On Mon, Jul 1, 2024 at 4:38 AM Oscar Benjamin <oscar.j.benja...@gmail.com> wrote: > > Hi all, > > Some issues were reported with the SymPy 1.13.0rc1 release candidate. > Since then I pushed 1.13.0rc2 and 1.13.0rc3 and I'm about to push > 1.13.0rc4. Thanks to everyone who tested the release candidate and > reported any problems. > > I think this is a list of issues and PRs that have been included (or > will be in rc4): > > https://github.com/sympy/sympy/issues/26731 > https://github.com/sympy/sympy/issues/26688 > https://github.com/sympy/sympy/issues/26690 > https://github.com/sympy/sympy/pull/26627 > https://github.com/sympy/sympy/issues/26692 > > There is one outstanding thing that I would like to resolve before the > final release which is the handling of the python-flint version. > > Currently the situation with rc3 is that if you install python-flint > 0.6.0 or hypothetically in future 0.6.1, 0.6.2, etc then when you > import sympy it will automatically use python-flint. If you have a > newer version of python-flint like 0.7.0 (which does not exist yet) or > an older version like 0.5.0 then it will print a warning and default > to not using python-flint (falling back on gmpy2 or just on using > Python types). > > This warning looks like: > > >>> import sympy > ~/sympy/sympy/external/gmpy.py:126: UserWarning: python-flint 0.5.0 > is installed but only version 0.6.* will be used by default. Falling > back to other ground types. Use SYMPY_GROUND_TYPES=flint to force the > use of python-flint. > warn(f"python-flint {_flint_version} is installed but only version " > > The reason for not using python-flint 0.5.0 is that SymPy 1.13 uses > some features that were added in 0.6.0. The reason for not using > python-flint 0.7.0 is that hypothetically future releases of > python-flint might make significant changes making them not as fully > compatible with SymPy 1.13. This is only a hypothetical though and > there are not any plans for python-flint to break in a backwards > compatible way. > > There are two questions here: > > 1. Should SymPy 1.13 just use newer versions of python-flint anyway? > 2. If not, should SymPy warn about newer versions of python-flint? > > I think that the answer to the warning question is that no there > should be no warning. These warnings end up being annoying for > downstream people who don't really care. That might mean that in > future some people are surprised by the fact that they have > python-flint installed but SymPy does not use it. I think that is > better than annoying other people with the warning. > > For the question about using newer versions of python-flint I think > that at least SymPy 1.13 should be able to use a few future versions > of python-flint but I am not sure what is a reasonable future range of > versions. Possibly SymPy could use any python-flint >= 0.6.0, < 1.0 > meaning that python-flint should not break any compatibility for SymPy > until it reaches version 1.0. We would need something somewhere to > test new versions of python-flint against potentially old versions of > SymPy though. I could add something in python-flint's CI to make sure > that it tests all changes against SymPy 1.13. > > I think my preference is: > > 1. Use python-flint >= 0.6.0, < 1.0 automatically. > 2. Don't use python-flint < 0.5.0 or >= 1.0 and don't give any warning about > it. > 3. Unless the SYMPY_GROUND_TYPES=flint variable is set in which case > use python-flint regardless of the version. > 4. Keep future versions of python-flint compatible with SymPy 1.13 > until python-flint 1.0. > 5. Once SymPy 1.13 is released add a job in python-flint CI to test > against SymPy 1.13. > > Does anyone have any views on this?
I guess the most ideal would be if we could synchronize with the upstream and have them promise that any release with a breaking change would bump, say, the major version number. That we we could pin <1.0 and be sure that will be the right thing. Also presumably if an issue does arise we can release a patch release that adds a pin, so maybe it isn't a huge deal as long as we are able to stay on top of it. Aaron Meurer > > As far as I know this is the only outstanding item before final release. > > -- > Oscar > > On Sat, 8 Jun 2024 at 08:34, Aaron Meurer <asmeu...