Thanks for releasing this new beta version.

On Cygwin, tried `make -j4`, and fpylll failed to build.
Perhaps one dependency is not properly indicated?
Running `make` again gave the `sagemath_doc_html`
usual failure on that machine, see #33139.

On Debian, tried `make fpylll` to see if the build would
also fail on fpylll, but instead it failed on numpy.
Perhaps another dependency not properly indicated?

That's all for now.

[Justin C. Walker]

> On Feb 27, 2022, at 16:37 , Volker Braun <vbrau...@gmail.com> wrote:
>
> 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

Thanks for your efforts in putting together these releases!

Beta3 built w/o problems on macOS 10.15.7 (2019 MBP (8-core Core i9). All tests (‘ptestlong’) again passed!!!

Beta3 built w/o problems on macOS 10.14.6 (2017 MBP (4-core Core i7). Testing (‘ptestlong’) yielded one failure, the previously reported “sageVAR” issue. Relevant piece of the log file attached.

Justin

[Kenji Iohara]

On my mac with Intelcore i5 and OS 12.2.1, it compiled from the first built, but 

with make ptest, it didn’t:

----------------------------------------------------------------------

sage -t --warn-long 100.6 --random-seed=265458366727004151924559316480357086755 src/doc/en/constructions/calculus.rst  # Timed out

sage -t --warn-long 100.6 --random-seed=265458366727004151924559316480357086755 src/sage/symbolic/relation.py  # Timed out

----------------------------------------------------------------------

Total time for all tests: 4203.0 seconds

    cpu time: 8279.1 seconds

    cumulative wall time: 15760.4 seconds

Features detected for doctesting: dvipng,graphviz,imagemagick,pandoc,pdf2svg,pdftocairo,sage.combinat,sage.geometry.polyhedron,sage.graphs,sage.groups,sage.plot,sage.rings.number_field,sage.rings.padics,sage.rings.real_double,sage.symbolic,sagemath_doc_html,sphinx

Pytest is not installed, skip checking tests that rely on it.

make: *** [ptest] Error 4

Here is its log file: 

[Emmanuel Charpentier]

FWIW, on Debian testing running on core i7 + 16 GB RAM, upgrading 9.6.beta1 to 9.6.beta3 and running ptestlong gets me one temporary failure :

sage -t --long --warn-long 245.4 --random-seed=61034896171412036953396016057425531267 src/sage/schemes/cyclic_covers/cycliccover_finite_field.py  # 1 doctest failed

and one permanent failure :

sage -t --long --warn-long 245.4 --random-seed=61034896171412036953396016057425531267 src/sage/graphs/graph.py  # 1 doctest failed

The second one seems to be numerical noise :

charpent@zen-book-flip:/usr/local/sage-9$ sage -t --long --warn-long 245.4 --random-seed=61034896171412036953396016057425531267 src/sage/graphs/graph.py  # 1 doctest failed
Running doctests with ID 2022-03-01-10-43-41-a97dd6df.
Git branch: develop
Using --optional=debian,dot2tex,fricas,gap_packages,libsemigroups,pip,sage,sage_spkg
Features to be detected: 4ti2,benzene,bliss,buckygen,conway_polynomials,csdp,database_cremona_ellcurve,database_cremona_mini_ellcurve,database_jones_numfield,database_knotinfo,dvipng,graphviz,imagemagick,jupymake,kenzo,latte_int,lrslib,mcqd,meataxe,pandoc,pdf2svg,pdftocairo,plantri,pynormaliz,python_igraph,rubiks,sage.combinat,sage.geometry.polyhedron,sage.graphs,sage.groups,sage.plot,sage.rings.number_field,sage.rings.padics,sage.rings.real_double,sage.symbolic,sage_numerical_backends_coin,sagemath_doc_html,sphinx,tdlib
Doctesting 1 file.
sage -t --long --warn-long 245.4 --random-seed=61034896171412036953396016057425531267 src/sage/graphs/graph.py
**********************************************************************
File "src/sage/graphs/graph.py", line 8586, in sage.graphs.graph.Graph.effective_resistance
Failed example:
    H.effective_resistance(1, 5, base_ring=RDF)
Expected:
    1.20000...
Got:
    1.1999999999999997
**********************************************************************
1 item had failures:
   1 of  22 in sage.graphs.graph.Graph.effective_resistance
    [1233 tests, 1 failure, 20.14 s]
----------------------------------------------------------------------
sage -t --long --warn-long 245.4 --random-seed=61034896171412036953396016057425531267 src/sage/graphs/graph.py  # 1 doctest failed
----------------------------------------------------------------------
Total time for all tests: 20.3 seconds
    cpu time: 18.7 seconds
    cumulative wall time: 20.1 seconds
Features detected for doctesting: sage.groups,sage.symbolic
Pytest is not installed, skip checking tests that rely on it.

