[2.5.0] Center of charge test failures on macOS runners #4211

IAlibay · 2023-07-26T09:50:42Z

Expected behavior

All tests on our latest production release should pass across expected machine types (especially ubuntu, and macos).

Actual behavior

As reported by @orbeckst here (and reproduced in #4210), tests for CoC unwrapping are failing.

This isn't seen on develop.

Code to reproduce the behavior

Run tests on the 2.5.0 on an intel macOS machine

orbeckst · 2023-07-26T17:32:23Z

Do you even have an idea what could cause the issue?

Is it specific to Intel Mac??

IAlibay · 2023-07-26T17:41:10Z

Yeah Intel Mac only. It's odd because the tests haven't changed, which makes me think that the base numpy version we use to cythonize might be to blame here. I want to check that once the new CI checks are in place.

IAlibay · 2023-07-29T12:05:47Z

Actually I don't know if it's specific to x86_64 intel - can anyone with a macos-arm64 confirm that this happens if they pull from conda-forge?

Note: this is only happening on conda-forge. I'm currently thinking it might be down to clang 15.

IAlibay · 2023-07-29T12:06:56Z

I will add here that the osx-arm64 builds on the feedstock aren't facing the issue, so maybe it is just x86_64: https://dev.azure.com/conda-forge/feedstock-builds/_build/results?buildId=753485&view=logs&j=9c5ef928-2cd6-52e5-dbe6-9d173a7d951b&t=c66c4975-bbef-551b-2ee6-85005a231b12

IAlibay · 2023-07-29T12:35:57Z

tagging @RMeli who has both an intel and arm macos machine

IAlibay · 2023-07-29T12:53:09Z

Where I'm at on all this:

Conda-forge builds have the above failure, but not PyPi ones
Building from PyPi source is also fine
All other platforms but x86_64 macos seem fine
The relevant code in 2.6.0 does not seem to have changed.
The previous conda-forge builds of 2.5.0 also fail (so it's unlikely to be something related to building with Cython 3)
Building with Cython 3 using PyPi does not seem to cause any issues
Using clang 14 (as is used by default on gh actions) instead of clang 15 (conda-forge default) does not change the issue

IAlibay · 2023-07-29T12:54:35Z

Short of someone giving me access to a macos x86_64 machine, I'm rapidly running out of options (and weekend patience..) on this one.

RMeli · 2023-07-29T19:10:09Z

I can confirm that I don't see this issue on osx-arm64:

mamba create -n mda-2.5.0-test MDAnalysis=2.5.0 MDAnalysisTests pytest pytest-xdist
mamba activate mda-2.5.0-test
pytest --disable-pytest-warnings --pyargs MDAnalysisTests --numprocesses 4

================================ 18559 passed, 403 skipped, 8 xfailed, 2 xpassed, 9533 warnings in 230.61s (0:03:50) =================================

I'll try x86_64 next and see what I can dig up.

hmacdope · 2023-07-30T10:53:24Z

Where I'm at on all this:

Conda-forge builds have the above failure, but not PyPi ones

Building from PyPi source is also fine

All other platforms but x86_64 macos seem fine

The relevant code in 2.6.0 does not seem to have changed.

The previous conda-forge builds of 2.5.0 also fail (so it's unlikely to be something related to building with Cython 3)

Building with Cython 3 using PyPi does not seem to cause any issues

Using clang 14 (as is used by default on gh actions) instead of clang 15 (conda-forge default) does not change the issue

This smells of compiler flags to me, I'll try and see if any extra optimisations on CF.

IAlibay · 2023-07-30T11:08:05Z

This smells of compiler flags to me, I'll try and see if any extra optimisations on CF.