@gmail.com> wrote: > > > > Another important change in this release is that Rational (and > > Integer) and Float objects no longer compare equal to one another with > > ==. > > > > Previously: > > > > >>> Integer(1) == Float(1.0) > > True > > >>> Rational(1, 2) == Float(0.5) > > True > > > > Now: > > > > >>> Integer(1) == Float(1.0) > > False > > >>> Rational(1, 2) == Float(0.5) > > False > > > > This was previously applied inconsistently. For example, it always > > worked this way inside of expressions > > > > >>> x**2 == x**2.0 > > False > > > > The motivation for this change is that == means structural, not > > mathematical equality in SymPy, and rational and floating-point > > numbers have different properties that make it inappropriate to > > consider them to be equivalent. It also fixes a long standing issue > > where Floats and Rationals would compare equal but not hash to the > > same values. > > > > This is a change that I expect could affect upstream projects, so if > > you depend on SymPy I would suggest trying 1.13rc1 out and letting us > > know if it affects you in a serious way. > > > > Aaron Meurer > > > > On Thu, Jun 6, 2024 at 9:08 AM Oscar Benjamin > > <oscar.j.benja...@gmail.com> wrote: > > > > > > Hi all, > > > > > > I have just pushed SymPy 1.13.0rc1 to PyPI. This is a prerelease that > > > is being made available for early testing. > > > > > > You can install this with: > > > > > > pip install sympy==1.13.0rc1 > > > > > > Or alternatively: > > > > > > pip install --upgrade --pre sympy > > > > > > The release files can also be downloaded from GitHub: > > > > > > https://github.com/sympy/sympy/releases/tag/1.13.0rc1 > > > > > > I expect that a full release of 1.13.0 will be made fairly soon and I > > > don't anticipate that there would be significant changes between rc1 > > > and final release. > > > > > > There are many changes since the previous release 1.12 which was just > > > over a year ago. you can find the release notes here although they > > > still need cleaning up: > > > > > > https://github.com/sympy/sympy/wiki/Release-Notes-for-1.13 > > > > > > One particular highlight of this release is that it can use > > > python-flint to accelerate some polynomial and other operations. It > > > would be great if people could test this out by installing > > > python-flint as well: > > > > > > pip install python-flint==0.6 > > > > > > You can set the environment variable SYMPY_GROUND_TYPES to any of > > > python, gmpy2 or flint to control which backend is used. The most > > > easily noticeable difference python-flint makes for now is speeding up > > > operations with univariate polynomials: > > > > > > $ SYMPY_GROUND_TYPES=python isympy > > > ... > > > IPython console for SymPy 1.13.0rc1 (Python 3.12.0-64-bit) (ground > > > types: python) > > > ... > > > In [1]: p = Poly(x + 1) > > > > > > In [2]: %time factor_list(p ** 100) > > > CPU times: user 117 ms, sys: 183 µs, total: 117 ms > > > Wall time: 115 ms > > > Out[2]: (1, [(Poly(x + 1, x, domain='ZZ'), 100)]) > > > > > > $ SYMPY_GROUND_TYPES=flint isympy > > > ... > > > IPython console for SymPy 1.13.0rc1 (Python 3.12.0-64-bit) (ground > > > types: flint) > > > ... > > > In [1]: p = Poly(x + 1) > > > > > > In [2]: %time factor_list(p ** 100) > > > CPU times: user 4.05 ms, sys: 16 µs, total: 4.07 ms > > > Wall time: 4.01 ms > > > Out[2]: (1, [(Poly(x + 1, x, domain='ZZ'), 100)]) > > > > > > That particular example shows a 30x speedup. Future releases of SymPy > > > will use flint for more operations so this is still in development but > > > it would be great for people to test this because although I have > > > tested SymPy with python-flint a lot I am not sure how many other > > > people have. > > > > > > As always please report any problems with the release either here or > > > on Github. There are so many changes in this release that it is > > > basically guaranteed that there will be some issues discovered > > > afterwards but it would be great if we could pick up on some of them > > > before the final release rather than after. > > > > > > ## Authors > > > > > > The following people contributed at least one patch to this release > > > (names are > > > given in alphabetical order by last name). A total of 146 people > > > contributed to this release. People with a * by their names contributed a > > > patch for the first time for this release; 100 people contributed > > > for the first time for this release. > > > > > > Thanks to everyone who contributed to this release! > > > > > > - Daan Koning (he/him)* > > > - Anton Akhmerov > > > - Han Wei Ang* > > > - anutosh491 > > > - Isidora Araya* > > > - atharvParlikar* > > > - Oscar Benjamin > > > - Evandro Bernardes > > > - Anurag Bhat > > > - Francesco Bonazzi > > > - Augusto Borges* > > > - João Bravo* > > > - Sam Brockie > > > - Pontus von Brömssen > > > - Steven Burns* > > > - Liwei Cai* > > > - Corey Cerovsek* > > > - HeeJae Chang* > > > - Abhishek