but I have no easy explanation for the first (temporary) one, whose ptestlong.log record is as follows :

sage -t --long --warn-long 245.4 --random-seed=61034896171412036953396016057425531267 src/sage/schemes/cyclic_covers/cycliccover_finite_field.py
**********************************************************************
File "src/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 1139, in sage.schemes.cyclic_covers.cycliccover_finite_field.CyclicCover_finite_field.frobenius_polynomial
Failed example:
    CyclicCover(11, PolynomialRing(GF(1129), 'x')([-1] + [0]*(5-1) + [1])).frobenius_polynomial() # long time
Exception raised:
    Traceback (most recent call last):
      File "sage/misc/cachefunc.pyx", line 1943, in sage.misc.cachefunc.CachedMethodCaller.__call__ (build/cythonized/sage/misc/cachefunc.c:10410)
        return cache[k]
    KeyError: ((11,), ())

    During handling of the above exception, another exception occurred:

    Traceback (most recent call last):
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/doctest/forker.py", line 694, in _run
        self.compile_and_execute(example, compiler, test.globs)
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/doctest/forker.py", line 1088, in compile_and_execute
        exec(compiled, globs)
      File "<doctest sage.schemes.cyclic_covers.cycliccover_finite_field.CyclicCover_finite_field.frobenius_polynomial[23]>", line 1, in <module>
        CyclicCover(Integer(11), PolynomialRing(GF(Integer(1129)), 'x')([-Integer(1)] + [Integer(0)]*(Integer(5)-Integer(1)) + [Integer(1)])).frobenius_polynomial() # long time
      File "sage/misc/cachefunc.pyx", line 2310, in sage.misc.cachefunc.CachedMethodCallerNoArgs.__call__ (build/cythonized/sage/misc/cachefunc.c:13020)
        self.cache = f(self._instance)
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 1234, in frobenius_polynomial
        F = self.frobenius_matrix(self._N0)
      File "sage/misc/cachefunc.pyx", line 1948, in sage.misc.cachefunc.CachedMethodCaller.__call__ (build/cythonized/sage/misc/cachefunc.c:10546)
        w = self._instance_call(*args, **kwds)
      File "sage/misc/cachefunc.pyx", line 1824, in sage.misc.cachefunc.CachedMethodCaller._instance_call (build/cythonized/sage/misc/cachefunc.c:10012)
        return self.f(self._instance, *args, **kwds)
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 1072, in frobenius_matrix
        FrobP = _frobenius_matrix_p(self._N0)
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 1068, in _frobenius_matrix_p
        ] = self._frob(i, j + self._epsilon * self._r, N0)
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 994, in _frob
        self._initialize_fat_horizontal(
      File "/usr/local/sage-9/local/var/lib/sage/venv-python3.9/lib/python3.9/site-packages/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 738, in _initialize_fat_horizontal
        DH = [elt[0, 0] for elt in interval_products(D0, D1, targets)]
      File "sage/schemes/hyperelliptic_curves/hypellfrob.pyx", line 140, in sage.schemes.hyperelliptic_curves.hypellfrob.interval_products (build/cythonized/sage/schemes/hyperelliptic_curves/hypellfrob.cpp:3999)
        sig_on()
    SystemError: calling remove_from_pari_stack() inside sig_on()
**********************************************************************
1 item had failures:
   1 of  60 in sage.schemes.cyclic_covers.cycliccover_finite_field.CyclicCover_finite_field.frobenius_polynomial

Of note : this ptestlong run, on a lightly loaded laptop freshly rebooted, entailed a very high use of memory : it used about 18 GB of swap file above the 16 GB of RAM. I don’t remember seeing such an high RAM usage during ptestlong of the previous versions.

