from __future__ import annotations import copy import operator as op import pickle import warnings import pytest from pint import DimensionalityError, OffsetUnitCalculusError, UnitStrippedWarning from pint.compat import np from pint.testsuite import helpers from pint.testsuite.test_umath import TestUFuncs @helpers.requires_numpy class TestNumpyMethods: @classmethod def setup_class(cls): from pint import _DEFAULT_REGISTRY cls.ureg = _DEFAULT_REGISTRY cls.Q_ = cls.ureg.Quantity @classmethod def teardown_class(cls): cls.ureg = None cls.Q_ = None @property def q(self): return [[1, 2], [3, 4]] * self.ureg.m @property def q_scalar(self): return np.array(5) * self.ureg.m @property def q_nan(self): return [[1, 2], [3, np.nan]] * self.ureg.m @property def q_zero_or_nan(self): return [[0, 0], [0, np.nan]] * self.ureg.m @property def q_temperature(self): return self.Q_([[1, 2], [3, 4]], self.ureg.degC) def assertNDArrayEqual(self, actual, desired): # Assert that the given arrays are equal, and are not Quantities np.testing.assert_array_equal(actual, desired) assert not isinstance(actual, self.Q_) assert not isinstance(desired, self.Q_) class TestNumpyArrayCreation(TestNumpyMethods): # https://docs.scipy.org/doc/numpy/reference/routines.array-creation.html @helpers.requires_array_function_protocol() def test_ones_like(self): self.assertNDArrayEqual(np.ones_like(self.q), np.array([[1, 1], [1, 1]])) @helpers.requires_array_function_protocol() def test_zeros_like(self): self.assertNDArrayEqual(np.zeros_like(self.q), np.array([[0, 0], [0, 0]])) @helpers.requires_array_function_protocol() def test_empty_like(self): ret = np.empty_like(self.q) assert ret.shape == (2, 2) assert isinstance(ret, np.ndarray) @helpers.requires_array_function_protocol() def test_full_like(self): helpers.assert_quantity_equal( np.full_like(self.q, self.Q_(0, self.ureg.degC)), self.Q_([[0, 0], [0, 0]], self.ureg.degC), ) self.assertNDArrayEqual(np.full_like(self.q, 2), np.array([[2, 2], [2, 2]])) class TestNumpyArrayManipulation(TestNumpyMethods): # TODO # https://www.numpy.org/devdocs/reference/routines.array-manipulation.html # copyto # broadcast # asarray asanyarray asmatrix asfarray asfortranarray ascontiguousarray asarray_chkfinite asscalar require # Changing array shape def test_flatten(self): helpers.assert_quantity_equal(self.q.flatten(), [1, 2, 3, 4] * self.ureg.m) def test_flat(self): for q, v in zip(self.q.flat, [1, 2, 3, 4]): assert q == v * self.ureg.m def test_reshape(self): helpers.assert_quantity_equal( self.q.reshape([1, 4]), [[1, 2, 3, 4]] * self.ureg.m ) def test_ravel(self): helpers.assert_quantity_equal(self.q.ravel(), [1, 2, 3, 4] * self.ureg.m) @helpers.requires_array_function_protocol() def test_ravel_numpy_func(self): helpers.assert_quantity_equal(np.ravel(self.q), [1, 2, 3, 4] * self.ureg.m) # Transpose-like operations @helpers.requires_array_function_protocol() def test_moveaxis(self): helpers.assert_quantity_equal( np.moveaxis(self.q, 1, 0), np.array([[1, 2], [3, 4]]).T * self.ureg.m ) @helpers.requires_array_function_protocol() def test_rollaxis(self): helpers.assert_quantity_equal( np.rollaxis(self.q, 1), np.array([[1, 2], [3, 4]]).T * self.ureg.m ) @helpers.requires_array_function_protocol() def test_swapaxes(self): helpers.assert_quantity_equal( np.swapaxes(self.q, 1, 0), np.array([[1, 2], [3, 4]]).T * self.ureg.m ) def test_transpose(self): helpers.assert_quantity_equal( self.q.transpose(), [[1, 3], [2, 4]] * self.ureg.m ) @helpers.requires_array_function_protocol() def test_transpose_numpy_func(self): helpers.assert_quantity_equal( np.transpose(self.q), [[1, 3], [2, 4]] * self.ureg.m ) @helpers.requires_array_function_protocol() def test_flip_numpy_func(self): helpers.assert_quantity_equal( np.flip(self.q, axis=0), [[3, 4], [1, 2]] * self.ureg.m ) # Changing number of dimensions @helpers.requires_array_function_protocol() def test_atleast_1d(self): actual = np.atleast_1d(self.Q_(0, self.ureg.degC), self.q.flatten()) expected = (self.Q_(np.array([0]), self.ureg.degC), self.q.flatten()) for ind_actual, ind_expected in zip(actual, expected): helpers.assert_quantity_equal(ind_actual, ind_expected) helpers.assert_quantity_equal(np.atleast_1d(self.q), self.q) @helpers.requires_array_function_protocol() def test_atleast_2d(self): actual = np.atleast_2d(self.Q_(0, self.ureg.degC), self.q.flatten()) expected = ( self.Q_(np.array([[0]]), self.ureg.degC), np.array([[1, 2, 3, 4]]) * self.ureg.m, ) for ind_actual, ind_expected in zip(actual, expected): helpers.assert_quantity_equal(ind_actual, ind_expected) helpers.assert_quantity_equal(np.atleast_2d(self.q), self.q) @helpers.requires_array_function_protocol() def test_atleast_3d(self): actual = np.atleast_3d(self.Q_(0, self.ureg.degC), self.q.flatten()) expected = ( self.Q_(np.array([[[0]]]), self.ureg.degC), np.array([[[1], [2], [3], [4]]]) * self.ureg.m, ) for ind_actual, ind_expected in zip(actual, expected): helpers.assert_quantity_equal(ind_actual, ind_expected) helpers.assert_quantity_equal( np.atleast_3d(self.q), np.array([[[1], [2]], [[3], [4]]]) * self.ureg.m ) @helpers.requires_array_function_protocol() def test_broadcast_to(self): helpers.assert_quantity_equal( np.broadcast_to(self.q[:, 1], (2, 2)), np.array([[2, 4], [2, 4]]) * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_expand_dims(self): helpers.assert_quantity_equal( np.expand_dims(self.q, 0), np.array([[[1, 2], [3, 4]]]) * self.ureg.m ) @helpers.requires_array_function_protocol() def