Here are the flags apparently set by CF:

2023-07-29T12:06:29.9787810Z +CXXFLAGS=-march=core2 -mtune=haswell -mssse3 -ftree-vectorize -fPIC -fPIE -fstack-protector-strong -O2 -pipe -stdlib=libc++ -fvisibility-inlines-hidden -fmessage-length=0 -isystem $PREFIX/include -fdebug-prefix-map=$SRC_DIR=/usr/local/src/conda/mdanalysis-2.5.0 -fdebug-prefix-map=$PREFIX=/usr/local/src/conda-prefix

hmacdope · 2023-07-30T23:12:15Z

Strange that they use -mtune=haswell rather than generic but shouldn't effect the accuracy. Could we try drop -msse3 and see what happens? Could be an extra SIMD optimisation that is changing something.

IAlibay · 2023-07-30T23:22:21Z

:/ I think we're going to have to create a custom build.sh on conda-forge, that's a lot of work. Maybe @RMeli or anyone else with an Intel MacOS machine can first check locally if building mdanalysis with or without -msse3 does the trick?

RMeli · 2023-07-31T07:26:51Z

I'm AFK, but I can certainly try ASAP.

orbeckst · 2023-08-01T00:36:05Z

If someone gives me instructions that even an idiot (I mean, PI) could follow then I'll be happy to do it. Sorry, a bit short on time to figure it out.

hmacdope · 2023-08-01T00:46:54Z

Should be enough to uncomment extra_cflags here and add -msse3 ,
Should look like

extra_cflags = -msse3

then recompile on your Intel Mac and see if that breaks the tests

orbeckst · 2023-08-01T01:57:01Z

On x86_64 intel macOS 13.5 with Python 3.10 and the -msse3 flag I still get the same failures.

diff --git a/package/setup.cfg b/package/setup.cfg
index b4151d55f..18ca2f399 100644
--- a/package/setup.cfg
+++ b/package/setup.cfg
@@ -9,7 +9,7 @@
 ## Uncomment the below line to generate the annotated cython HTML output
 # annotate_cython=True
 ## any additional compiler flags for core C routines, excluding ENCORE
-#extra_cflags =
+extra_cflags = -msse3

 [build_sphinx]
 all-files = 1

$ clang --version
Apple clang version 14.0.3 (clang-1403.0.22.14.1)
Target: x86_64-apple-darwin22.6.0
Thread model: posix
InstalledDir: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin

set up and testing

mamba create -c conda-forge -n mda310y python==3.10 pytest pytest-xdist
mamba activate mda310y
pip install -v -e package/
pip install -v -e testsuite/

Inspecting the output shows that

Cython 3.0 is used (and complains about deprecation of DEF)
-msse3 shows up in clang lines

run tests

pytest --disable-pytest-warnings --pyargs MDAnalysisTests -k Testcenter_of_charge

results

(mda310y) deathstar:mdanalysis oliver$ pytest --disable-pytest-warnings --pyargs MDAnalysisTests -k Testcenter_of_charge
======================================================= test session starts ========================================================
platform darwin -- Python 3.10.0, pytest-7.4.0, pluggy-1.2.0
rootdir: ~/MDAnalysis/mdanalysis
plugins: hypothesis-6.82.0, xdist-3.3.1
collected 19037 items / 19022 deselected / 7 skipped / 15 selected

testsuite/MDAnalysisTests/core/test_topologyattrs.py ..........FFFFF                                                         [100%]

============================================================= FAILURES =============================================================
___________________________________________ Testcenter_of_charge.test_coc_unwrap[group] ____________________________________________

self = <MDAnalysisTests.core.test_topologyattrs.Testcenter_of_charge object at 0x130740430>, u = <Universe with 4 atoms>
compound = 'group'

    @pytest.mark.parametrize('compound', compounds)
    def test_coc_unwrap(self, u, compound):
        u.atoms.wrap
        coc = u.atoms[:2].center_of_charge(compound=compound, unwrap=True)
>       assert_equal(coc.flatten(), [0, -0.5, 0])

~/MDAnalysis/mdanalysis/testsuite/MDAnalysisTests/core/test_topologyattrs.py:693:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

args = (<built-in function eq>, array([0.        , 0.49999997, 0.        ]), [0, -0.5, 0])
kwds = {'err_msg': '', 'header': 'Arrays are not equal', 'strict': False, 'verbose': True}

