6 from pybind11_tests
import numpy_array
as m
8 np = pytest.importorskip(
"numpy")
14 for size_check
in m.get_platform_dtype_size_checks():
16 assert size_check.size_cpp == size_check.size_numpy, size_check
18 for check
in m.get_concrete_dtype_checks():
20 assert check.numpy == check.pybind11, check
21 if check.numpy.num != check.pybind11.num:
23 "NOTE: typenum mismatch for {}: {} != {}".format(
24 check, check.numpy.num, check.pybind11.num
29 @pytest.fixture(scope=
"function")
31 return np.array([[1, 2, 3], [4, 5, 6]],
"=u2")
37 assert all(m.shape(a) == [])
38 assert all(m.strides(a) == [])
39 with pytest.raises(IndexError)
as excinfo:
41 assert str(excinfo.value) ==
"invalid axis: 0 (ndim = 0)"
42 with pytest.raises(IndexError)
as excinfo:
44 assert str(excinfo.value) ==
"invalid axis: 0 (ndim = 0)"
47 assert m.itemsize(a) == 8
48 assert m.nbytes(a) == 8
51 a = np.array([[1, 2, 3], [4, 5, 6]],
"u2").view()
52 a.flags.writeable =
False
54 assert all(m.shape(a) == [2, 3])
55 assert m.shape(a, 0) == 2
56 assert m.shape(a, 1) == 3
57 assert all(m.strides(a) == [6, 2])
58 assert m.strides(a, 0) == 6
59 assert m.strides(a, 1) == 2
60 with pytest.raises(IndexError)
as excinfo:
62 assert str(excinfo.value) ==
"invalid axis: 2 (ndim = 2)"
63 with pytest.raises(IndexError)
as excinfo:
65 assert str(excinfo.value) ==
"invalid axis: 2 (ndim = 2)"
66 assert not m.writeable(a)
68 assert m.itemsize(a) == 2
69 assert m.nbytes(a) == 12
70 assert not m.owndata(a)
73 @pytest.mark.parametrize(
74 "args, ret", [([], 0), ([0], 0), ([1], 3), ([0, 1], 1), ([1, 2], 5)]
77 assert m.index_at(arr, *args) == ret
78 assert m.index_at_t(arr, *args) == ret
79 assert m.offset_at(arr, *args) == ret * arr.dtype.itemsize
80 assert m.offset_at_t(arr, *args) == ret * arr.dtype.itemsize
94 with pytest.raises(IndexError)
as excinfo:
96 assert str(excinfo.value) ==
"too many indices for an array: 3 (ndim = 2)"
99 @pytest.mark.parametrize(
102 ([], [1, 2, 3, 4, 5, 6]),
104 ([0, 1], [2, 3, 4, 5, 6]),
109 from sys
import byteorder
111 assert all(m.data_t(arr, *args) == ret)
112 assert all(m.data(arr, *args)[(0
if byteorder ==
"little" else 1) :: 2] == ret)
113 assert all(m.data(arr, *args)[(1
if byteorder ==
"little" else 0) :: 2] == 0)
116 @pytest.mark.parametrize(
"dim", [0, 1, 3])
118 for func
in m.at_t, m.mutate_at_t:
119 with pytest.raises(IndexError)
as excinfo:
120 func(arr, *([0] * dim))
121 assert str(excinfo.value) ==
"index dimension mismatch: {} (ndim = 2)".format(
127 assert m.at_t(arr, 0, 2) == 3
128 assert m.at_t(arr, 1, 0) == 4
130 assert all(m.mutate_at_t(arr, 0, 2).ravel() == [1, 2, 4, 4, 5, 6])
131 assert all(m.mutate_at_t(arr, 1, 0).ravel() == [1, 2, 4, 5, 5, 6])
135 arr.flags.writeable =
False
138 (m.mutate_data_t, ()),
139 (m.mutate_at_t, (0, 0)),
141 with pytest.raises(ValueError)
as excinfo:
143 assert str(excinfo.value) ==
"array is not writeable"
147 assert all(m.mutate_data(arr).ravel() == [2, 4, 6, 8, 10, 12])
148 assert all(m.mutate_data(arr).ravel() == [4, 8, 12, 16, 20, 24])
149 assert all(m.mutate_data(arr, 1).ravel() == [4, 8, 12, 32, 40, 48])
150 assert all(m.mutate_data(arr, 0, 1).ravel() == [4, 16, 24, 64, 80, 96])
151 assert all(m.mutate_data(arr, 1, 2).ravel() == [4, 16, 24, 64, 80, 192])
153 assert all(m.mutate_data_t(arr).ravel() == [5, 17, 25, 65, 81, 193])
154 assert all(m.mutate_data_t(arr).ravel() == [6, 18, 26, 66, 82, 194])
