MAINT: Check essential data functions

Author: mtsokol

array_api_tests/test_creation_functions.py

@@ -3,6 +3,7 @@
 from itertools import count
 from typing import Iterator, NamedTuple, Union
 
+import pytest
 from hypothesis import assume, given, note
 from hypothesis import strategies as st
 
@@ -76,6 +77,7 @@ def reals(min_value=None, max_value=None) -> st.SearchStrategy[Union[int, float]
     )
 
 
+@pytest.mark.has_setup_funcs
 @given(dtype=st.none() | hh.real_dtypes, data=st.data())
 def test_arange(dtype, data):
     if dtype is None or dh.is_float_dtype(dtype):
@@ -194,6 +196,7 @@ def test_arange(dtype, data):
             ), f"out[0]={out[0]}, but should be {_start} {f_func}"
 
 
+@pytest.mark.has_setup_funcs
 @given(shape=hh.shapes(min_side=1), data=st.data())
 def test_asarray_scalars(shape, data):
     kw = data.draw(
@@ -257,6 +260,7 @@ def scalar_eq(s1: Scalar, s2: Scalar) -> bool:
         return s1 == s2
 
 
+@pytest.mark.has_setup_funcs
 @given(
     shape=hh.shapes(),
     dtypes=hh.oneway_promotable_dtypes(dh.all_dtypes),
@@ -424,6 +428,7 @@ def test_full(shape, fill_value, kw):
     ph.assert_fill("full", fill_value=fill_value, dtype=dtype, out=out, kw=dict(fill_value=fill_value))
 
 
+@pytest.mark.has_setup_funcs
 @given(kw=hh.kwargs(dtype=st.none() | hh.all_dtypes), data=st.data())
 def test_full_like(kw, data):
     dtype = kw.get("dtype", None) or data.draw(hh.all_dtypes, label="dtype")
@@ -442,6 +447,7 @@ def test_full_like(kw, data):
 finite_kw = {"allow_nan": False, "allow_infinity": False}
 
 
+@pytest.mark.has_setup_funcs
 @given(
     num=hh.sizes,
     dtype=st.none() | hh.real_floating_dtypes,
@@ -492,6 +498,7 @@ def test_linspace(num, dtype, endpoint, data):
         ph.assert_array_elements("linspace", out=out, expected=expected)
 
 
+@pytest.mark.has_setup_funcs
 @given(dtype=hh.numeric_dtypes, data=st.data())
 def test_meshgrid(dtype, data):
     # The number and size of generated arrays is arbitrarily limited to prevent

array_api_tests/test_manipulation_functions.py

@@ -347,6 +347,7 @@ def test_repeat(x, kw, data):
 
 reshape_shape = st.shared(hh.shapes(), key="reshape_shape")
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.unvectorized
 @given(
     x=hh.arrays(dtype=hh.all_dtypes, shape=reshape_shape),

array_api_tests/test_operators_and_elementwise_functions.py

@@ -781,6 +781,7 @@ def test_acosh(x):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx,", make_binary_params("add", dh.numeric_dtypes))
 @given(data=st.data())
 def test_add(ctx, data):
@@ -854,6 +855,7 @@ def test_atanh(x):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize(
     "ctx", make_binary_params("bitwise_and", dh.bool_and_all_int_dtypes)
 )
@@ -873,6 +875,7 @@ def test_bitwise_and(ctx, data):
     binary_param_assert_against_refimpl(ctx, left, right, res, "&", refimpl)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize(
     "ctx", make_binary_params("bitwise_left_shift", dh.all_int_dtypes)
 )
@@ -895,6 +898,7 @@ def test_bitwise_left_shift(ctx, data):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize(
     "ctx", make_unary_params("bitwise_invert", dh.bool_and_all_int_dtypes)
 )
@@ -913,6 +917,7 @@ def test_bitwise_invert(ctx, data):
     unary_assert_against_refimpl(ctx.func_name, x, out, refimpl, expr_template="~{}={}")
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize(
     "ctx", make_binary_params("bitwise_or", dh.bool_and_all_int_dtypes)
 )
@@ -932,6 +937,7 @@ def test_bitwise_or(ctx, data):
     binary_param_assert_against_refimpl(ctx, left, right, res, "|", refimpl)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize(
     "ctx", make_binary_params("bitwise_right_shift", dh.all_int_dtypes)
 )
@@ -953,6 +959,7 @@ def test_bitwise_right_shift(ctx, data):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize(
     "ctx", make_binary_params("bitwise_xor", dh.bool_and_all_int_dtypes)
 )
@@ -981,6 +988,7 @@ def test_ceil(x):
 