    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
E           AssertionError:
E           Arrays are not equal
E
E           Mismatched elements: 1 / 3 (33.3%)
E           Max absolute difference: 0.99999997
E           Max relative difference: 1.99999994
E            x: array([0. , 0.5, 0. ])
E            y: array([ 0. , -0.5,  0. ])

~/tmp/micromamba-root/envs/mda310y/lib/python3.10/contextlib.py:79: AssertionError
__________________________________________ Testcenter_of_charge.test_coc_unwrap[segments] __________________________________________

self = <MDAnalysisTests.core.test_topologyattrs.Testcenter_of_charge object at 0x1307405b0>, u = <Universe with 4 atoms>
compound = 'segments'

    @pytest.mark.parametrize('compound', compounds)
    def test_coc_unwrap(self, u, compound):
        u.atoms.wrap
        coc = u.atoms[:2].center_of_charge(compound=compound, unwrap=True)
>       assert_equal(coc.flatten(), [0, -0.5, 0])

~/MDAnalysis/mdanalysis/testsuite/MDAnalysisTests/core/test_topologyattrs.py:693:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

args = (<built-in function eq>, array([0.        , 0.49999997, 0.        ]), [0, -0.5, 0])
kwds = {'err_msg': '', 'header': 'Arrays are not equal', 'strict': False, 'verbose': True}

    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
E           AssertionError:
E           Arrays are not equal
E
E           Mismatched elements: 1 / 3 (33.3%)
E           Max absolute difference: 0.99999997
E           Max relative difference: 1.99999994
E            x: array([0. , 0.5, 0. ])
E            y: array([ 0. , -0.5,  0. ])

~/tmp/micromamba-root/envs/mda310y/lib/python3.10/contextlib.py:79: AssertionError
__________________________________________ Testcenter_of_charge.test_coc_unwrap[residues] __________________________________________

self = <MDAnalysisTests.core.test_topologyattrs.Testcenter_of_charge object at 0x130740640>, u = <Universe with 4 atoms>
compound = 'residues'

    @pytest.mark.parametrize('compound', compounds)
    def test_coc_unwrap(self, u, compound):
        u.atoms.wrap
        coc = u.atoms[:2].center_of_charge(compound=compound, unwrap=True)
>       assert_equal(coc.flatten(), [0, -0.5, 0])

~/MDAnalysis/mdanalysis/testsuite/MDAnalysisTests/core/test_topologyattrs.py:693:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

args = (<built-in function eq>, array([0.        , 0.49999997, 0.        ]), [0, -0.5, 0])
kwds = {'err_msg': '', 'header': 'Arrays are not equal', 'strict': False, 'verbose': True}

    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
E           AssertionError:
E           Arrays are not equal
E
E           Mismatched elements: 1 / 3 (33.3%)
E           Max absolute difference: 0.99999997
E           Max relative difference: 1.99999994
E            x: array([0. , 0.5, 0. ])
E            y: array([ 0. , -0.5,  0. ])

~/tmp/micromamba-root/envs/mda310y/lib/python3.10/contextlib.py:79: AssertionError
_________________________________________ Testcenter_of_charge.test_coc_unwrap[molecules] __________________________________________

self = <MDAnalysisTests.core.test_topologyattrs.Testcenter_of_charge object at 0x1307406d0>, u = <Universe with 4 atoms>
compound = 'molecules'

    @pytest.mark.parametrize('compound', compounds)
    def test_coc_unwrap(self, u, compound):
        u.atoms.wrap
        coc = u.atoms[:2].center_of_charge(compound=compound, unwrap=True)
>       assert_equal(coc.flatten(), [0, -0.5, 0])

~/MDAnalysis/mdanalysis/testsuite/MDAnalysisTests/core/test_topologyattrs.py:693:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

args = (<built-in function eq>, array([0.        , 0.49999997, 0.        ]), [0, -0.5, 0])
kwds = {'err_msg': '', 'header': 'Arrays are not equal', 'strict': False, 'verbose': True}

