Bug 2137443

Description of problem:
In https://src.fedoraproject.org/rpms/python-hypothesis/pull-request/17 I'd like to update hypothesis to 6.56.3. However, python-astropy-healpix fails to build with it. I've also tried updating hypothesis to 6.49.1, but python-astropy-healpix failed there as well.

https://copr.fedorainfracloud.org/coprs/churchyard/hypothesis-6.56.3/package/python-astropy-healpix/
https://copr.fedorainfracloud.org/coprs/churchyard/hypothesis-6.49.1/package/python-astropy-healpix/

_________________________________ test_vec2pix _________________________________

    @given(nside_pow=integers(0, 29), nest=booleans(),
>          x=floats(-1, 1, allow_nan=False, allow_infinity=False).filter(lambda x: abs(x) > 1e-10),
           y=floats(-1, 1, allow_nan=False, allow_infinity=False).filter(lambda y: abs(y) > 1e-10),
           z=floats(-1, 1, allow_nan=False, allow_infinity=False).filter(lambda z: abs(z) > 1e-10))

astropy_healpix/tests/test_healpy.py:116: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

nside_pow = 2, x = 1.0, y = 1.0, z = 0.5, nest = False

    @given(nside_pow=integers(0, 29), nest=booleans(),
           x=floats(-1, 1, allow_nan=False, allow_infinity=False).filter(lambda x: abs(x) > 1e-10),
           y=floats(-1, 1, allow_nan=False, allow_infinity=False).filter(lambda y: abs(y) > 1e-10),
           z=floats(-1, 1, allow_nan=False, allow_infinity=False).filter(lambda z: abs(z) > 1e-10))
    @settings(max_examples=2000, derandomize=True)
    def test_vec2pix(nside_pow, x, y, z, nest):
        nside = 2 ** nside_pow
        ipix1 = hp_compat.vec2pix(nside, x, y, z, nest=nest)
        ipix2 = hp.vec2pix(nside, x, y, z, nest=nest)
>       assert ipix1 == ipix2
E       assert 42 == 58
E       Falsifying example: test_vec2pix(
E           nside_pow=2, nest=False, x=1.0, y=1.0, z=0.5,
E       )

astropy_healpix/tests/test_healpy.py:124: AssertionError
_________________________________ test_vec2ang _________________________________

    @given(vectors=arrays(float, (3,), elements=floats(-1, 1)).filter(not_at_origin),
>          lonlat=booleans(), ndim=integers(0, 4))

astropy_healpix/tests/test_healpy.py:210: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

vectors = array([  0.00000000e+000,   1.11137937e-162,   1.00000000e+000])
lonlat = False, ndim = 0

    @given(vectors=arrays(float, (3,), elements=floats(-1, 1)).filter(not_at_origin),
           lonlat=booleans(), ndim=integers(0, 4))
    @settings(max_examples=500, derandomize=True)
    def test_vec2ang(vectors, lonlat, ndim):
        vectors = np.broadcast_to(vectors, (2,) * ndim + (3,))
        theta1, phi1 = hp_compat.vec2ang(vectors, lonlat=lonlat)
        theta2, phi2 = hp.vec2ang(vectors, lonlat=lonlat)
        # Healpy sometimes returns NaNs for phi (somewhat incorrectly)
        phi2 = np.nan_to_num(phi2)
        assert_allclose(theta1, theta1, atol=1e-10)
>       assert_allclose(phi1, phi2, atol=1e-10)
E       AssertionError: 
E       Not equal to tolerance rtol=1e-07, atol=1e-10
E       
E       Mismatched elements: 1 / 1 (100%)
E       Max absolute difference: 1.5707963267948966
E       Max relative difference: 1.0
E        x: array([ 0.])
E        y: array([ 1.570796])
E       Falsifying example: test_vec2ang(
E           vectors=array([  0.00000000e+000,   1.11137937e-162,   1.00000000e+000]),
E           lonlat=False,
E           ndim=0,
E       )

astropy_healpix/tests/test_healpy.py:219: AssertionError



Version-Release number of selected component: 0.7-1


How reproducible: consistently


Steps to Reproduce:
1. fetch hypothesis update from the PR or one of the coprs linked above
2. build python-astropy-healpix from rawhide dist-git

Actual results: build fails


Expected results: build succeeds


Additional info:
Please let me know if I can ship the update, or if you'd like to deep dive into the failure before I do that.