diff options
| -rw-r--r-- | numpy/core/tests/test_numeric.py | 32 |
1 files changed, 14 insertions, 18 deletions
diff --git a/numpy/core/tests/test_numeric.py b/numpy/core/tests/test_numeric.py index fcedf33fc..1315364bc 100644 --- a/numpy/core/tests/test_numeric.py +++ b/numpy/core/tests/test_numeric.py @@ -2357,34 +2357,30 @@ class TestClip: That accounts for most of the function; the actual test is just three lines to calculate and compare actual vs expected results! """ + numeric_dtypes = hynp.integer_dtypes() | hynp.floating_dtypes() # Generate shapes for the bounds which can be broadcast with each other # and with the base shape. Below, we might decide to use scalar bounds, # but it's clearer to generate these shapes unconditionally in advance. in_shapes, result_shape = data.draw( hynp.mutually_broadcastable_shapes( - num_shapes=2, - base_shape=arr.shape, - # Commenting out the min_dims line allows zero-dimensional arrays, - # and zero-dimensional arrays containing NaN make the test fail. - min_dims=1 + num_shapes=2, base_shape=arr.shape ) ) - # This test can fail if we allow either bound to be a scalar `nan`, or - # if bounds are of a different (still integer or float) dtype than the - # array. At some point we should investigate and fix those problems. - amin = data.draw( - hynp.from_dtype(arr.dtype, allow_nan=False) - | hynp.arrays(dtype=arr.dtype, shape=in_shapes[0]) - ) - amax = data.draw( - hynp.from_dtype(arr.dtype, allow_nan=False) - | hynp.arrays(dtype=arr.dtype, shape=in_shapes[1]) - ) + # Scalar `nan` is deprecated due to the differing behaviour it shows. + s = numeric_dtypes.flatmap( + lambda x: hynp.from_dtype(x, allow_nan=False)) + amin = data.draw(s | hynp.arrays(dtype=numeric_dtypes, + shape=in_shapes[0], elements={"allow_nan": False})) + amax = data.draw(s | hynp.arrays(dtype=numeric_dtypes, + shape=in_shapes[1], elements={"allow_nan": False})) # Then calculate our result and expected result and check that they're - # equal! See gh-12519 for discussion deciding on this property. + # equal! See gh-12519 and gh-19457 for discussion deciding on this + # property and the result_type argument. result = np.clip(arr, amin, amax) - expected = np.minimum(amax, np.maximum(arr, amin)) + t = np.result_type(arr, amin, amax) + expected = np.minimum(amax, np.maximum(arr, amin, dtype=t), dtype=t) + assert result.dtype == t assert_array_equal(result, expected) |