155 assert all(m.mutate_data_t(arr, 1).ravel() == [6, 18, 26, 67, 83, 195])
156 assert all(m.mutate_data_t(arr, 0, 1).ravel() == [6, 19, 27, 68, 84, 196])
157 assert all(m.mutate_data_t(arr, 1, 2).ravel() == [6, 19, 27, 68, 84, 197])
171 with pytest.raises(IndexError)
as excinfo:
173 assert str(excinfo.value) ==
"index 2 is out of bounds for axis 0 with size 2"
174 with pytest.raises(IndexError)
as excinfo:
176 assert str(excinfo.value) ==
"index 4 is out of bounds for axis 1 with size 3"
180 assert m.make_c_array().flags.c_contiguous
181 assert not m.make_c_array().flags.f_contiguous
182 assert m.make_f_array().flags.f_contiguous
183 assert not m.make_f_array().flags.c_contiguous
187 m.make_empty_shaped_array()
190 assert m.scalar_int().ndim == 0
191 assert m.scalar_int().shape == ()
192 assert m.scalar_int() == 42
196 def assert_references(a, b, base=None):
197 from distutils.version
import LooseVersion
202 assert a.__array_interface__[
"data"][0] == b.__array_interface__[
"data"][0]
203 assert a.shape == b.shape
204 assert a.strides == b.strides
205 assert a.flags.c_contiguous == b.flags.c_contiguous
206 assert a.flags.f_contiguous == b.flags.f_contiguous
207 assert a.flags.writeable == b.flags.writeable
208 assert a.flags.aligned == b.flags.aligned
209 if LooseVersion(np.__version__) >= LooseVersion(
"1.14.0"):
210 assert a.flags.writebackifcopy == b.flags.writebackifcopy
212 assert a.flags.updateifcopy == b.flags.updateifcopy
213 assert np.all(a == b)
214 assert not b.flags.owndata
215 assert b.base
is base
216 if a.flags.writeable
and a.ndim == 2:
218 assert b[0, 0] == 1234
220 a1 = np.array([1, 2], dtype=np.int16)
221 assert a1.flags.owndata
and a1.base
is None
223 assert_references(a1, a2)
225 a1 = np.array([[1, 2], [3, 4]], dtype=np.float32, order=
"F")
226 assert a1.flags.owndata
and a1.base
is None
228 assert_references(a1, a2)
230 a1 = np.array([[1, 2], [3, 4]], dtype=np.float32, order=
"C")
231 a1.flags.writeable =
False
233 assert_references(a1, a2)
235 a1 = np.random.random((4, 4, 4))
237 assert_references(a1, a2)
241 assert_references(a1t, a2, a1)
245 assert_references(a1d, a2, a1)
247 a1m = a1[::-1, ::-1, ::-1]
249 assert_references(a1m, a2, a1)
255 ac_view_1 = ac.numpy_view()
256 ac_view_2 = ac.numpy_view()
257 assert np.all(ac_view_1 == np.array([1, 2], dtype=np.int32))
264 ArrayClass::numpy_view()
265 ArrayClass::numpy_view()
270 assert ac_view_2[0] == 4
271 assert ac_view_2[1] == 3
286 m.function_taking_uint64(123)
287 m.function_taking_uint64(np.uint64(123))
291 assert m.isinstance_untyped(np.array([1, 2, 3]),
"not an array")
292 assert m.isinstance_typed(np.array([1.0, 2.0, 3.0]))
296 defaults = m.default_constructors()
297 for a
in defaults.values():
299 assert defaults[
"array"].dtype == np.array([]).dtype
300 assert defaults[
"array_t<int32>"].dtype == np.int32
301 assert defaults[
"array_t<double>"].dtype == np.float64
303 results = m.converting_constructors([1, 2, 3])
304 for a
in results.values():
305 np.testing.assert_array_equal(a, [1, 2, 3])
306 assert results[
"array"].dtype == np.int_
307 assert results[
"array_t<int32>"].dtype == np.int32
308 assert results[
"array_t<double>"].dtype == np.float64
313 assert m.overloaded(np.array([1], dtype=
"float64")) ==
"double"
314 assert m.overloaded(np.array([1], dtype=
"float32")) ==
"float"
315 assert m.overloaded(np.array([1], dtype=
"ushort")) ==
"unsigned short"
