sanurielf / kvxopt Goto Github PK
View Code? Open in Web Editor NEWThis project forked from cvxopt/cvxopt
KVXOPT -- Python Software for Convex Optimization containing more wrappers suite-sparse.
License: Other
This project forked from cvxopt/cvxopt
KVXOPT -- Python Software for Convex Optimization containing more wrappers suite-sparse.
License: Other
Great package! Just wonder whether the KLU solver supports custom ordering instead of the default AMD? I noticed cholmod supports this feature *1, but awkwardly I cannot find any documents about this for KLU.
BTW, I tried to search KLU as key words in this repo but cannot find any code related. Where should I find the relevant doc and source code? Sorry for this dumb question and thank you in advance!
Again, great package! KLU is FAST!
Hello Uriel,
I hope all is well with you! I'm seeing some weird CI error from the ANDES Conda-forge builder.
Here's the log of a recent CI output:
2024-02-04T21:00:18.2055603Z
2024-02-04T21:00:18.2056437Z The following NEW packages will be INSTALLED:
2024-02-04T21:00:18.2056725Z
2024-02-04T21:00:18.2057418Z _libgcc_mutex: 0.1-conda_forge conda-forge
2024-02-04T21:00:18.2058129Z _openmp_mutex: 4.5-2_gnu conda-forge
2024-02-04T21:00:18.2058596Z andes: 1.9.0-pyhd8ed1ab_0 local
2024-02-04T21:00:18.2059053Z brotli: 1.1.0-hd590300_1 conda-forge
2024-02-04T21:00:18.2059525Z brotli-bin: 1.1.0-hd590300_1 conda-forge
2024-02-04T21:00:18.2059974Z bzip2: 1.0.8-hd590300_5 conda-forge
2024-02-04T21:00:18.2060455Z ca-certificates: 2024.2.2-hbcca054_0 conda-forge
2024-02-04T21:00:18.2060930Z certifi: 2024.2.2-pyhd8ed1ab_0 conda-forge
2024-02-04T21:00:18.2061398Z chardet: 5.2.0-py312h7900ff3_1 conda-forge
2024-02-04T21:00:18.2061862Z colorama: 0.4.6-pyhd8ed1ab_0 conda-forge
2024-02-04T21:00:18.2062343Z coloredlogs: 15.0.1-pyhd8ed1ab_3 conda-forge
2024-02-04T21:00:18.2062806Z contourpy: 1.2.0-py312h8572e83_0 conda-forge
2024-02-04T21:00:18.2063320Z cycler: 0.12.1-pyhd8ed1ab_0 conda-forge
2024-02-04T21:00:18.2063785Z dill: 0.3.8-pyhd8ed1ab_0 conda-forge
2024-02-04T21:00:18.2064243Z dsdp: 5.8-hd9d9efa_1203 conda-forge
2024-02-04T21:00:18.2064698Z et_xmlfile: 1.1.0-pyhd8ed1ab_0 conda-forge
2024-02-04T21:00:18.2065152Z fftw: 3.3.10-nompi_hc118613_108 conda-forge
2024-02-04T21:00:18.2065876Z fonttools: 4.47.2-py312h98912ed_0 conda-forge
2024-02-04T21:00:18.2066385Z freetype: 2.12.1-h267a509_2 conda-forge
2024-02-04T21:00:18.2066846Z glpk: 5.0-h445213a_0 conda-forge
2024-02-04T21:00:18.2067299Z gmp: 6.3.0-h59595ed_0 conda-forge
2024-02-04T21:00:18.2067748Z gsl: 2.7-he838d99_0 conda-forge
2024-02-04T21:00:18.2068190Z humanfriendly: 10.0-pyhd8ed1ab_6 conda-forge
2024-02-04T21:00:18.2068653Z icu: 73.2-h59595ed_0 conda-forge
2024-02-04T21:00:18.2069113Z kiwisolver: 1.4.5-py312h8572e83_1 conda-forge
2024-02-04T21:00:18.2069581Z kvxopt: 1.3.2.0-py312h5c6e270_2 conda-forge
... (omitted)
2024-02-04T21:00:18.2103747Z
2024-02-04T21:00:27.1061043Z Preparing transaction: ...working... done
2024-02-04T21:00:30.4037169Z Verifying transaction: ...working... done
2024-02-04T21:00:42.1839036Z Executing transaction: ...working... done
2024-02-04T21:00:42.2230809Z export PREFIX=/home/conda/feedstock_root/build_artifacts/andes_1707080235100/_test_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_
2024-02-04T21:00:42.2232005Z export SRC_DIR=/home/conda/feedstock_root/build_artifacts/andes_1707080235100/test_tmp
2024-02-04T21:00:42.6251963Z import: 'andes'
2024-02-04T21:00:43.2720397Z import: 'andes.core'
2024-02-04T21:00:43.2721665Z import: 'andes'
2024-02-04T21:00:43.2728591Z import: 'andes.core'
2024-02-04T21:00:43.3655459Z + pip check
2024-02-04T21:00:44.0219538Z andes 1.9.0 has requirement kvxopt>=1.3, but you have kvxopt 0.0.0.