test_squeeze(self): helpers.assert_quantity_equal(np.squeeze(self.q), self.q) helpers.assert_quantity_equal( self.q.reshape([1, 4]).squeeze(), [1, 2, 3, 4] * self.ureg.m ) # Changing number of dimensions # Joining arrays @helpers.requires_array_function_protocol() def test_concat_stack(self, subtests): for func in (np.concatenate, np.stack, np.hstack, np.vstack, np.dstack): with subtests.test(func=func): helpers.assert_quantity_equal( func([self.q] * 2), self.Q_(func([self.q.m] * 2), self.ureg.m) ) # One or more of the args is a bare array full of zeros or NaNs helpers.assert_quantity_equal( func([self.q_zero_or_nan.m, self.q]), self.Q_(func([self.q_zero_or_nan.m, self.q.m]), self.ureg.m), ) # One or more of the args is a bare array with at least one non-zero, # non-NaN element nz = self.q_zero_or_nan nz.m[0, 0] = 1 with pytest.raises(DimensionalityError): func([nz.m, self.q]) @helpers.requires_array_function_protocol() def test_block_column_stack(self, subtests): for func in (np.block, np.column_stack): with subtests.test(func=func): helpers.assert_quantity_equal( func([self.q[:, 0], self.q[:, 1]]), self.Q_(func([self.q[:, 0].m, self.q[:, 1].m]), self.ureg.m), ) # One or more of the args is a bare array full of zeros or NaNs helpers.assert_quantity_equal( func( [ self.q_zero_or_nan[:, 0].m, self.q[:, 0], self.q_zero_or_nan[:, 1].m, ] ), self.Q_( func( [ self.q_zero_or_nan[:, 0].m, self.q[:, 0].m, self.q_zero_or_nan[:, 1].m, ] ), self.ureg.m, ), ) # One or more of the args is a bare array with at least one non-zero, # non-NaN element nz = self.q_zero_or_nan nz.m[0, 0] = 1 with pytest.raises(DimensionalityError): func([nz[:, 0].m, self.q[:, 0]]) @helpers.requires_array_function_protocol() def test_append(self): helpers.assert_quantity_equal( np.append(self.q, [[0, 0]] * self.ureg.m, axis=0), [[1, 2], [3, 4], [0, 0]] * self.ureg.m, ) def test_astype(self): actual = self.q.astype(np.float32) expected = self.Q_(np.array([[1.0, 2.0], [3.0, 4.0]], dtype=np.float32), "m") helpers.assert_quantity_equal(actual, expected) assert actual.m.dtype == expected.m.dtype def test_item(self): helpers.assert_quantity_equal(self.Q_([[0]], "m").item(), 0 * self.ureg.m) def test_broadcast_arrays(self): x = self.Q_(np.array([[1, 2, 3]]), "m") y = self.Q_(np.array([[4], [5]]), "nm") result = np.broadcast_arrays(x, y) expected = self.Q_( [ [[1.0, 2.0, 3.0], [1.0, 2.0, 3.0]], [[4e-09, 4e-09, 4e-09], [5e-09, 5e-09, 5e-09]], ], "m", ) helpers.assert_quantity_equal(result, expected) result = np.broadcast_arrays(x, y, subok=True) helpers.assert_quantity_equal(result, expected) def test_roll(self): helpers.assert_quantity_equal( np.roll(self.q, 1), [[4, 1], [2, 3]] * self.ureg.m ) class TestNumpyMathematicalFunctions(TestNumpyMethods): # https://www.numpy.org/devdocs/reference/routines.math.html # Trigonometric functions @helpers.requires_array_function_protocol() def test_unwrap(self): helpers.assert_quantity_equal( np.unwrap([0, 3 * np.pi] * self.ureg.radians), [0, np.pi] ) helpers.assert_quantity_equal( np.unwrap([0, 540] * self.ureg.deg), [0, 180] * self.ureg.deg ) # Rounding @helpers.requires_array_function_protocol() def test_fix(self): helpers.assert_quantity_equal(np.fix(3.13 * self.ureg.m), 3.0 * self.ureg.m) helpers.assert_quantity_equal(np.fix(3.0 * self.ureg.m), 3.0 * self.ureg.m) helpers.assert_quantity_equal( np.fix([2.1, 2.9, -2.1, -2.9] * self.ureg.m), [2.0, 2.0, -2.0, -2.0] * self.ureg.m, ) # Sums, products, differences @helpers.requires_array_function_protocol() def test_prod(self): axis = 0 where = [[True, False], [True, True]] helpers.assert_quantity_equal(self.q.prod(), 24 * self.ureg.m**4) helpers.assert_quantity_equal(self.q.prod(axis=axis), [3, 8] * self.ureg.m**2) helpers.assert_quantity_equal(self.q.prod(where=where), 12 * self.ureg.m**3) @helpers.requires_array_function_protocol() def test_prod_numpy_func(self): axis = 0 where = [[True, False], [True, True]] helpers.assert_quantity_equal(np.prod(self.q), 24 * self.ureg.m**4) helpers.assert_quantity_equal( np.prod(self.q, axis=axis), [3, 8] * self.ureg.m**2 ) helpers.assert_quantity_equal(np.prod(self.q, where=where), 12 * self.ureg.m**3) with pytest.raises(DimensionalityError): np.prod(self.q, axis=axis, where=where) helpers.assert_quantity_equal( np.prod(self.q, axis=axis, where=[[True, False], [False, True]]), [1, 4] * self.ureg.m, ) helpers.assert_quantity_equal( np.prod(self.q, axis=axis, where=[True, False]), [3, 1] * self.ureg.m**2 ) @helpers.requires_array_function_protocol() def test_nanprod_numpy_func(self): helpers.assert_quantity_equal(np.nanprod(self.q_nan), 6 * self.ureg.m**3) helpers.assert_quantity_equal( np.nanprod(self.q_nan, axis=0), [3, 2] * self.ureg.m**2 ) helpers.assert_quantity_equal( np.nanprod(self.q_nan, axis=1), [2, 3] * self.ureg.m**2 ) def test_sum(self): assert self.q.sum() == 10 * self.ureg.m helpers.assert_quantity_equal(self.q.sum(0), [4, 6] * self.ureg.m) helpers.assert_quantity_equal(self.q.sum(1), [3, 7] * self.ureg.m) @helpers.requires_array_function_protocol() def test_sum_numpy_func(self): helpers.assert_quantity_equal(np.sum(self.q, axis=0), [4, 6] * self.ureg.m) with pytest.raises(OffsetUnitCalculusError): np.sum(self.q_temperature) @helpers.requires_array_function_protocol() def test_nansum_numpy_func(self): helpers.assert_quantity_equal( np.nansum(self.q_nan, axis=0), [4, 2] * self.ureg.m ) def test_cumprod(self): with pytest.raises(DimensionalityError): self.q.cumprod() helpers.assert_quantity_equal((self.q / self.ureg.m).cumprod(), [1, 2, 6, 24]) @helpers.requires_array_function_protocol() def