316 assert m.overloaded(np.array([1], dtype=
"intc")) ==
"int"
317 assert m.overloaded(np.array([1], dtype=
"longlong")) ==
"long long"
318 assert m.overloaded(np.array([1], dtype=
"complex")) ==
"double complex"
319 assert m.overloaded(np.array([1], dtype=
"csingle")) ==
"float complex"
322 assert m.overloaded(np.array([1], dtype=
"uint8")) ==
"double"
324 with pytest.raises(TypeError)
as excinfo:
325 m.overloaded(
"not an array")
329 overloaded(): incompatible function arguments. The following argument types are supported:
330 1. (arg0: numpy.ndarray[numpy.float64]) -> str
331 2. (arg0: numpy.ndarray[numpy.float32]) -> str
332 3. (arg0: numpy.ndarray[numpy.int32]) -> str
333 4. (arg0: numpy.ndarray[numpy.uint16]) -> str
334 5. (arg0: numpy.ndarray[numpy.int64]) -> str
335 6. (arg0: numpy.ndarray[numpy.complex128]) -> str
336 7. (arg0: numpy.ndarray[numpy.complex64]) -> str
338 Invoked with: 'not an array'
342 assert m.overloaded2(np.array([1], dtype=
"float64")) ==
"double"
343 assert m.overloaded2(np.array([1], dtype=
"float32")) ==
"float"
344 assert m.overloaded2(np.array([1], dtype=
"complex64")) ==
"float complex"
345 assert m.overloaded2(np.array([1], dtype=
"complex128")) ==
"double complex"
346 assert m.overloaded2(np.array([1], dtype=
"float32")) ==
"float"
348 assert m.overloaded3(np.array([1], dtype=
"float64")) ==
"double"
349 assert m.overloaded3(np.array([1], dtype=
"intc")) ==
"int"
351 overloaded3(): incompatible function arguments. The following argument types are supported:
352 1. (arg0: numpy.ndarray[numpy.int32]) -> str
353 2. (arg0: numpy.ndarray[numpy.float64]) -> str
357 with pytest.raises(TypeError)
as excinfo:
358 m.overloaded3(np.array([1], dtype=
"uintc"))
359 assert msg(excinfo.value) == expected_exc +
repr(np.array([1], dtype=
"uint32"))
360 with pytest.raises(TypeError)
as excinfo:
361 m.overloaded3(np.array([1], dtype=
"float32"))
362 assert msg(excinfo.value) == expected_exc +
repr(np.array([1.0], dtype=
"float32"))
363 with pytest.raises(TypeError)
as excinfo:
364 m.overloaded3(np.array([1], dtype=
"complex"))
365 assert msg(excinfo.value) == expected_exc +
repr(np.array([1.0 + 0.0j]))
368 assert m.overloaded4(np.array([1], dtype=
"double")) ==
"double"
369 assert m.overloaded4(np.array([1], dtype=
"longlong")) ==
"long long"
373 assert m.overloaded4(np.array([1], dtype=
"float32")) ==
"double"
374 assert m.overloaded4(np.array([1], dtype=
"short")) ==
"long long"
376 assert m.overloaded5(np.array([1], dtype=
"double")) ==
"double"
377 assert m.overloaded5(np.array([1], dtype=
"uintc")) ==
"unsigned int"
378 assert m.overloaded5(np.array([1], dtype=
"float32")) ==
"unsigned int"
382 """Tests fix for #685 - ndarray shouldn't go to std::string overload"""
384 assert m.issue685(
"abc") ==
"string"
385 assert m.issue685(np.array([97, 98, 99], dtype=
"b")) ==
"array"
386 assert m.issue685(123) ==
"other"
390 z1 = np.array([[1, 2], [3, 4]], dtype=
"float64")
392 assert np.all(z1 == [[11, 12], [13, 14]])
394 with pytest.raises(ValueError)
as excinfo:
395 m.proxy_add2(np.array([1.0, 2, 3]), 5.0)
397 msg(excinfo.value) ==
"array has incorrect number of dimensions: 1; expected 2"
400 expect_c = np.ndarray(shape=(3, 3, 3), buffer=np.array(range(3, 30)), dtype=
"int")
401 assert np.all(m.proxy_init3(3.0) == expect_c)
402 expect_f = np.transpose(expect_c)
403 assert np.all(m.proxy_init3F(3.0) == expect_f)