2024-02-04T21:00:44.2317708Z WARNING: Tests failed for andes-1.9.0-pyhd8ed1ab_0.conda - moving package to /home/conda/feedstock_root/build_artifacts/broken
For some reason, the installed KVXOPT 1.3.2.0 is recognized as 0.0.0. I wonder if this is due to the use of non-conventional, four-digit semver by KVXOPT.
When running the tests with valgrind, the following memory leaks were found (repeated several times):
==697== 8 bytes in 1 blocks are still reachable in loss record 10 of 4,759
==697== at 0x483B7F3: malloc (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so)
==697== by 0x9700A2B: cholmod_l_malloc (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x96FBCF9: cholmod_l_allocate_work (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x9714A41: cholmod_l_analyze_p2 (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x9682F92: symbolic (cholmod.c:274)
==697== by 0x178751: _PyMethodDef_RawFastCallKeywords (call.c:690)
==697== by 0x1788DA: _PyCFunction_FastCallKeywords (call.c:730)
==697== by 0x222D3A: call_function (ceval.c:4568)
==697== by 0x22ED3F: _PyEval_EvalFrameDefault (ceval.c:3093)
==697== by 0x223E0F: PyEval_EvalFrameEx (ceval.c:547)
==697== by 0x224892: _PyEval_EvalCodeWithName (ceval.c:3930)
==697== by 0x178075: _PyFunction_FastCallKeywords (call.c:433)
and
==697== 40 bytes in 1 blocks are still reachable in loss record 98 of 4,759
==697== at 0x483B723: malloc (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so)
==697== by 0x483E017: realloc (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so)
==697== by 0x9A182AC: ??? (in /usr/lib/x86_64-linux-gnu/libgomp.so.1.0.0)
==697== by 0x9A27581: ??? (in /usr/lib/x86_64-linux-gnu/libgomp.so.1.0.0)
==697== by 0x9A1E8E0: GOMP_parallel (in /usr/lib/x86_64-linux-gnu/libgomp.so.1.0.0)
==697== by 0x97502D8: ??? (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x9753D27: cholmod_l_super_numeric (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x971645F: cholmod_l_factorize_p (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x971671A: cholmod_l_factorize (in /usr/lib/x86_64-linux-gnu/libcholmod.so.3.0.14)
==697== by 0x96813C7: numeric (cholmod.c:359)
==697== by 0x178751: _PyMethodDef_RawFastCallKeywords (call.c:690)
==697== by 0x1788DA: _PyCFunction_FastCallKeywords (call.c:730)
Just want to check if 3.11 is officially supported.
MWE:
from kvxopt import spdiag, matrix
c = spdiag(matrix(range(1,1000)))
c[0:10000:30]
Segfaults on Python 3.9.7 on Ubuntu 20.04. KVXOPT version is 1.2.7.3. The code works fine with CVXOPT 1.2.7.
This is a tremendous project, it solved the problem I was having with my Apple M1 processor.
However, I have run into a problem when generating a Docker image. The image creation fails when adding this package, as it does not have wheels for Linux ARM. Is there a possibility to add wheels for Linux aarch64?
Hi, I'm just curious as to why version 1.2.7.5 is, as of writing, only available for OSX ARM64.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.