    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
E           AssertionError:
E           Arrays are not equal
E
E           Mismatched elements: 1 / 3 (33.3%)
E           Max absolute difference: 0.99999997
E           Max relative difference: 1.99999994
E            x: array([0. , 0.5, 0. ])
E            y: array([ 0. , -0.5,  0. ])

~/tmp/micromamba-root/envs/mda310y/lib/python3.10/contextlib.py:79: AssertionError
_________________________________________ Testcenter_of_charge.test_coc_unwrap[fragments] __________________________________________

self = <MDAnalysisTests.core.test_topologyattrs.Testcenter_of_charge object at 0x130740760>, u = <Universe with 4 atoms>
compound = 'fragments'

    @pytest.mark.parametrize('compound', compounds)
    def test_coc_unwrap(self, u, compound):
        u.atoms.wrap
        coc = u.atoms[:2].center_of_charge(compound=compound, unwrap=True)
>       assert_equal(coc.flatten(), [0, -0.5, 0])

~/MDAnalysis/mdanalysis/testsuite/MDAnalysisTests/core/test_topologyattrs.py:693:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

args = (<built-in function eq>, array([0.        , 0.49999997, 0.        ]), [0, -0.5, 0])
kwds = {'err_msg': '', 'header': 'Arrays are not equal', 'strict': False, 'verbose': True}

    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
E           AssertionError:
E           Arrays are not equal
E
E           Mismatched elements: 1 / 3 (33.3%)
E           Max absolute difference: 0.99999997
E           Max relative difference: 1.99999994
E            x: array([0. , 0.5, 0. ])
E            y: array([ 0. , -0.5,  0. ])

~/tmp/micromamba-root/envs/mda310y/lib/python3.10/contextlib.py:79: AssertionError
===================================================== short test summary info ======================================================
FAILED testsuite/MDAnalysisTests/core/test_topologyattrs.py::Testcenter_of_charge::test_coc_unwrap[group] - AssertionError:
FAILED testsuite/MDAnalysisTests/core/test_topologyattrs.py::Testcenter_of_charge::test_coc_unwrap[segments] - AssertionError:
FAILED testsuite/MDAnalysisTests/core/test_topologyattrs.py::Testcenter_of_charge::test_coc_unwrap[residues] - AssertionError:
FAILED testsuite/MDAnalysisTests/core/test_topologyattrs.py::Testcenter_of_charge::test_coc_unwrap[molecules] - AssertionError:
FAILED testsuite/MDAnalysisTests/core/test_topologyattrs.py::Testcenter_of_charge::test_coc_unwrap[fragments] - AssertionError:
============================== 5 failed, 10 passed, 7 skipped, 19022 deselected, 6 warnings in 27.97s ==============================

hmacdope · 2023-08-01T01:58:37Z

Sorry @orbeckst my theory was that the flag causes the issue. Do you get the issue without the flag ?

orbeckst · 2023-08-01T02:02:29Z

Ok, I'll try a build without first.

orbeckst · 2023-08-01T02:04:53Z

(FYI:

warning: MDAnalysis/lib/nsgrid.pyx:86:0: The 'DEF' statement is deprecated and will be removed in a future Cython version. Consider using global variables, constants, and in-place literals instead. See cython/cython#4310

from Cython...)

orbeckst · 2023-08-01T02:09:16Z

I built without -msse3 and the tests still fail in the same way.

orbeckst · 2023-08-01T02:15:34Z

I removed -ffast-math and suddenly the test passes (without -msse3).

We wanted to remove it anyway #3821 — can I submit a PR?

diff --git a/package/setup.py b/package/setup.py
index 898ab5b3b..f6467bc4b 100755
--- a/package/setup.py
+++ b/package/setup.py
@@ -268,7 +268,7 @@ def extensions(config):
     use_openmp = config.get('use_openmp', default=True)
     annotate_cython = config.get('annotate_cython', default=False)

