def test_kwargs_size_or_shape(): a = normal(loc=10.0, scale=0.1, shape=(10, 10), chunks=(5, 5)) b = normal(loc=10.0, scale=0.1, size=(10, 10), chunks=(5, 5)) assert isinstance(a, Array) assert isinstance(b, Array) assert a.chunks == b.chunks assert np.array(a).shape == np.array(b).shape
def test_kwargs_size_or_shape(): a = normal(loc=10.0, scale=0.1, shape=(10, 10), blockshape=(5, 5)) b = normal(loc=10.0, scale=0.1, size=(10, 10), blockshape=(5, 5)) assert isinstance(a, Array) assert isinstance(b, Array) assert a.blockdims == b.blockdims assert np.array(a).shape == np.array(b).shape
def test_add_gaussian_noise(self): s = self.signal s.change_dtype("float64") kwargs = {} if s._lazy: data = s.data.compute() from dask.array.random import seed, normal kwargs["chunks"] = s.data.chunks else: data = s.data.copy() from numpy.random import seed, normal seed(1) s.add_gaussian_noise(std=1.0) seed(1) if s._lazy: s.compute() np.testing.assert_array_almost_equal( s.data - data, normal(scale=1.0, size=data.shape, **kwargs))
def test_can_make_really_big_random_array(): x = normal(10, 1, (1000000, 1000000), chunks=(100000, 100000))
def test_kwargs(): a = normal(loc=10.0, scale=0.1, size=(10, 10), chunks=(5, 5)) assert isinstance(a, Array) x = np.array(a) assert 8 < x.mean() < 12
def test_can_make_really_big_random_array(): normal(10, 1, (1000000, 1000000), chunks=(100000, 100000))
def test_kwargs(): a = normal(loc=10.0, scale=0.1, size=(10, 10), blockshape=(5, 5)) assert isinstance(a, Array) x = into(np.ndarray, a) assert 8 < x.mean() < 12