405 assert m.proxy_squared_L2_norm(np.array(range(6))) == 55
406 assert m.proxy_squared_L2_norm(np.array(range(6), dtype=
"float64")) == 55
408 assert m.proxy_auxiliaries2(z1) == [11, 11,
True, 2, 8, 2, 2, 4, 32]
409 assert m.proxy_auxiliaries2(z1) == m.array_auxiliaries2(z1)
411 assert m.proxy_auxiliaries1_const_ref(z1[0, :])
412 assert m.proxy_auxiliaries2_const_ref(z1)
416 z1 = np.array([[1, 2], [3, 4]], dtype=
"float64")
417 m.proxy_add2_dyn(z1, 10)
418 assert np.all(z1 == [[11, 12], [13, 14]])
420 expect_c = np.ndarray(shape=(3, 3, 3), buffer=np.array(range(3, 30)), dtype=
"int")
421 assert np.all(m.proxy_init3_dyn(3.0) == expect_c)
423 assert m.proxy_auxiliaries2_dyn(z1) == [11, 11,
True, 2, 8, 2, 2, 4, 32]
424 assert m.proxy_auxiliaries2_dyn(z1) == m.array_auxiliaries2(z1)
428 with pytest.raises(ValueError)
as excinfo:
430 assert str(excinfo.value) ==
"cannot create a pybind11::array from a nullptr"
432 with pytest.raises(ValueError)
as excinfo:
433 m.array_t_fail_test()
434 assert str(excinfo.value) ==
"cannot create a pybind11::array_t from a nullptr"
436 with pytest.raises(ValueError)
as excinfo:
437 m.array_fail_test_negative_size()
438 assert str(excinfo.value) ==
"negative dimensions are not allowed"
442 assert m.array_initializer_list1().shape == (1,)
443 assert m.array_initializer_list2().shape == (1, 2)
444 assert m.array_initializer_list3().shape == (1, 2, 3)
445 assert m.array_initializer_list4().shape == (1, 2, 3, 4)
449 a = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=
"float64")
452 assert np.all(a == [[1, 2, 3], [4, 5, 6], [7, 8, 9]])
455 m.array_resize3(a, 4,
False)
459 m.array_resize3(a, 3,
True)
460 except ValueError
as e:
461 assert str(e).startswith(
"cannot resize an array")
465 m.array_resize3(b, 3,
False)
466 except ValueError
as e:
467 assert str(e).startswith(
"cannot resize this array: it does not own its data")
470 assert b.shape == (8, 8)
473 @pytest.mark.xfail(
"env.PYPY")
475 a = m.create_and_resize(2)
477 assert np.all(a == 42.0)
481 a = m.index_using_ellipsis(np.zeros((5, 6, 7)))
482 assert a.shape == (6,)
485 @pytest.mark.parametrize(
"forcecast", [
False,
True])
486 @pytest.mark.parametrize(
"contiguity", [
None,
"C",
"F"])
487 @pytest.mark.parametrize(
"noconvert", [
False,
True])
488 @pytest.mark.filterwarnings(
489 "ignore:Casting complex values to real discards the imaginary part:numpy.ComplexWarning"
492 function_name =
"accept_double"
493 if contiguity ==
"C":
494 function_name +=
"_c_style"
495 elif contiguity ==
"F":
496 function_name +=
"_f_style"
498 function_name +=
"_forcecast"
500 function_name +=
"_noconvert"
501 function =
getattr(m, function_name)
503 for dtype
in [np.dtype(
"float32"), np.dtype(
"float64"), np.dtype(
"complex128")]:
504 for order
in [
"C",
"F"]:
505 for shape
in [(2, 2), (1, 3, 1, 1), (1, 1, 1), (0,)]:
510 should_raise = dtype.name ==
"complex128" and not forcecast
515 trivially_contiguous = sum(1
for d
in shape
if d > 1) <= 1
516 should_raise = dtype.name !=
"float64" or (
517 contiguity
is not None
518 and contiguity != order
519 and not trivially_contiguous
522 array = np.zeros(shape, dtype=dtype, order=order)
527 TypeError, match=
"incompatible function arguments"
532 @pytest.mark.xfail(
"env.PYPY")
534 from sys
import getrefcount
536 dtype = np.dtype(np.float_)
537 a = np.array([1], dtype=dtype)
538 before = getrefcount(dtype)
540 after = getrefcount(dtype)
541 assert after == before