 
 @pytest.mark.min_version("2023.12")
+@pytest.mark.has_setup_funcs
 @given(x=hh.arrays(dtype=hh.real_dtypes, shape=hh.shapes()), data=st.data())
 def test_clip(x, data):
     # Ensure that if both min and max are arrays that all three of x, min, max
@@ -1145,6 +1153,7 @@ def test_cosh(x):
     unary_assert_against_refimpl("cosh", x, out, refimpl)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("divide", dh.all_float_dtypes))
 @given(data=st.data())
 def test_divide(ctx, data):
@@ -1168,6 +1177,7 @@ def test_divide(ctx, data):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("equal", dh.all_dtypes))
 @given(data=st.data())
 def test_equal(ctx, data):
@@ -1242,6 +1252,7 @@ def refimpl(z):
     unary_assert_against_refimpl("floor", x, out, refimpl, strict_check=True)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("floor_divide", dh.real_dtypes))
 @given(data=st.data())
 def test_floor_divide(ctx, data):
@@ -1261,6 +1272,7 @@ def test_floor_divide(ctx, data):
     binary_param_assert_against_refimpl(ctx, left, right, res, "//", operator.floordiv)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("greater", dh.real_dtypes))
 @given(data=st.data())
 def test_greater(ctx, data):
@@ -1281,6 +1293,7 @@ def test_greater(ctx, data):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("greater_equal", dh.real_dtypes))
 @given(data=st.data())
 def test_greater_equal(ctx, data):
@@ -1352,6 +1365,7 @@ def test_isnan(x):
     unary_assert_against_refimpl("isnan", x, out, refimpl, res_stype=bool)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("less", dh.real_dtypes))
 @given(data=st.data())
 def test_less(ctx, data):
@@ -1372,6 +1386,7 @@ def test_less(ctx, data):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("less_equal", dh.real_dtypes))
 @given(data=st.data())
 def test_less_equal(ctx, data):
@@ -1463,6 +1478,7 @@ def logaddexp_refimpl(l: float, r: float) -> float:
 
 
 @pytest.mark.min_version("2023.12")
+@pytest.mark.has_setup_funcs
 @given(*hh.two_mutual_arrays(dh.real_float_dtypes))
 def test_logaddexp(x1, x2):
     out = xp.logaddexp(x1, x2)
@@ -1476,6 +1492,7 @@ def test_logaddexp(x1, x2):
     )
 
 
+@pytest.mark.has_setup_funcs
 @given(hh.arrays(dtype=xp.bool, shape=hh.shapes()))
 def test_logical_not(x):
     out = xp.logical_not(x)
@@ -1486,6 +1503,7 @@ def test_logical_not(x):
     )
 
 
+@pytest.mark.has_setup_funcs
 @given(*hh.two_mutual_arrays([xp.bool]))
 def test_logical_and(x1, x2):
     out = xp.logical_and(x1, x2)
@@ -1500,6 +1518,7 @@ def test_logical_and(x1, x2):
     )
 
 
+@pytest.mark.has_setup_funcs
 @given(*hh.two_mutual_arrays([xp.bool]))
 def test_logical_or(x1, x2):
     out = xp.logical_or(x1, x2)
@@ -1514,6 +1533,7 @@ def test_logical_or(x1, x2):
     )
 