test_cumprod_numpy_func(self): with pytest.raises(DimensionalityError): np.cumprod(self.q) helpers.assert_quantity_equal(np.cumprod(self.q / self.ureg.m), [1, 2, 6, 24]) helpers.assert_quantity_equal( np.cumprod(self.q / self.ureg.m, axis=1), [[1, 2], [3, 12]] ) @helpers.requires_array_function_protocol() def test_nancumprod_numpy_func(self): with pytest.raises(DimensionalityError): np.nancumprod(self.q_nan) helpers.assert_quantity_equal( np.nancumprod(self.q_nan / self.ureg.m), [1, 2, 6, 6] ) @helpers.requires_array_function_protocol() def test_diff(self): helpers.assert_quantity_equal(np.diff(self.q, 1), [[1], [1]] * self.ureg.m) helpers.assert_quantity_equal( np.diff(self.q_temperature, 1), [[1], [1]] * self.ureg.delta_degC ) @helpers.requires_array_function_protocol() def test_ediff1d(self): helpers.assert_quantity_equal(np.ediff1d(self.q), [1, 1, 1] * self.ureg.m) helpers.assert_quantity_equal( np.ediff1d(self.q_temperature), [1, 1, 1] * self.ureg.delta_degC ) @helpers.requires_array_function_protocol() def test_gradient(self): grad = np.gradient([[1, 1], [3, 4]] * self.ureg.m, 1 * self.ureg.J) helpers.assert_quantity_equal( grad[0], [[2.0, 3.0], [2.0, 3.0]] * self.ureg.m / self.ureg.J ) helpers.assert_quantity_equal( grad[1], [[0.0, 0.0], [1.0, 1.0]] * self.ureg.m / self.ureg.J ) grad = np.gradient(self.Q_([[1, 1], [3, 4]], self.ureg.degC), 1 * self.ureg.J) helpers.assert_quantity_equal( grad[0], [[2.0, 3.0], [2.0, 3.0]] * self.ureg.delta_degC / self.ureg.J ) helpers.assert_quantity_equal( grad[1], [[0.0, 0.0], [1.0, 1.0]] * self.ureg.delta_degC / self.ureg.J ) @helpers.requires_array_function_protocol() def test_cross(self): a = [[3, -3, 1]] * self.ureg.kPa b = [[4, 9, 2]] * self.ureg.m**2 helpers.assert_quantity_equal( np.cross(a, b), [[-15, -2, 39]] * self.ureg.kPa * self.ureg.m**2 ) # NP2: Remove this when we only support np>=2.0 @helpers.requires_array_function_protocol() def test_trapz(self): helpers.assert_quantity_equal( np.trapz([1.0, 2.0, 3.0, 4.0] * self.ureg.J, dx=1 * self.ureg.m), 7.5 * self.ureg.J * self.ureg.m, ) @helpers.requires_array_function_protocol() # NP2: Remove this when we only support np>=2.0 # trapezoid added in numpy 2.0 @helpers.requires_numpy_at_least("2.0") def test_trapezoid(self): helpers.assert_quantity_equal( np.trapezoid([1.0, 2.0, 3.0, 4.0] * self.ureg.J, dx=1 * self.ureg.m), 7.5 * self.ureg.J * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_dot(self): helpers.assert_quantity_equal( self.q.ravel().dot(np.array([1, 0, 0, 1])), 5 * self.ureg.m ) @helpers.requires_array_function_protocol() def test_dot_numpy_func(self): helpers.assert_quantity_equal( np.dot(self.q.ravel(), [0, 0, 1, 0] * self.ureg.dimensionless), 3 * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_einsum(self): a = np.arange(25).reshape(5, 5) * self.ureg.m b = np.arange(5) * self.ureg.m helpers.assert_quantity_equal(np.einsum("ii", a), 60 * self.ureg.m) helpers.assert_quantity_equal( np.einsum("ii->i", a), np.array([0, 6, 12, 18, 24]) * self.ureg.m ) helpers.assert_quantity_equal(np.einsum("i,i", b, b), 30 * self.ureg.m**2) helpers.assert_quantity_equal( np.einsum("ij,j", a, b), np.array([30, 80, 130, 180, 230]) * self.ureg.m**2, ) @helpers.requires_array_function_protocol() def test_solve(self): A = self.q b = [[3], [7]] * self.ureg.s x = np.linalg.solve(A, b) helpers.assert_quantity_almost_equal(x, self.Q_([[1], [1]], "s / m")) helpers.assert_quantity_almost_equal(np.dot(A, x), b) # Arithmetic operations def test_addition_with_scalar(self): a = np.array([0, 1, 2]) b = 10.0 * self.ureg("gram/kilogram") helpers.assert_quantity_almost_equal( a + b, self.Q_([0.01, 1.01, 2.01], self.ureg.dimensionless) ) helpers.assert_quantity_almost_equal( b + a, self.Q_([0.01, 1.01, 2.01], self.ureg.dimensionless) ) def test_addition_with_incompatible_scalar(self): a = np.array([0, 1, 2]) b = 1.0 * self.ureg.m with pytest.raises(DimensionalityError): op.add(a, b) with pytest.raises(DimensionalityError): op.add(b, a) def test_power(self): arr = np.array(range(3), dtype=float) q = self.Q_(arr, "meter") for op_ in (op.pow, op.ipow, np.power): q_cp = copy.copy(q) with pytest.raises(DimensionalityError): op_(2.0, q_cp) arr_cp = copy.copy(arr) arr_cp = copy.copy(arr) q_cp = copy.copy(q) with pytest.raises(DimensionalityError): op_(q_cp, arr_cp) q_cp = copy.copy(q) q2_cp = copy.copy(q) with pytest.raises(DimensionalityError): op_(q_cp, q2_cp) helpers.assert_quantity_equal( np.power(self.q, self.Q_(2)), self.Q_([[1, 4], [9, 16]], "m**2") ) helpers.assert_quantity_equal( self.q ** self.Q_(2), self.Q_([[1, 4], [9, 16]], "m**2") ) self.assertNDArrayEqual(arr ** self.Q_(2), np.array([0, 1, 4])) def test_sqrt(self): q = self.Q_(100, "m**2") helpers.assert_quantity_equal(np.sqrt(q), self.Q_(10, "m")) def test_cbrt(self): q = self.Q_(1000, "m**3") helpers.assert_quantity_equal(np.cbrt(q), self.Q_(10, "m")) @pytest.mark.xfail @helpers.requires_numpy def test_exponentiation_array_exp_2(self): arr = np.array(range(3), dtype=float) # q = self.Q_(copy.copy(arr), None) q = self.Q_(copy.copy(arr), "meter") arr_cp = copy.copy(arr) q_cp = copy.copy(q) # this fails as expected since numpy 1.8.0 but... with pytest.raises(DimensionalityError): op.pow(arr_cp, q_cp) # ..not for op.ipow ! # q_cp is treated as if it is an array. The units are ignored. # Quantity.