-    extra_compile_args = ['-std=c99', '-ffast-math', '-O3', '-funroll-loops',
+    extra_compile_args = ['-std=c99', '-O3', '-funroll-loops',
                           '-fsigned-zeros'] # see #2722
     define_macros = []
     if config.get('debug_cflags', default=False):

result

(mda310y) deathstar:mdanalysis oliver$ pytest --disable-pytest-warnings --pyargs MDAnalysisTests -k Testcenter_of_charge
======================================================= test session starts ========================================================
platform darwin -- Python 3.10.0, pytest-7.4.0, pluggy-1.2.0
rootdir: ~/MDAnalysis/mdanalysis
plugins: hypothesis-6.82.0, xdist-3.3.1
collected 19037 items / 19022 deselected / 7 skipped / 15 selected

testsuite/MDAnalysisTests/core/test_topologyattrs.py ...............                                                         [100%]

=================================== 15 passed, 7 skipped, 19022 deselected, 6 warnings in 6.68s ====================================

orbeckst · 2023-08-01T02:19:42Z

@hmacdope I also tested with -msse3 and tests still pass, as long as we don't use fast-math.

diff --git a/package/setup.cfg b/package/setup.cfg
index b4151d55f..18ca2f399 100644
--- a/package/setup.cfg
+++ b/package/setup.cfg
@@ -9,7 +9,7 @@
 ## Uncomment the below line to generate the annotated cython HTML output
 # annotate_cython=True
 ## any additional compiler flags for core C routines, excluding ENCORE
-#extra_cflags =
+extra_cflags = -msse3

 [build_sphinx]
 all-files = 1
diff --git a/package/setup.py b/package/setup.py
index 898ab5b3b..f6467bc4b 100755
--- a/package/setup.py
+++ b/package/setup.py
@@ -268,7 +268,7 @@ def extensions(config):
     use_openmp = config.get('use_openmp', default=True)
     annotate_cython = config.get('annotate_cython', default=False)

-    extra_compile_args = ['-std=c99', '-ffast-math', '-O3', '-funroll-loops',
+    extra_compile_args = ['-std=c99', '-O3', '-funroll-loops',
                           '-fsigned-zeros'] # see #2722
     define_macros = []
     if config.get('debug_cflags', default=False):

- fix #4211 - fix #3821 - Without -ffast-math, the AtomGroup.center_of_charge(..., unwrap=True) pass again on Intel x86_64 macOS. - updated CHANGELOG

orbeckst · 2023-08-01T02:36:33Z

@hmacdope is the output of the tests on intel mac actually wrong or just showing up on the other side of the box, i.e., still correct under PBC?

hmacdope · 2023-08-01T07:49:44Z

Playing with it on paper it seems that way. I will spare you my shoddily drawn diagrams. I swear we agreed to remove --fast-math a while ago. Thanks for investigating.

RMeli · 2023-08-01T20:27:26Z

Thanks @orbeckst for stepping in, I was on holidays the past couple of days.

Great investigation, thanks for all the details! I can reproduce your investigation, so #4220 LGTM (since we wanted to remove the flag anyways).

IAlibay · 2023-08-02T00:55:42Z

Probably something to add here, I think the reason that our PyPi cron checks weren't failing but the conda-forge ones are is that the former is defaulting to gcc whilst the latter is using clang.

We probably need to rethink what our build matrix's default compilers are at some point (I started doing that at some point and it became super messy).

hmacdope · 2023-08-02T00:58:19Z

I would advocate for clang if possible.

IAlibay · 2023-08-02T01:00:19Z

🤦🏽 I think I know what I did wrong. This is a me forgetting what I did in the past issue.

orbeckst · 2023-08-02T14:11:54Z

@hmacdope just to clarify your comment #4211 (comment) : I read it as that you're confirming that the failing tests here are still showing scientifically correct results but that they are producing alternative results (i.e., it's more of a reproducibility issue). Is that a correct interpretation?

(I am asking because that seems to determine how we will handle PR #4220 .)

hmacdope · 2023-08-02T20:30:47Z

@hmacdope just to clarify your comment #4211 (comment) : I read it as that you're confirming that the failing tests here are still showing scientifically correct results but that they are producing alternative results (i.e., it's more of a reproducibility issue). Is that a correct interpretation?