HTH,

​

[Samuel Lelièvre]

2022-03-01 10:00 UTC, Emmanuel Charpentier:

>
> FWIW, on Debian testing running on core i7 + 16 GB RAM,
> upgrading 9.6.beta1 to 9.6.beta3 and running ptestlong
>gets me one temporary failure :
>

> ... src/sage/schemes/cyclic_covers/cycliccover_finite_field.py # 1 doctest failed
>
> and one permanent failure :
>
> ... src/sage/graphs/graph.py # 1 doctest failed

>
> The second one seems to be numerical noise :
>

> **********************************************************************
> File "src/sage/graphs/graph.py", line 8586, in sage.graphs.graph.Graph.effective_resistance
> Failed example:
> H.effective_resistance(1, 5, base_ring=RDF)
> Expected:
> 1.20000...
> Got:
> 1.1999999999999997
> **********************************************************************

Taken care of by Sage Trac ticket 33427:

numerical noise in effective_resistance involving RDF
https://trac.sagemath.org/ticket/33427

where a fix already has been positively reviewed.

> but I have no easy explanation for the first (temporary) one,
> whose ptestlong.log record is as follows :
>

> **********************************************************************
> File "src/sage/schemes/cyclic_covers/cycliccover_finite_field.py", line 1139,
> in sage.schemes.cyclic_covers.cycliccover_finite_field.CyclicCover_finite_field.frobenius_polynomial
> Failed example:
> CyclicCover(11, PolynomialRing(GF(1129), 'x')([-1] + [0]*(5-1) + [1])).frobenius_polynomial() # long time
> Exception raised:
> Traceback (most recent call last):
> File "sage/misc/cachefunc.pyx", line 1943,
> in sage.misc.cachefunc.CachedMethodCaller.__call__
> (build/cythonized/sage/misc/cachefunc.c:10410)
> return cache[k]
> KeyError: ((11,), ())
>
> During handling of the above exception, another exception occurred:

> ...

> SystemError: calling remove_from_pari_stack() inside sig_on()
> **********************************************************************
> 1 item had failures:
> 1 of 60 in sage.schemes.cyclic_covers.cycliccover_finite_field.CyclicCover_finite_field.frobenius_polynomial
>
> Of note : this ptestlong run, on a lightly loaded laptop freshly rebooted,
> entailed a very high use of memory : it used about 18 GB of swap file

> above the 16 GB of RAM. I don’t remember seeing such a high

> RAM usage during ptestlong of the previous versions.

Was that while testing the same file

src/sage/schemes/cyclic_covers/cycliccover_finite_field.py

that ended up causing a system error? Or you could have hit
something similar to the issue reported at Sage Trac ticket 33363:

(Too) long doctest in sage/matrix/matrix_integer_dense.pyx
https://trac.sagemath.org/ticket/33363

where a test could sometimes take extremely long (the ticket
does not comment on the memory usage in that case).

[Eric Gourgoulhon]

On Ubuntu 20.04 running on Xeon E5-2623 + 16 GB RAM:

- incremental build (-j16) from Sage 9.6.beta2 with system python (3.8.10): OK

- make ptestlong --> "All tests passed!"

Eric.