__ipow__ is never called arr_cp = copy.copy(arr) q_cp = copy.copy(q) with pytest.raises(DimensionalityError): op.ipow(arr_cp, q_cp) class TestNumpyUnclassified(TestNumpyMethods): def test_tolist(self): with pytest.raises(AttributeError): (5 * self.ureg.m).tolist() assert self.q.tolist() == [ [1 * self.ureg.m, 2 * self.ureg.m], [3 * self.ureg.m, 4 * self.ureg.m], ] def test_fill(self): tmp = self.q tmp.fill(6 * self.ureg.ft) helpers.assert_quantity_equal(tmp, [[6, 6], [6, 6]] * self.ureg.ft) tmp.fill(5 * self.ureg.m) helpers.assert_quantity_equal(tmp, [[5, 5], [5, 5]] * self.ureg.m) def test_take(self): helpers.assert_quantity_equal(self.q.take([0, 1, 2, 3]), self.q.flatten()) def test_put(self): q = [1.0, 2.0, 3.0, 4.0] * self.ureg.m q.put([0, 2], [10.0, 20.0] * self.ureg.m) helpers.assert_quantity_equal(q, [10.0, 2.0, 20.0, 4.0] * self.ureg.m) q = [1.0, 2.0, 3.0, 4.0] * self.ureg.m q.put([0, 2], [1.0, 2.0] * self.ureg.mm) helpers.assert_quantity_equal(q, [0.001, 2.0, 0.002, 4.0] * self.ureg.m) q = [1.0, 2.0, 3.0, 4.0] * self.ureg.m / self.ureg.mm q.put([0, 2], [1.0, 2.0]) helpers.assert_quantity_equal( q, [0.001, 2.0, 0.002, 4.0] * self.ureg.m / self.ureg.mm ) q = [1.0, 2.0, 3.0, 4.0] * self.ureg.m with pytest.raises(DimensionalityError): q.put([0, 2], [4.0, 6.0] * self.ureg.J) with pytest.raises(DimensionalityError): q.put([0, 2], [4.0, 6.0]) def test_repeat(self): helpers.assert_quantity_equal( self.q.repeat(2), [1, 1, 2, 2, 3, 3, 4, 4] * self.ureg.m ) def test_sort(self): q = [4, 5, 2, 3, 1, 6] * self.ureg.m q.sort() helpers.assert_quantity_equal(q, [1, 2, 3, 4, 5, 6] * self.ureg.m) @helpers.requires_array_function_protocol() def test_sort_numpy_func(self): q = [4, 5, 2, 3, 1, 6] * self.ureg.m helpers.assert_quantity_equal(np.sort(q), [1, 2, 3, 4, 5, 6] * self.ureg.m) def test_argsort(self): q = [1, 4, 5, 6, 2, 9] * self.ureg.MeV self.assertNDArrayEqual(q.argsort(), [0, 4, 1, 2, 3, 5]) @helpers.requires_array_function_protocol() def test_argsort_numpy_func(self): self.assertNDArrayEqual(np.argsort(self.q, axis=0), np.array([[0, 0], [1, 1]])) def test_diagonal(self): q = [[1, 2, 3], [1, 2, 3], [1, 2, 3]] * self.ureg.m helpers.assert_quantity_equal(q.diagonal(offset=1), [2, 3] * self.ureg.m) @helpers.requires_array_function_protocol() def test_diagonal_numpy_func(self): q = [[1, 2, 3], [1, 2, 3], [1, 2, 3]] * self.ureg.m helpers.assert_quantity_equal(np.diagonal(q, offset=-1), [1, 2] * self.ureg.m) def test_compress(self): helpers.assert_quantity_equal( self.q.compress([False, True], axis=0), [[3, 4]] * self.ureg.m ) helpers.assert_quantity_equal( self.q.compress([False, True], axis=1), [[2], [4]] * self.ureg.m ) @helpers.requires_array_function_protocol() def test_compress_nep18(self): helpers.assert_quantity_equal( np.compress([False, True], self.q, axis=1), [[2], [4]] * self.ureg.m ) def test_searchsorted(self): q = self.q.flatten() self.assertNDArrayEqual(q.searchsorted([1.5, 2.5] * self.ureg.m), [1, 2]) q = self.q.flatten() with pytest.raises(DimensionalityError): q.searchsorted([1.5, 2.5]) @helpers.requires_array_function_protocol() def test_searchsorted_numpy_func(self): """Test searchsorted as numpy function.""" q = self.q.flatten() self.assertNDArrayEqual(np.searchsorted(q, [1.5, 2.5] * self.ureg.m), [1, 2]) def test_nonzero(self): q = [1, 0, 5, 6, 0, 9] * self.ureg.m self.assertNDArrayEqual(q.nonzero()[0], [0, 2, 3, 5]) @helpers.requires_array_function_protocol() def test_nonzero_numpy_func(self): q = [1, 0, 5, 6, 0, 9] * self.ureg.m self.assertNDArrayEqual(np.nonzero(q)[0], [0, 2, 3, 5]) @helpers.requires_array_function_protocol() def test_any_numpy_func(self): q = [0, 1] * self.ureg.m assert np.any(q) with pytest.raises(ValueError): np.any(self.q_temperature) @helpers.requires_array_function_protocol() def test_all_numpy_func(self): q = [0, 1] * self.ureg.m assert not np.all(q) with pytest.raises(ValueError): np.all(self.q_temperature) @helpers.requires_array_function_protocol() def test_count_nonzero_numpy_func(self): q = [1, 0, 5, 6, 0, 9] * self.ureg.m assert np.count_nonzero(q) == 4 def test_max(self): assert self.q.max() == 4 * self.ureg.m def test_max_numpy_func(self): assert np.max(self.q) == 4 * self.ureg.m @helpers.requires_array_function_protocol() def test_max_with_axis_arg(self): helpers.assert_quantity_equal(np.max(self.q, axis=1), [2, 4] * self.ureg.m) @helpers.requires_array_function_protocol() def test_max_with_initial_arg(self): helpers.assert_quantity_equal( np.max(self.q[..., None], axis=2, initial=3 * self.ureg.m), [[3, 3], [3, 4]] * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_nanmax(self): assert np.nanmax(self.q_nan) == 3 * self.ureg.m def test_argmax(self): assert self.q.argmax() == 3 @helpers.requires_array_function_protocol() def test_argmax_numpy_func(self): self.assertNDArrayEqual(np.argmax(self.q, axis=0), np.array([1, 1])) @helpers.requires_array_function_protocol() def test_nanargmax_numpy_func(self): self.assertNDArrayEqual(np.nanargmax(self.q_nan, axis=0), np.array([1, 0])) def test_maximum(self): helpers.assert_quantity_equal( np.maximum(self.q, self.Q_([0, 5], "m")), self.Q_([[1, 5], [3, 5]], "m") ) def test_min(self): assert self.q.min() == 1 * self.ureg.m @helpers.requires_array_function_protocol() def test_min_numpy_func(self): assert np.min(self.q) == 1 * self.ureg.m @helpers.requires_array_function_protocol() def test_min_with_axis_arg(self): helpers.assert_quantity_equal(np.min(self.q, axis=1), [1, 3] * self.ureg.m) @helpers.requires_array_function_protocol() def test_min_with_initial_arg(self): helpers.assert_quantity_equal( np.min(self.q[..., None], axis=2, initial=3 * self.ureg.m), [[1, 2], [3, 3]] * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_nanmin(self): assert np.nanmin(self.q_nan) == 1 * self.ureg.m def test_argmin(self): assert self.q.argmin() == 0 @helpers.requires_array_function_protocol() def test_argmin_numpy_func(self): self.assertNDArrayEqual(np.argmin(self.q, axis=0), np.array([0, 0])) @helpers.requires_array_function_protocol() def test_nanargmin_numpy_func(self): self.assertNDArrayEqual(np.nanargmin(self.q_nan, axis=0), np.array([0, 0])) def test_minimum(self): helpers.assert_quantity_equal( np.minimum(self.q, self.Q_([0, 5], "m")), self.Q_([[0, 2], [0, 4]], "m") ) # NP2: Can remove Q_(arr).ptp test when we only support numpy>=2 def test_ptp(self): if not np.lib.NumpyVersion(np.