 
+@pytest.mark.has_setup_funcs
 @given(*hh.two_mutual_arrays([xp.bool]))
 def test_logical_xor(x1, x2):
     out = xp.logical_xor(x1, x2)
@@ -1546,6 +1566,7 @@ def test_minimum(x1, x2):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("multiply", dh.numeric_dtypes))
 @given(data=st.data())
 def test_multiply(ctx, data):
@@ -1577,6 +1598,7 @@ def test_negative(ctx, data):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("not_equal", dh.all_dtypes))
 @given(data=st.data())
 def test_not_equal(ctx, data):
@@ -1598,6 +1620,7 @@ def test_not_equal(ctx, data):
 
 
 @pytest.mark.min_version("2024.12")
+@pytest.mark.has_setup_funcs
 @given(
     shapes=hh.two_mutually_broadcastable_shapes,
     dtype=hh.real_floating_dtypes,
@@ -1617,6 +1640,8 @@ def test_nextafter(shapes, dtype, data):
         out=out
     )
 
+
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_unary_params("positive", dh.numeric_dtypes))
 @given(data=st.data())
 def test_positive(ctx, data):
@@ -1629,6 +1654,7 @@ def test_positive(ctx, data):
     ph.assert_array_elements(ctx.func_name, out=out, expected=x)
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("pow", dh.numeric_dtypes))
 @given(data=st.data())
 def test_pow(ctx, data):
@@ -1676,6 +1702,7 @@ def test_reciprocal(x):
 
 
 @pytest.mark.skip(reason="flaky")
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("remainder", dh.real_dtypes))
 @given(data=st.data())
 def test_remainder(ctx, data):
@@ -1770,6 +1797,7 @@ def test_sqrt(x):
     )
 
 
+@pytest.mark.has_setup_funcs
 @pytest.mark.parametrize("ctx", make_binary_params("subtract", dh.numeric_dtypes))
 @given(data=st.data())
 def test_subtract(ctx, data):
@@ -1923,6 +1951,7 @@ def test_binary_with_scalars_bitwise_shifts(func_data, x1x2):
 
 
 @pytest.mark.min_version("2024.12")
+@pytest.mark.has_setup_funcs
 @pytest.mark.unvectorized
 @given(
     x1x2=hh.array_and_py_scalar([xp.int32]),

array_api_tests/test_utility_functions.py

@@ -67,6 +67,7 @@ def test_any(x, data):
 
 @pytest.mark.unvectorized
 @pytest.mark.min_version("2024.12")
+@pytest.mark.has_setup_funcs
 @given(
     x=hh.arrays(hh.numeric_dtypes, hh.shapes(min_dims=1, min_side=1)),
     data=st.data(),

conftest.py

@@ -98,6 +98,11 @@ def pytest_configure(config):
         "markers",
         "unvectorized: asserts against values via element-wise iteration (not performative!)",
     )
+    config.addinivalue_line(
+        "markers",
+        "has_setup_funcs: run when essential draw data setup functions used "
+        "by Hypothesis are available in the namespace",
+    )
     # Hypothesis
     deadline = None if config.getoption("--hypothesis-disable-deadline") else 800
     settings.register_profile(
@@ -245,6 +250,15 @@ def pytest_collection_modifyitems(config, items):
                         reason=f"requires ARRAY_API_TESTS_VERSION >= {min_version}"
                     )
                 )
+        # skip if namespace doesn't support essential draw data setup functions
+        if any(m.name == "has_setup_funcs" for m in markers):
+            essential_functions = ["asarray", "isnan", "reshape", "zeros"]
+            if not all(hasattr(xp, func_name) for func_name in essential_functions):
+                item.add_marker(
+                    mark.skip(reason="At least one of the essential data setup "
+                                     "functions is not present in the namespace: "
+                                     f"{essential_functions}")
+                )
         # reduce max generated Hypothesis example for unvectorized tests
         if any(m.name == "unvectorized" for m in markers):
             # TODO: limit generated examples when settings already applied