[Volker Braun]

The SystemError: calling remove_from_pari_stack() inside sig_on() in cycliccover_finite_field.py is https://trac.sagemath.org/ticket/33304

[Matthias Köppe]

Tests ran at https://github.com/sagemath/sage/runs/5353144443 for Linux

ubuntu (from -trusty all the way to -jammy): Clean

- also bionic-i386: Clean

debian (from -stretch to -sid): Clean

... except debian-bullseye-standard (https://github.com/sagemath/sage/runs/5353145196?check_suite_focus=true) - severe numerics issues, already tracked at https://trac.sagemath.org/ticket/31621 
linuxmint (from -17  to -20.3): Clean

... except linuxmint-19-standard, 19.3-standard (same numerics issues)

fedora (from -26 to -36): Clean

... except fedora-28-standard (https://github.com/sagemath/sage/runs/5353145559?check_suite_focus=true) groebner_fan

... except fedora-34-standard (https://github.com/sagemath/sage/runs/5353145727?check_suite_focus=true) trouble with giac

... except fedora-36-standard (https://github.com/sagemath/sage/runs/5353145792?check_suite_focus=true) sage.quadratic_forms.binary_qf.BinaryQF.solve_integer ZeroDivisionError

centos (-7 and -stream-8): Clean

gentoo: Clean

archlinux: Clean

opensuse (-15.3): Clean; -tumbleweed: fplll issue - https://trac.sagemath.org/ticket/33448

slackware: Clean

[Matthias Köppe]

Also cygwin-standard (https://github.com/sagemath/sage/actions/runs/1907624649) looks fine.

[G. M.-S.]

Thanks, Volker.

On two Apple Silicon M1 Macs with macOS 11.6.4 (Big Sur) and macOS 12.2.1 (Monterey), with Homebrew up to date, after

source .homebrew-build-env 

./configure

make succeeds.

And make ptestlong gives only 2 errors:

----------------------------------------------------------------------
sage -t --long --warn-long 25.2 --random-seed=300585988638342567798698437588651751284 src/sage/graphs/graph.py  # 1 doctest failed
sage -t --long --warn-long 25.2 --random-seed=300585988638342567798698437588651751284 src/sage/matrix/matrix_gf2e_dense.pyx  # 1 doctest failed
----------------------------------------------------------------------

namely

**********************************************************************
File "src/sage/graphs/graph.py", line 8586, in sage.graphs.graph.Graph.effective_resistance
Failed example:
    H.effective_resistance(1, 5, base_ring=RDF)
Expected:
    1.20000...
Got:

    1.1999999999999995
**********************************************************************

due to numerical noise, see https://trac.sagemath.org/ticket/33427 (already closed)

and

**********************************************************************
File "src/sage/matrix/matrix_gf2e_dense.pyx", line 524, in sage.matrix.matrix_gf2e_dense.Matrix_gf2e_dense._multiply_karatsuba
Failed example:
    A._multiply_karatsuba(B) == A._multiply_classical(B)
Expected:
    True
Got:
    False
**********************************************************************

due to the fact that M4RIE fails on this processor (see https://bitbucket.org/malb/m4rie/issues/23/trying-to-compile-on-apple-m1).

On Monterey there were as usual 5 alerts "Pyhton quit unexpectedly", which were ignored.

Guillermo

[Emmanuel Charpentier]

Le mardi 1 mars 2022 à 11:59:22 UTC+1, Samuel Lelievre a écrit :

[ Snip… ]

Cant’ tell…

Or you could have hit
something similar to the issue reported at Sage Trac ticket 33363:

(Too) long doctest in sage/matrix/matrix_integer_dense.pyx
https://trac.sagemath.org/ticket/33363

where a test could sometimes take extremely long (the ticket
does not comment on the memory usage in that case).

Possibly.

FWIW, I get exactly the same result on another, slightly smaller, machine (Debian testing running on core i5 + 8 GB RAM) : temporary failure on cycliccover_finite_field.py, permanent failure on graph.py, (more moderate) overconsumption of the swap file.

HTH,

​

[Samuel Lelièvre]

2022-03-03 10:29 UTC+1, Emmanuel Charpentier:

>
> FWIW, I get exactly the same result on another,
> slightly smaller, machine (Debian testing running
> on core i5 + 8 GB RAM) : temporary failure on
> cycliccover_finite_field.py, permanent failure
> on graph.py, (more moderate) overconsumption
> of the swap file.

Possibly related tickets:

- Random failure in cycliccover_finite_field.py
https://trac.sagemath.org/ticket/33304
- Random failure in cycliccover_finite_field.py
https://trac.sagemath.org/ticket/30419