__version__) >= "2.0.0b1": assert self.q.ptp() == 3 * self.ureg.m # NP2: Keep this test for numpy>=2, it's only arr.ptp() that is deprecated @helpers.requires_array_function_protocol() def test_ptp_numpy_func(self): helpers.assert_quantity_equal(np.ptp(self.q, axis=0), [2, 2] * self.ureg.m) def test_clip(self): helpers.assert_quantity_equal( self.q.clip(max=2 * self.ureg.m), [[1, 2], [2, 2]] * self.ureg.m ) helpers.assert_quantity_equal( self.q.clip(min=3 * self.ureg.m), [[3, 3], [3, 4]] * self.ureg.m ) helpers.assert_quantity_equal( self.q.clip(min=2 * self.ureg.m, max=3 * self.ureg.m), [[2, 2], [3, 3]] * self.ureg.m, ) helpers.assert_quantity_equal( self.q.clip(3 * self.ureg.m, None), [[3, 3], [3, 4]] * self.ureg.m ) helpers.assert_quantity_equal( self.q.clip(3 * self.ureg.m), [[3, 3], [3, 4]] * self.ureg.m ) with pytest.raises(DimensionalityError): self.q.clip(self.ureg.J) with pytest.raises(DimensionalityError): self.q.clip(1) @helpers.requires_array_function_protocol() def test_clip_numpy_func(self): helpers.assert_quantity_equal( np.clip(self.q, 150 * self.ureg.cm, None), [[1.5, 2], [3, 4]] * self.ureg.m ) def test_round(self): q = [1, 1.33, 5.67, 22] * self.ureg.m helpers.assert_quantity_equal(q.round(0), [1, 1, 6, 22] * self.ureg.m) helpers.assert_quantity_equal(q.round(-1), [0, 0, 10, 20] * self.ureg.m) helpers.assert_quantity_equal(q.round(1), [1, 1.3, 5.7, 22] * self.ureg.m) @helpers.requires_array_function_protocol() def test_round_numpy_func(self): helpers.assert_quantity_equal( np.around(1.0275 * self.ureg.m, decimals=2), 1.03 * self.ureg.m ) helpers.assert_quantity_equal( np.round(1.0275 * self.ureg.m, decimals=2), 1.03 * self.ureg.m ) def test_trace(self): assert self.q.trace() == (1 + 4) * self.ureg.m def test_cumsum(self): helpers.assert_quantity_equal(self.q.cumsum(), [1, 3, 6, 10] * self.ureg.m) @helpers.requires_array_function_protocol() def test_cumsum_numpy_func(self): helpers.assert_quantity_equal( np.cumsum(self.q, axis=0), [[1, 2], [4, 6]] * self.ureg.m ) @helpers.requires_array_function_protocol() def test_nancumsum_numpy_func(self): helpers.assert_quantity_equal( np.nancumsum(self.q_nan, axis=0), [[1, 2], [4, 2]] * self.ureg.m ) def test_mean(self): assert self.q.mean() == 2.5 * self.ureg.m @helpers.requires_array_function_protocol() def test_mean_numpy_func(self): assert np.mean(self.q) == 2.5 * self.ureg.m assert np.mean(self.q_temperature) == self.Q_(2.5, self.ureg.degC) @helpers.requires_array_function_protocol() def test_nanmean_numpy_func(self): assert np.nanmean(self.q_nan) == 2 * self.ureg.m @helpers.requires_array_function_protocol() def test_average_numpy_func(self): helpers.assert_quantity_almost_equal( np.average(self.q, axis=0, weights=[1, 2]), [2.33333, 3.33333] * self.ureg.m, rtol=1e-5, ) @helpers.requires_array_function_protocol() def test_median_numpy_func(self): assert np.median(self.q) == 2.5 * self.ureg.m @helpers.requires_array_function_protocol() def test_nanmedian_numpy_func(self): assert np.nanmedian(self.q_nan) == 2 * self.ureg.m def test_var(self): assert self.q.var() == 1.25 * self.ureg.m**2 @helpers.requires_array_function_protocol() def test_var_numpy_func(self): assert np.var(self.q) == 1.25 * self.ureg.m**2 @helpers.requires_array_function_protocol() def test_nanvar_numpy_func(self): helpers.assert_quantity_almost_equal( np.nanvar(self.q_nan), 0.66667 * self.ureg.m**2, rtol=1e-5 ) def test_std(self): helpers.assert_quantity_almost_equal( self.q.std(), 1.11803 * self.ureg.m, rtol=1e-5 ) @helpers.requires_array_function_protocol() def test_std_numpy_func(self): helpers.assert_quantity_almost_equal( np.std(self.q), 1.11803 * self.ureg.m, rtol=1e-5 ) with pytest.raises(OffsetUnitCalculusError): np.std(self.q_temperature) def test_cumprod(self): with pytest.raises(DimensionalityError): self.q.cumprod() helpers.assert_quantity_equal((self.q / self.ureg.m).cumprod(), [1, 2, 6, 24]) @helpers.requires_array_function_protocol() def test_nanstd_numpy_func(self): helpers.assert_quantity_almost_equal( np.nanstd(self.q_nan), 0.81650 * self.ureg.m, rtol=1e-5 ) def test_conj(self): helpers.assert_quantity_equal((self.q * (1 + 1j)).conj(), self.q * (1 - 1j)) helpers.assert_quantity_equal( (self.q * (1 + 1j)).conjugate(), self.q * (1 - 1j) ) def test_getitem(self): with pytest.raises(IndexError): self.q.