darwin aarch64 skrf/calibration/tests/test_calibration.py failing with divide by zero (determinate calculations)

[ghost]

Describe the bug
updating to scikit-rf version 0.29.1 PR NixOS/nixpkgs#266389 fails in the unit test skrf/calibration/tests/test_calibration.py with divide by zero errors.

numpy version is: 1.25.2, python is 3.11, darwin (macos) aarch64

unitfail.txt

1st failure pasted inline.

FAILED skrf/calibration/tests/test_calibration.py::DetermineTest::test_determine_reflect_matched_thru_and_line - RuntimeWarning: divide by zero encountered in det
FAILED skrf/calibration/tests/test_calibration.py::DetermineTest::test_determine_reflect_matched_thru_and_line_ideal_reflect - RuntimeWarning: divide by zero encountered in det
FAILED skrf/calibration/tests/test_calibration.py::DetermineTest::test_determine_reflect_regression - RuntimeWarning: divide by zero encountered in det
=================================== FAILURES ===================================
__________ DetermineTest.test_determine_reflect_matched_thru_and_line __________
self = <test_calibration.DetermineTest testMethod=test_determine_reflect_matched_thru_and_line>
    def test_determine_reflect_matched_thru_and_line(self):
        freq = rf.F(.25, .7, 40, unit = 'GHz')
        medium = Coaxial.from_attenuation_VF(freq, att = 3.0, VF = .69)
    
        thru = medium.line(0, 'm')
        line = medium.line(0.12, 'm')
        short = medium.short()
    
        rng = npy.random.default_rng(12)
        short.s[:, 0, 0] = short.s[:, 0, 0]+rng.uniform(-.02, 0.02, freq.f.size) + rng.uniform(-.02, 0.02, freq.f.size)*1j
    
>       r = determine_reflect(thru, rf.two_port_reflect(short, short), line)
skrf/calibration/tests/test_calibration.py:101: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
skrf/calibration/calibration.py:6406: in determine_reflect
    d = -det(thru_m.s)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 
a = array([[[5.e-08+0.j, 1.e+00+0.j],
        [1.e+00+0.j, 5.e-08+0.j]],
       [[5.e-08+0.j, 1.e+00+0.j],
        [1.e+0... 1.e+00+0.j],
        [1.e+00+0.j, 5.e-08+0.j]],
       [[5.e-08+0.j, 1.e+00+0.j],
        [1.e+00+0.j, 5.e-08+0.j]]])
    @array_function_dispatch(_unary_dispatcher)
    def det(a):
        """
        Compute the determinant of an array.
    
        Parameters
        ----------
        a : (..., M, M) array_like
            Input array to compute determinants for.
    
        Returns
        -------
        det : (...) array_like
            Determinant of `a`.
    
        See Also
        --------
        slogdet : Another way to represent the determinant, more suitable
          for large matrices where underflow/overflow may occur.
        scipy.linalg.det : Similar function in SciPy.
    
        Notes
        -----
    
        .. versionadded:: 1.8.0
    
        Broadcasting rules apply, see the `numpy.linalg` documentation for
        details.
    
        The determinant is computed via LU factorization using the LAPACK
        routine ``z/dgetrf``.
    
        Examples
        --------
        The determinant of a 2-D array [[a, b], [c, d]] is ad - bc:
    
        >>> a = np.array([[1, 2], [3, 4]])
        >>> np.linalg.det(a)
        -2.0 # may vary
    
        Computing determinants for a stack of matrices:
    
        >>> a = np.array([ [[1, 2], [3, 4]], [[1, 2], [2, 1]], [[1, 3], [3, 1]] ])
        >>> a.shape
        (3, 2, 2)
        >>> np.linalg.det(a)
        array([-2., -3., -8.])
    
        """
        a = asarray(a)
        _assert_stacked_2d(a)
        _assert_stacked_square(a)
        t, result_t = _commonType(a)
        signature = 'D->D' if isComplexType(t) else 'd->d'
>       r = _umath_linalg.det(a, signature=signature)
E       RuntimeWarning: divide by zero encountered in det

/nix/store/afrric2fpgzq45bvpn5rabs26v02nn3w-python3.11-numpy-1.25.2/lib/python3.11/site-packages/numpy/linalg/linalg.py:2180: RuntimeWarning

To Reproduce
Steps to reproduce the behavior:

Expected behavior
A clear and concise description of what you expected to happen.

Data
If applicable, add the data leading to the issue (a minimum example is appreciated)

Screenshots
If applicable, add screenshots to help explain your problem.

System

• OS: macos aarch64
• scikit-rf version 0.29.1

Additional context
Add any other context about the problem here if relevant.
numpy version is: 1.25.2
python is 3.11

[ghost]

on macos aarch64 this simplified code produces a divide by zero warning but runs fine on x64 linux -- similar to the failing unit tests above.

#!/usr/bin/env python3
import numpy as np
print(np.__version__)
a = np.array([[5.e-8+0.j, 1+0.j], [1+0.j, 5.e-8+0.j]])
print(a)
print(np.linalg.det(a))

./test.py
1.25.2
[[5.e-08+0.j 1.e+00+0.j]
 [1.e+00+0.j 5.e-08+0.j]]
/nix/store/rpdx49knxarbr4lmkwfi5l0461lyp6bd-python3-3.11.6-env/lib/python3.11/site-packages/numpy/linalg/linalg.py:2180: RuntimeWarning: divide by zero encountered in det
  r = _umath_linalg.det(a, signature=signature)
/nix/store/rpdx49knxarbr4lmkwfi5l0461lyp6bd-python3-3.11.6-env/lib/python3.11/site-packages/numpy/linalg/linalg.py:2180: RuntimeWarning: invalid value encountered in det
  r = _umath_linalg.det(a, signature=signature)

[pbsds]

note to future self: re-enable the failing test in nixpkgs if a fix is found

[FranzForstmayr]

This should be reported to https://github.com/numpy/numpy itself, but this probably related numpy/numpy#22025.

[FranzForstmayr]

Nevertheless we could modify the unit test slightly to avoid this issue. I'm afraid we are not able to find the root cause here.