(I am asking because that seems to determine how we will handle PR #4220 .)

Yep that is correct. It's basically a sign bit issue due to some denormal to 0 squashing. Results are scientifically correct.

* Fix AtomGroup.center_of_charge(..., unwrap=True) giving inconsistent (but scientifically correct) results on Intel macOS (Issue #4211) * Removed `-ffast-math` compiler flag to avoid potentially incorrect floating point results in MDAnalysis *and* other codes when MDAnalysis shared libraries are loaded --- see https://moyix.blogspot.com/2022/09/someones-been-messing-with-my-subnormals.html (Issues #3821, #4211)

IAlibay added Priority-Critical testing labels Jul 26, 2023

IAlibay mentioned this issue Jul 26, 2023

[CI] Add post deployment conda and pypi cron tests #4210

Merged

4 tasks

orbeckst added a commit that referenced this issue Aug 1, 2023

remove -ffast-math compiler flag

8581a9a

- fix #4211 - fix #3821 - Without -ffast-math, the AtomGroup.center_of_charge(..., unwrap=True) pass again on Intel x86_64 macOS. - updated CHANGELOG

orbeckst mentioned this issue Aug 1, 2023

remove -ffast-math compiler flag #4220

Merged

4 tasks

IAlibay mentioned this issue Aug 2, 2023

Add setup-os contents to deployed MDA checks in CRON CI #4223

Closed

RMeli closed this as completed in #4220 Aug 2, 2023

orbeckst mentioned this issue Aug 22, 2023

Remove fast-math MDAnalysis/mdaencore#6

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[2.5.0] Center of charge test failures on macOS runners #4211

[2.5.0] Center of charge test failures on macOS runners #4211

IAlibay commented Jul 26, 2023

orbeckst commented Jul 26, 2023

IAlibay commented Jul 26, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

RMeli commented Jul 29, 2023

hmacdope commented Jul 30, 2023

IAlibay commented Jul 30, 2023

hmacdope commented Jul 30, 2023

IAlibay commented Jul 30, 2023 •

edited

Loading

RMeli commented Jul 31, 2023

orbeckst commented Aug 1, 2023

hmacdope commented Aug 1, 2023

orbeckst commented Aug 1, 2023 •

edited

Loading

hmacdope commented Aug 1, 2023

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023 •

edited

Loading

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023

hmacdope commented Aug 1, 2023

RMeli commented Aug 1, 2023

IAlibay commented Aug 2, 2023

hmacdope commented Aug 2, 2023

IAlibay commented Aug 2, 2023

orbeckst commented Aug 2, 2023

hmacdope commented Aug 2, 2023

[2.5.0] Center of charge test failures on macOS runners #4211

[2.5.0] Center of charge test failures on macOS runners #4211

Comments

IAlibay commented Jul 26, 2023

Expected behavior

Actual behavior

Code to reproduce the behavior

orbeckst commented Jul 26, 2023

IAlibay commented Jul 26, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

IAlibay commented Jul 29, 2023

RMeli commented Jul 29, 2023

hmacdope commented Jul 30, 2023

IAlibay commented Jul 30, 2023

hmacdope commented Jul 30, 2023

IAlibay commented Jul 30, 2023 • edited Loading

RMeli commented Jul 31, 2023

orbeckst commented Aug 1, 2023

hmacdope commented Aug 1, 2023

orbeckst commented Aug 1, 2023 • edited Loading

set up and testing

results

hmacdope commented Aug 1, 2023

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023 • edited Loading

result

orbeckst commented Aug 1, 2023

orbeckst commented Aug 1, 2023

hmacdope commented Aug 1, 2023

RMeli commented Aug 1, 2023

IAlibay commented Aug 2, 2023

hmacdope commented Aug 2, 2023

IAlibay commented Aug 2, 2023

orbeckst commented Aug 2, 2023

hmacdope commented Aug 2, 2023

IAlibay commented Jul 30, 2023 •

edited

Loading

orbeckst commented Aug 1, 2023 •

edited

Loading

orbeckst commented Aug 1, 2023 •

edited

Loading