__getitem__((0, 10)) helpers.assert_quantity_equal(self.q[0], [1, 2] * self.ureg.m) assert self.q[1, 1] == 4 * self.ureg.m def test_setitem(self): with pytest.raises(TypeError): self.q[0, 0] = 1 with pytest.raises(DimensionalityError): self.q[0, 0] = 1 * self.ureg.J with pytest.raises(DimensionalityError): self.q[0] = 1 with pytest.raises(DimensionalityError): self.q[0] = np.ndarray([1, 2]) with pytest.raises(DimensionalityError): self.q[0] = 1 * self.ureg.J q = self.q.copy() q[0] = 1 * self.ureg.m helpers.assert_quantity_equal(q, [[1, 1], [3, 4]] * self.ureg.m) q = self.q.copy() q[...] = 1 * self.ureg.m helpers.assert_quantity_equal(q, [[1, 1], [1, 1]] * self.ureg.m) q = self.q.copy() q[:] = 1 * self.ureg.m helpers.assert_quantity_equal(q, [[1, 1], [1, 1]] * self.ureg.m) # check and see that dimensionless numbers work correctly q = [0, 1, 2, 3] * self.ureg.dimensionless q[0] = 1 helpers.assert_quantity_equal(q, np.asarray([1, 1, 2, 3])) q[0] = self.ureg.m / self.ureg.mm helpers.assert_quantity_equal(q, np.asarray([1000, 1, 2, 3])) q = [0.0, 1.0, 2.0, 3.0] * self.ureg.m / self.ureg.mm q[0] = 1.0 helpers.assert_quantity_equal(q, [0.001, 1, 2, 3] * self.ureg.m / self.ureg.mm) # Check that this properly masks the first item without warning q = self.ureg.Quantity( np.ma.array([0.0, 1.0, 2.0, 3.0], mask=[False, True, False, False]), "m" ) with warnings.catch_warnings(record=True) as w: q[0] = np.ma.masked # Check for no warnings assert not w assert q.mask[0] def test_setitem_mixed_masked(self): masked = np.ma.array( [ 1, 2, ], mask=[True, False], ) q = self.Q_(np.ones(shape=(2,)), "m") with pytest.raises(DimensionalityError): q[:] = masked masked_q = self.Q_(masked, "mm") q[:] = masked_q helpers.assert_quantity_equal(q, [1.0, 0.002] * self.ureg.m) def test_iterator(self): for q, v in zip(self.q.flatten(), [1, 2, 3, 4]): assert q == v * self.ureg.m def test_iterable(self): assert np.iterable(self.q) assert not np.iterable(1 * self.ureg.m) def test_reversible_op(self): """ """ x = self.q.magnitude u = self.Q_(np.ones(x.shape)) helpers.assert_quantity_equal(x / self.q, u * x / self.q) helpers.assert_quantity_equal(x * self.q, u * x * self.q) helpers.assert_quantity_equal(x + u, u + x) helpers.assert_quantity_equal(x - u, -(u - x)) def test_pickle(self, subtests): for protocol in range(pickle.HIGHEST_PROTOCOL + 1): with subtests.test(protocol): q1 = [10, 20] * self.ureg.m q2 = pickle.loads(pickle.dumps(q1, protocol)) self.assertNDArrayEqual(q1.magnitude, q2.magnitude) assert q1.units == q2.units def test_equal(self): x = self.q.magnitude u = self.Q_(np.ones(x.shape)) true = np.ones_like(x, dtype=np.bool_) false = np.zeros_like(x, dtype=np.bool_) helpers.assert_quantity_equal(u, u) helpers.assert_quantity_equal(u == u, u.magnitude == u.magnitude) helpers.assert_quantity_equal(u == 1, u.magnitude == 1) v = self.Q_(np.zeros(x.shape), "m") w = self.Q_(np.ones(x.shape), "m") self.assertNDArrayEqual(v == 1, false) self.assertNDArrayEqual( self.Q_(np.zeros_like(x), "m") == self.Q_(np.zeros_like(x), "s"), false, ) self.assertNDArrayEqual(v == v, true) self.assertNDArrayEqual(v == w, false) self.assertNDArrayEqual(v == w.to("mm"), false) self.assertNDArrayEqual(u == v, false) def test_shape(self): u = self.Q_(np.arange(12)) u.shape = 4, 3 assert u.magnitude.shape == (4, 3) def test_dtype(self): u = self.Q_(np.arange(12, dtype="uint32")) assert u.dtype == "uint32" @helpers.requires_array_function_protocol() def test_shape_numpy_func(self): assert np.shape(self.q) == (2, 2) @helpers.requires_array_function_protocol() def test_len_numpy_func(self): assert len(self.q) == 2 @helpers.requires_array_function_protocol() def test_ndim_numpy_func(self): assert np.ndim(self.q) == 2 @helpers.requires_array_function_protocol() def test_copy_numpy_func(self): q_copy = np.copy(self.q) helpers.assert_quantity_equal(self.q, q_copy) assert self.q is not q_copy @helpers.requires_array_function_protocol() def test_trim_zeros_numpy_func(self): q = [0, 4, 3, 0, 2, 2, 0, 0, 0] * self.ureg.m helpers.assert_quantity_equal(np.trim_zeros(q), [4, 3, 0, 2, 2] * self.ureg.m) @helpers.requires_array_function_protocol() def test_result_type_numpy_func(self): assert np.result_type(self.q) == np.dtype("int") @helpers.requires_array_function_protocol() def test_nan_to_num_numpy_func(self): helpers.assert_quantity_equal( np.nan_to_num(self.q_nan, nan=-999 * self.ureg.mm), [[1, 2], [3, -0.999]] * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_meshgrid_numpy_func(self): x = [1, 2] * self.ureg.m y = [0, 50, 100] * self.ureg.mm xx, yy = np.meshgrid(x, y) helpers.assert_quantity_equal(xx, [[1, 2], [1, 2], [1, 2]] * self.ureg.m) helpers.assert_quantity_equal(yy, [[0, 0], [50, 50], [100, 100]] * self.ureg.mm) @helpers.requires_array_function_protocol() def test_isclose_numpy_func(self): q2 = [[1000.05, 2000], [3000.00007, 4001]] * self.ureg.mm self.assertNDArrayEqual( np.isclose(self.q, q2), np.array([[False, True], [True, False]]) ) self.assertNDArrayEqual( np.isclose(self.q, q2, atol=1e-5 * self.ureg.mm, rtol=1e-7), np.array([[False, True], [True, False]]), ) self.assertNDArrayEqual( np.isclose(self.q, q2, atol=1e-5, rtol=1e-7), np.array([[False, True], [True, False]]), ) @helpers.requires_array_function_protocol() def test_interp_numpy_func(self): x = [1, 4] * self.ureg.m xp = np.linspace(0, 3, 5) * self.ureg.m fp = self.Q_([0, 5, 10, 15, 20], self.ureg.degC) helpers.assert_quantity_almost_equal( np.interp(x, xp, fp), self.Q_([6.66667, 20.0], self.ureg.degC), rtol=1e-5 ) x_ = np.array([1, 4]) xp_ = np.linspace(0, 3, 5) fp_ = [0, 5, 10, 15, 20] helpers.assert_quantity_almost_equal( np.interp(x_, xp_, fp), self.Q_([6.6667, 20.0], self.ureg.degC), rtol=1e-5 ) helpers.assert_quantity_almost_equal( np.interp(x, xp, fp_), [6.6667, 20.0], rtol=1e-5 ) def test_comparisons(self): self.assertNDArrayEqual( self.q > 2 * self.ureg.m, np.array([[False, False], [True, True]]) ) self.assertNDArrayEqual( self.q < 2 * self.ureg.m, np.array([[True, False], [False, False]]) ) @helpers.requires_array_function_protocol() def test_where(self): helpers.assert_quantity_equal( np.where(self.q >= 2 * self.ureg.m, self.q, 20 * self.ureg.m), [[20, 2], [3, 4]] * self.ureg.m, ) helpers.assert_quantity_equal( np.where(self.q >= 2 * self.ureg.m, self.q, 0), [[0, 2], [3, 4]] * self.ureg.m, ) helpers.assert_quantity_equal( np.where(self.q >= 2 * self.ureg.m, self.q, np.nan), [[np.nan, 2], [3, 4]] * self.ureg.m, ) helpers.assert_quantity_equal( np.where(self.q >= 3 * self.ureg.m, 0, self.q), [[1, 2], [0, 0]] * self.ureg.m, ) helpers.assert_quantity_equal( np.where(self.q >= 3 * self.ureg.m, np.nan, self.q), [[1, 2], [np.nan, np.nan]] * self.ureg.m, ) helpers.assert_quantity_equal( np.where(self.q >= 2 * self.ureg.m, self.q, np.array(np.nan)), [[np.nan, 2], [3, 4]] * self.ureg.m, ) helpers.assert_quantity_equal( np.where(self.q >= 3 * self.ureg.m, np.array(np.nan), self.q), [[1, 2], [np.nan, np.nan]] * self.ureg.m, ) with pytest.raises(DimensionalityError): np.where( self.q < 2 * self.ureg.m, self.q, 0 * self.ureg.J, ) helpers.assert_quantity_equal( np.where([-1, 0, 1] * self.ureg.m, [1, 2, 1] * self.ureg.s, np.nan), [1, np.nan, 1] * self.ureg.s, ) with pytest.raises( ValueError, match=".*Boolean value of Quantity with offset unit is ambiguous", ): np.where( self.ureg.Quantity([-1, 0, 1], "degC"), [1, 2, 1] * self.ureg.s, np.nan ) @helpers.requires_array_function_protocol() def test_fabs(self): helpers.assert_quantity_equal( np.fabs(self.q - 2 * self.ureg.m), self.Q_([[1, 0], [1, 2]], "m") ) @helpers.requires_array_function_protocol() def test_isin(self): self.assertNDArrayEqual( np.isin(self.q, self.Q_([0, 2, 4], "m")), np.array([[False, True], [False, True]]), ) self.assertNDArrayEqual( np.isin(self.q, self.Q_([0, 2, 4], "J")), np.array([[False, False], [False, False]]), ) self.assertNDArrayEqual( np.isin(self.q, [self.Q_(2, "m"), self.Q_(4, "J")]), np.array([[False, True], [False, False]]), ) self.assertNDArrayEqual( np.isin(self.q, self.q.m), np.array([[False, False], [False, False]]) ) self.assertNDArrayEqual( np.isin(self.q / self.ureg.cm, [1, 3]), np.array([[True, False], [True, False]]), ) with pytest.raises(ValueError): np.isin(self.q.m, self.q) @helpers.requires_array_function_protocol() def test_percentile(self): helpers.assert_quantity_equal(np.percentile(self.q, 25), self.Q_(1.75, "m")) @helpers.requires_array_function_protocol() def test_nanpercentile(self): helpers.assert_quantity_equal( np.nanpercentile(self.q_nan, 25), self.Q_(1.5, "m") ) @helpers.requires_array_function_protocol() def test_quantile(self): helpers.assert_quantity_equal(np.quantile(self.q, 0.25), self.Q_(1.75, "m")) @helpers.requires_array_function_protocol() def test_nanquantile(self): helpers.assert_quantity_equal( np.nanquantile(self.q_nan, 0.25), self.Q_(1.5, "m") ) @helpers.requires_array_function_protocol() def test_copyto(self): a = self.q.m q = copy.copy(self.q) np.copyto(q, 2 * q, where=[True, False]) helpers.assert_quantity_equal(q, self.Q_([[2, 2], [6, 4]], "m")) np.copyto(q, 0, where=[[False, False], [True, False]]) helpers.assert_quantity_equal(q, self.Q_([[2, 2], [0, 4]], "m")) np.copyto(a, q) self.assertNDArrayEqual(a, np.array([[2, 2], [0, 4]])) @helpers.requires_array_function_protocol() def test_tile(self): helpers.assert_quantity_equal( np.tile(self.q, 2), np.array([[1, 2, 1, 2], [3, 4, 3, 4]]) * self.ureg.m ) @helpers.requires_numpy_at_least("1.20") @helpers.requires_array_function_protocol() def test_sliding_window_view(self): q = self.Q_([[1, 2, 2, 1], [2, 1, 1, 2], [1, 2, 2, 1]], "m") actual = np.lib.stride_tricks.sliding_window_view(q, window_shape=(3, 3)) expected = self.Q_( [[[[1, 2, 2], [2, 1, 1], [1, 2, 2]], [[2, 2, 1], [1, 1, 2], [2, 2, 1]]]], "m", ) helpers.assert_quantity_equal(actual, expected) @helpers.requires_array_function_protocol() def test_rot90(self): helpers.assert_quantity_equal( np.rot90(self.q), np.array([[2, 4], [1, 3]]) * self.ureg.m ) @helpers.requires_array_function_protocol() def test_insert(self): helpers.assert_quantity_equal( np.insert(self.q, 1, 0 * self.ureg.m, axis=1), np.array([[1, 0, 2], [3, 0, 4]]) * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_delete(self): q = self.Q_(np.array([[1, 2, 3, 4], [5, 6, 7, 8], [9, 10, 11, 12]]), "m") helpers.assert_quantity_equal( np.delete(q, 1, axis=0), np.array([[1, 2, 3, 4], [9, 10, 11, 12]]) * self.ureg.m, ) helpers.assert_quantity_equal( np.delete(q, np.s_[::2], 1), np.array([[2, 4], [6, 8], [10, 12]]) * self.ureg.m, ) helpers.assert_quantity_equal( np.delete(q, [1, 3, 5], None), np.array([1, 3, 5, 7, 8, 9, 10, 11, 12]) * self.ureg.m, ) def test_ndarray_downcast(self): with pytest.warns(UnitStrippedWarning): np.asarray(self.q) def test_ndarray_downcast_with_dtype(self): with pytest.warns(UnitStrippedWarning): qarr = np.asarray(self.q, dtype=np.float64) assert qarr.dtype == np.float64 def test_array_protocol_unavailable(self): for attr in ("__array_struct__", "__array_interface__"): with pytest.raises(AttributeError): getattr(self.q, attr) @helpers.requires_array_function_protocol() def test_resize(self): helpers.assert_quantity_equal( np.resize(self.q, (2, 4)), [[1, 2, 3, 4], [1, 2, 3, 4]] * self.ureg.m ) @helpers.requires_array_function_protocol() def test_pad(self): # Tests reproduced with modification from NumPy documentation a = [1, 2, 3, 4, 5] * self.ureg.m b = self.Q_([4.0, 6.0, 8.0, 9.0, -3.0], "degC") helpers.assert_quantity_equal( np.pad(a, (2, 3), "constant"), [0, 0, 1, 2, 3, 4, 5, 0, 0, 0] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "constant", constant_values=(0, 600 * self.ureg.cm)), [0, 0, 1, 2, 3, 4, 5, 6, 6, 6] * self.ureg.m, ) helpers.assert_quantity_equal( np.pad( b, (2, 