Package Details: python-cvxpy 1.1.17-1

Git Clone URL: https://aur.archlinux.org/python-cvxpy.git (read-only, click to copy)
Package Base: python-cvxpy
Description: A domain-specific language for modeling convex optimization problems
Upstream URL: http://github.com/cvxgrp/cvxpy/
Licenses: Apache
Submitter: None
Maintainer: thrasibule
Last Packager: thrasibule
Votes: 8
Popularity: 0.093831
First Submitted: 2016-09-15 09:01
Last Updated: 2021-11-16 16:12

Latest Comments

doublesack commented on 2021-08-31 13:00

Fixed python-osqp. Tests should work just fine now.

JohnRobson commented on 2021-08-24 15:58

FAILED cvxpy/tests/test_benchmarks.py::TestBenchmarks::test_least_squares - cvxpy.error.SolverError: The solver OSQP is not installed.

bidskii commented on 2021-08-12 05:46

This fails to build due to failing tests. If python-cvxopt is installed the failing tests will pass.

=================================================================================================================================== FAILURES ===================================================================================================================================
__________________________________________________________________________________________________________________ TestAllSolvers.test_mixed_integer_behavior __________________________________________________________________________________________________________________
cvxpy.error.SolverError: 

                    You need a mixed-integer solver for this model. Refer to the documentation
                        https://www.cvxpy.org/tutorial/advanced/index.html#mixed-integer-programs
                    for discussion on this topic.

                    Quick fix 1: if you install the python package CVXOPT (pip install cvxopt),
                    then CVXPY can use the open-source mixed-integer solver `GLPK`.

                    Quick fix 2: you can explicitly specify solver='ECOS_BB'. This may result
                    in incorrect solutions and is not recommended.


During handling of the above exception, another exception occurred:

self = <cvxpy.tests.test_conic_solvers.TestAllSolvers testMethod=test_mixed_integer_behavior>

    def test_mixed_integer_behavior(self) -> None:
        x = cp.Variable(2, name='x', integer=True)
        objective = cp.Minimize(cp.sum(x))
        prob = cp.Problem(objective, [x >= 0])
        if INSTALLED_MI_SOLVERS == [cp.ECOS_BB]:
            with self.assertRaisesRegex(cp.error.SolverError, "You need a mixed-integer"
                                                              "solver for this model.*"):
>               prob.solve()
E               AssertionError: "You need a mixed-integersolver for this model.*" does not match "
E               
E                                   You need a mixed-integer solver for this model. Refer to the documentation
E                                       https://www.cvxpy.org/tutorial/advanced/index.html#mixed-integer-programs
E                                   for discussion on this topic.
E               
E                                   Quick fix 1: if you install the python package CVXOPT (pip install cvxopt),
E                                   then CVXPY can use the open-source mixed-integer solver `GLPK`.
E               
E                                   Quick fix 2: you can explicitly specify solver='ECOS_BB'. This may result
E                                   in incorrect solutions and is not recommended.
E                               "

cvxpy/tests/test_conic_solvers.py:1328: AssertionError
____________________________________________________________________________________________________________________ TestECOS_BB.test_ecos_bb_explicit_only ____________________________________________________________________________________________________________________
cvxpy.error.SolverError: 

                    You need a mixed-integer solver for this model. Refer to the documentation
                        https://www.cvxpy.org/tutorial/advanced/index.html#mixed-integer-programs
                    for discussion on this topic.

                    Quick fix 1: if you install the python package CVXOPT (pip install cvxopt),
                    then CVXPY can use the open-source mixed-integer solver `GLPK`.

                    Quick fix 2: you can explicitly specify solver='ECOS_BB'. This may result
                    in incorrect solutions and is not recommended.


During handling of the above exception, another exception occurred:

self = <cvxpy.tests.test_conic_solvers.TestECOS_BB testMethod=test_ecos_bb_explicit_only>

    def test_ecos_bb_explicit_only(self) -> None:
        """Test that ECOS_BB isn't chosen by default.
        """
        x = cp.Variable(1, name='x', integer=True)
        objective = cp.Minimize(cp.sum(x))
        prob = cp.Problem(objective, [x >= 0])
        if INSTALLED_MI_SOLVERS != [cp.ECOS_BB]:
            prob.solve()
            assert prob.solver_stats.solver_name != cp.ECOS_BB
        else:
            with self.assertRaisesRegex(cp.error.SolverError, "You need a mixed-integer"
                                                              "solver for this model.*"):
>               prob.solve()
E               AssertionError: "You need a mixed-integersolver for this model.*" does not match "
E               
E                                   You need a mixed-integer solver for this model. Refer to the documentation
E                                       https://www.cvxpy.org/tutorial/advanced/index.html#mixed-integer-programs
E                                   for discussion on this topic.
E               
E                                   Quick fix 1: if you install the python package CVXOPT (pip install cvxopt),
E                                   then CVXPY can use the open-source mixed-integer solver `GLPK`.
E               
E                                   Quick fix 2: you can explicitly specify solver='ECOS_BB'. This may result
E                                   in incorrect solutions and is not recommended.
E                               "

cvxpy/tests/test_conic_solvers.py:1348: AssertionError

bidskii commented on 2020-12-02 22:04

I am getting this error while installing v1.1.7-1. Is this clash intended?

error: failed to commit transaction (conflicting files)
python-cvxpy: /usr/lib/python3.9/site-packages/examples/__init__.py exists in filesystem (owned by python-filetype)
python-cvxpy: /usr/lib/python3.9/site-packages/examples/__pycache__/__init__.cpython-39.opt-1.pyc exists in filesystem (owned by python-filetype)
python-cvxpy: /usr/lib/python3.9/site-packages/examples/__pycache__/__init__.cpython-39.pyc exists in filesystem (owned by python-filetype)
Errors occurred, no packages were upgraded.

doublesack commented on 2019-07-12 09:51

I don't think that openblas-lapack is required for this package. At least for me, selftests work with openblas, cblas and lapack...

@steinbuch: I don't know where your problem is coming from. It installs fine if you change the PKGBUILD to use the latest cvxpy version (1.0.24)

steinbuch commented on 2019-06-27 19:38

I'm getting "pkg_resources.DistributionNotFound: The 'ecos>=2' distribution was not found and is required by cvxpy" despite I've installed both ecos and python-ecos

jierumei commented on 2018-08-20 19:43

Having error when checking

======================================================================
FAIL: test_verbose (cvxpy.tests.test_problem.TestProblem)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/jrmei/.cache/yay/python-cvxpy/src/cvxpy-1.0.3/cvxpy/tests/test_problem.py", line 310, in test_verbose
    assert len(output) > 0
AssertionError: 
-------------------- >> begin captured stdout << ---------------------
ECOS

--------------------- >> end captured stdout << ----------------------

----------------------------------------------------------------------
Ran 1823 tests in 46.955s

FAILED (failures=1)
==> ERROR: A failure occurred in check().
    Aborting...
Error making: python-cvxpy

JohnRobson commented on 2018-06-05 23:59

To be able to install this package you must to install this 'blas' package: "openblas-lapack". (Don't install: blas, openblas lapack cblas lapacke, atlas-lapack).

Check your installation:

from numpy.distutils.system_info import get_info

get_info('blas_opt')

{'libraries': ['openblas', 'openblas'], 'library_dirs': ['/usr/lib64'], 'language': 'c', 'define_macros': [('HAVE_CBLAS', None)]}

thrasibule commented on 2016-12-06 21:59

Package should depend on fastcache https://pypi.python.org/pypi/fastcache/0.4.0, not in AUR yet.