Chaudhary* > > > - Abhinav Cillanki* > > > - codecruisader* > > > - Costor > > > - Björn Dahlgren > > > - Saikat Das* > > > - Ethan DeGuire* > > > - Theodore Dias* > > > - dodo* > > > - fazledyn-or* > > > - Emile Fourcini* > > > - Mark van Gelder* > > > - Riccardo Di Girolamo* > > > - Pascal Gitz* > > > - Aaron Gokaslan > > > - Kishore Gopalakrishnan > > > - Michael Greminger > > > - Oscar Gustafsson > > > - Nick Harder* > > > - Zac Hatfield-Dodds* > > > - Le Cong Minh Hieu* > > > - Warren Jacinto* > > > - Jerry James > > > - Hwayeon Kang* > > > - Samith Karunathilake* > > > - Harsh Kasat* > > > - Johannes Kasimir* > > > - Kaustubh* > > > - Steve Kieffer > > > - Evelyn King > > > - Matthias Köppe > > > - K. Kraus* > > > - Abhishek kumar* > > > - Abhishek Kumar* > > > - Shishir Kushwaha* > > > - S.Y. Lee > > > - Raphael Lehner* > > > - Matthias Liesenfeld* > > > - Qijia Liu > > > - Sam Lubelsky > > > - Fabio Luporini* > > > - Megan Ly > > > - Nikhil Maan > > > - Colin B. Macdonald > > > - Oriel Malihi* > > > - Mohak Malviya* > > > - Martin Manns* > > > - Tirthankar Mazumder > > > - Ehren Metcalfe > > > - Aaron Meurer > > > - mohammedouahman* > > > - Lukas Molleman* > > > - John Möller* > > > - Suman mondal* > > > - Jason Moore > > > - Harry Mountain* > > > - Arnab Nandi* > > > - Harrison Oates* > > > - omahs* > > > - Victory Omole* > > > - Julien Palard > > > - Bobby Palmer* > > > - Ishan Pandhare > > > - Jay Patankar* > > > - Prey Patel* > > > - Abhishek Patidar > > > - Idan Pazi* > > > - Ivan Petukhov* > > > - philwillnyc* > > > - George Pittock* > > > - platypus* > > > - Advait Pote > > > - James A. Preiss* > > > - Dean Price* > > > - Psycho-Pirate > > > - Sophia Pustova* > > > - Baiyuan Qiu* > > > - Ravindu-Hirimuthugoda* > > > - Juha Remes > > > - Mikhail Remnev* > > > - Tilo Reneau-Cardoso* > > > - Matthias Rettl* > > > - Mohamed Rezk* > > > - Konstantinos Riganas* > > > - Sam Ritchie* > > > - Richard Rodenbusch* > > > - Pedro Rosa* > > > - Alberto Jiménez Ruiz* > > > - Bhavik Sachdev* > > > - Taylan Sahin* > > > - Saicharan > > > - Pablo Galindo Salgado > > > - Richard Samuel* > > > - Davide Sandonà > > > - Raj Sapale > > > - Gilles Schintgen > > > - Nico Schlömer > > > - Hanspeter Schmid > > > - Alexis Schotte* > > > - Vladimir Sereda* > > > - Ankit Kumar Singh* > > > - Sachin Singh* > > > - Maciej Skórski* > > > - Chris Smith > > > - Henrique Soares* > > > - Vivek Soni* > > > - Timo Stienstra > > > - Grace Su* > > > - Kalevi Suominen > > > - Daiki Takahashi > > > - Diane Tchuindjo > > > - Gerald Teschl* > > > - Seb Tiburzio* > > > - Tommaso Vaccari* > > > - Geetika Vadali* > > > - Viraj Vekaria* > > > - Laurence Warne* > > > - Daniel Weindl* > > > - James Whitehead* > > > - Lorenz Winkler* > > > - Congxu Yang* > > > - 袁野 (Yuan Ye)* > > > - Lauren Yim* > > > - Zedmat* > > > - Shuai Zhou* > > > - Zhenxu Zhu* > > > - Alexander Zhura* > > > - zzc* > > > > > > The SHA-256 hashes for the release are: > > > > > > 37392a943e14db2886543141c2121330887b6f834604a740a476f4b516315d11 > > > sympy-1.13.0rc1.tar.gz > > > f6fce4b6b1ef4d89631fe81380f4e562e86615d2aef0bc3a17c4f1d0e3cd5dab > > > sympy-1.13.0rc1-py3-none-any.whl > > > b3b729284e791f4af8d792c67d7dbf30289563ff2acf3d6650c7eac7284e7f6d > > > sympy-docs-html-1.13.0rc1.zip > > > 14436f97e203918ca9655491e79b412b3452e46e3fefa61c5f3d544c874c5788 > > > sympy-docs-pdf-1.13.0rc1.pdf > > > > > > -- > > > Oscar > > > > > > -- > > > You received this message because you are subscribed to the Google Groups > > > "sympy" group. > > > To unsubscribe from this group and stop receiving emails from it, send an > > > email to sympy+unsubscr...@googlegroups.com. > > > To view this discussion on the web visit > > > https://groups.google.com/d/msgid/sympy/CAHVvXxRUsh8CDHkJDc3dsYj4V_FkxWxa9MQpLSzw497bVni_9Q%40mail.gmail.com. > > > > -- > > You received this message because you are subscribed to the Google Groups > > "sympy" group. > > To unsubscribe from this group and stop receiving emails from it, send an > > email to sympy+unsubscr...@googlegroups.com. > > To view this discussion on the web visit > > https://groups.google.com/d/msgid/sympy/CAKgW%3D6%2BPQcwkJaGHXSMtxTwPGzQMzVPYdv4QVt7cPiqEDeQJpg%40mail.gmail.com. > > -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sympy+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/sympy/CAHVvXxRSakG0A8%2BneCx3NRofBiQWkm8NoVo9z%2BFBnKnTg_OSgw%40mail.gmail.com. -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to sympy+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sympy/CAKgW%3D6%2B0-D2PTYDQUaTb2evFN68Hhu7%2BLnsW0-oLz6oSFnT7ng%40mail.gmail.com.