1), "constant", constant_values=(np.nan, self.Q_(10, "degC")) ), self.Q_([np.nan, np.nan, 4, 6, 8, 9, -3, 10], "degC"), ) with pytest.raises(DimensionalityError): np.pad(a, (2, 3), "constant", constant_values=4) helpers.assert_quantity_equal( np.pad(a, (2, 3), "edge"), [1, 1, 1, 2, 3, 4, 5, 5, 5, 5] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "linear_ramp"), [0, 0, 1, 2, 3, 4, 5, 3, 1, 0] * self.ureg.m, ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "linear_ramp", end_values=(5, -4) * self.ureg.m), [5, 3, 1, 2, 3, 4, 5, 2, -1, -4] * self.ureg.m, ) helpers.assert_quantity_equal( np.pad(a, (2,), "maximum"), [5, 5, 1, 2, 3, 4, 5, 5, 5] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(a, (2,), "mean"), [3, 3, 1, 2, 3, 4, 5, 3, 3] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(a, (2,), "median"), [3, 3, 1, 2, 3, 4, 5, 3, 3] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(self.q, ((3, 2), (2, 3)), "minimum"), [ [1, 1, 1, 2, 1, 1, 1], [1, 1, 1, 2, 1, 1, 1], [1, 1, 1, 2, 1, 1, 1], [1, 1, 1, 2, 1, 1, 1], [3, 3, 3, 4, 3, 3, 3], [1, 1, 1, 2, 1, 1, 1], [1, 1, 1, 2, 1, 1, 1], ] * self.ureg.m, ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "reflect"), [3, 2, 1, 2, 3, 4, 5, 4, 3, 2] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "reflect", reflect_type="odd"), [-1, 0, 1, 2, 3, 4, 5, 6, 7, 8] * self.ureg.m, ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "symmetric"), [2, 1, 1, 2, 3, 4, 5, 5, 4, 3] * self.ureg.m ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "symmetric", reflect_type="odd"), [0, 1, 1, 2, 3, 4, 5, 5, 6, 7] * self.ureg.m, ) helpers.assert_quantity_equal( np.pad(a, (2, 3), "wrap"), [4, 5, 1, 2, 3, 4, 5, 1, 2, 3] * self.ureg.m ) def pad_with(vector, pad_width, iaxis, kwargs): pad_value = kwargs.get("padder", 10) vector[: pad_width[0]] = pad_value vector[-pad_width[1] :] = pad_value b = self.Q_(np.arange(6).reshape((2, 3)), "degC") helpers.assert_quantity_equal( np.pad(b, 2, pad_with), self.Q_( [ [10, 10, 10, 10, 10, 10, 10], [10, 10, 10, 10, 10, 10, 10], [10, 10, 0, 1, 2, 10, 10], [10, 10, 3, 4, 5, 10, 10], [10, 10, 10, 10, 10, 10, 10], [10, 10, 10, 10, 10, 10, 10], ], "degC", ), ) helpers.assert_quantity_equal( np.pad(b, 2, pad_with, padder=100), self.Q_( [ [100, 100, 100, 100, 100, 100, 100], [100, 100, 100, 100, 100, 100, 100], [100, 100, 0, 1, 2, 100, 100], [100, 100, 3, 4, 5, 100, 100], [100, 100, 100, 100, 100, 100, 100], [100, 100, 100, 100, 100, 100, 100], ], "degC", ), ) # Note: Does not support Quantity pad_with vectorized callable use @helpers.requires_array_function_protocol() def test_allclose(self): assert np.allclose([1e10, 1e-8] * self.ureg.m, [1.00001e10, 1e-9] * self.ureg.m) assert np.allclose( [1e10, 1e-8] * self.ureg.m, [1.00001e13, 1e-6] * self.ureg.mm ) assert not np.allclose( [1e10, 1e-8] * self.ureg.m, [1.00001e10, 1e-9] * self.ureg.mm ) assert np.allclose( [1e10, 1e-8] * self.ureg.m, [1.00001e10, 1e-9] * self.ureg.m, atol=1e-8 * self.ureg.m, ) assert not np.allclose([1.0, np.nan] * self.ureg.m, [1.0, np.nan] * self.ureg.m) assert np.allclose( [1.0, np.nan] * self.ureg.m, [1.0, np.nan] * self.ureg.m, equal_nan=True ) assert np.allclose( [1e10, 1e-8] * self.ureg.m, [1.00001e10, 1e-9] * self.ureg.m, atol=1e-8 ) with pytest.raises(DimensionalityError): assert np.allclose( [1e10, 1e-8] * self.ureg.m, [1.00001e10, 1e-9] * self.ureg.m, atol=1e-8 * self.ureg.s, ) @helpers.requires_array_function_protocol() def test_intersect1d(self): helpers.assert_quantity_equal( np.intersect1d([1, 3, 4, 3] * self.ureg.m, [3, 1, 2, 1] * self.ureg.m), [1, 3] * self.ureg.m, ) @helpers.requires_array_function_protocol() def test_linalg_norm(self): q = np.array([[3, 5, 8], [4, 12, 15]]) * self.ureg.m expected = [5, 13, 17] * self.ureg.m helpers.assert_quantity_equal(np.linalg.norm(q, axis=0), expected) @pytest.mark.skip class TestBitTwiddlingUfuncs(TestUFuncs): """Universal functions (ufuncs) > Bittwiddling functions http://docs.scipy.org/doc/numpy/reference/ufuncs.html#bittwiddlingfunctions bitwise_and(x1, x2[, out]) Compute the bitwise AND of two arrays elementwise. bitwise_or(x1, x2[, out]) Compute the bitwise OR of two arrays elementwise. bitwise_xor(x1, x2[, out]) Compute the bitwise XOR of two arrays elementwise. invert(x[, out]) Compute bitwise inversion, or bitwise NOT, elementwise. left_shift(x1, x2[, out]) Shift the bits of an integer to the left. right_shift(x1, x2[, out]) Shift the bits of an integer to the right. Parameters ---------- Returns ------- """ @property def qless(self): return np.asarray([1, 2, 3, 4], dtype=np.uint8) * self.ureg.dimensionless @property def qs(self): return 8 * self.ureg.J @property def q1(self): return np.asarray([1, 2, 3, 4], dtype=np.uint8) * self.ureg.J @property def q2(self): return 2 * self.q1 @property def qm(self): return np.asarray([1, 2, 3, 4], dtype=np.uint8) * self.ureg.m def test_bitwise_and(self): self._test2(np.bitwise_and, self.q1, (self.q2, self.qs), (self.qm,), "same") def test_bitwise_or(self): self._test2( np.bitwise_or, self.q1, (self.q1, self.q2, self.qs), (self.qm,), "same" ) def test_bitwise_xor(self): self._test2( np.bitwise_xor, self.q1, (self.q1, self.q2, self.qs), (self.qm,), "same" ) def test_invert(self): self._test1(np.invert, (self.q1, self.q2, self.qs), (), "same") def test_left_shift(self): self._test2( np.left_shift, self.q1, (self.qless, 2), (self.q1, self.q2, self.qs), "same" ) def test_right_shift(self): self._test2( np.right_shift, self.q1, (self.qless, 2), (self.q1, self.q2, self.qs), "same", )