def test_single_point_on_edge(self):
"""Test behaviour for a non-zero grid cell on the edge.
Note that this behaviour is 'wrong' and is a result of
scipy.ndimage.correlate 'nearest' mode. We need to fix
this in the future.
"""
cube = set_up_cube(
zero_point_indices=[(0, 0, 7, 0)]) # On the (y) edge.
expected = np.ones_like(cube.data)
expected_centroid = np.array([
[0.92, 0.96, 0.992],
[0.848, 0.912, 0.968],
[0.824, 0.896, 0.96],
[0.848, 0.912, 0.968],
[0.92, 0.96, 0.992],
])
for index, slice_ in enumerate(expected_centroid):
expected[0][0][5 + index][0:3] = slice_
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_lat_long(self):
"""Test behaviour for a single grid cell on lat long grid."""
cube = set_up_cube_lat_long()
msg = "Invalid grid: projection_x/y coords required"
ranges = (3, 3)
with self.assertRaisesRegexp(ValueError, msg):
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges)
def test_basic(self):
"""Test that the plugin returns an iris.cube.Cube."""
cube = set_up_cube(
zero_point_indices=((0, 0, 2, 2),), num_time_points=1,
num_grid_points=5)
ranges = (2, 2)
result = (
CircularNeighbourhood(
unweighted_mode=True).apply_circular_kernel(cube, ranges))
self.assertIsInstance(result, Cube)
def test_single_point_range_1(self):
"""Test behaviour with a non-zero point with unit range."""
cube = set_up_cube()
expected = np.ones_like(cube.data)
expected[0][0][7][7] = 0.0
ranges = (1, 1)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point(self):
"""Test behaviour for a single non-zero grid cell."""
cube = set_up_cube()
expected = np.ones_like(cube.data)
for index, slice_ in enumerate(SINGLE_POINT_RANGE_3_CENTROID):
expected[0][0][5 + index][5:10] = slice_
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_almost_corner(self):
"""Test behaviour for a non-zero grid cell quite near a corner."""
cube = set_up_cube(
zero_point_indices=[(0, 0, 2, 2)]) # Just within corner range.
expected = np.ones_like(cube.data)
for index, slice_ in enumerate(SINGLE_POINT_RANGE_3_CENTROID):
expected[0][0][index][0:5] = slice_
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_adjacent_edge(self):
"""Test behaviour for a single non-zero grid cell near the edge."""
cube = set_up_cube(
zero_point_indices=[(0, 0, 7, 1)]) # Range 3 goes over the edge.
expected = np.ones_like(cube.data)
for index, slice_ in enumerate(SINGLE_POINT_RANGE_3_CENTROID):
expected[0][0][5 + index][0:4] = slice_[1:]
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_range_5(self):
"""Test behaviour with a non-zero point with a large range."""
cube = set_up_cube()
expected = np.ones_like(cube.data)
for time_index in range(len(expected)):
for index, slice_ in enumerate(SINGLE_POINT_RANGE_5_CENTROID):
expected[0][time_index][3 + index][3:12] = slice_
ranges = (5, 5)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_basic(self):
"""Test that a cube with correct data is produced by the run method"""
data = np.array([[0.992, 0.968, 0.96, 0.968, 0.992],
[0.968, 0.944, 0.936, 0.944, 0.968],
[0.96, 0.936, 0.928, 0.936, 0.96],
[0.968, 0.944, 0.936, 0.944, 0.968],
[0.992, 0.968, 0.96, 0.968, 0.992]])
cube = set_up_cube(
zero_point_indices=((0, 0, 2, 2),), num_grid_points=5)[0, 0]
result = CircularNeighbourhood().run(cube, self.RADIUS)
self.assertIsInstance(cube, Cube)
self.assertArrayAlmostEqual(result.data, data)
def test_multi_point_multitimes(self):
"""Test behaviour for points over multiple times."""
cube = set_up_cube(
zero_point_indices=[(0, 0, 10, 10), (0, 1, 7, 7)],
num_time_points=2
)
expected = np.ones_like(cube.data)
for index, slice_ in enumerate(SINGLE_POINT_RANGE_3_CENTROID):
expected[0][0][8 + index][8:13] = slice_
for index, slice_ in enumerate(SINGLE_POINT_RANGE_3_CENTROID):
expected[0][1][5 + index][5:10] = slice_
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_flat(self):
"""Test behaviour for a single non-zero grid cell, flat weighting.
Note that this gives one more grid cell range than weighted! As the
affected area is one grid cell more in each direction, an equivalent
range of 2 was chosen for this test.
"""
cube = set_up_cube()
expected = np.ones_like(cube.data)
for index, slice_ in enumerate(SINGLE_POINT_RANGE_2_CENTROID_FLAT):
expected[0][0][5 + index][5:10] = slice_
ranges = (2, 2)
result = (
CircularNeighbourhood(
unweighted_mode=True).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_point_pair(self):
"""Test behaviour for two nearby non-zero grid cells."""
cube = set_up_cube(
zero_point_indices=[(0, 0, 7, 6), (0, 0, 7, 8)])
expected_snippet = np.array([
[0.992, 0.968, 0.952, 0.936, 0.952, 0.968, 0.992],
[0.968, 0.944, 0.904, 0.888, 0.904, 0.944, 0.968],
[0.96, 0.936, 0.888, 0.872, 0.888, 0.936, 0.96],
[0.968, 0.944, 0.904, 0.888, 0.904, 0.944, 0.968],
[0.992, 0.968, 0.952, 0.936, 0.952, 0.968, 0.992]
])
expected = np.ones_like(cube.data)
for index, slice_ in enumerate(expected_snippet):
expected[0][0][5 + index][4:11] = slice_
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_range_5_small_domain(self):
"""Test behaviour - non-zero point, small domain, large range.
This exhibits the undesirable edge reflection behaviour.
"""
cube = set_up_cube(
zero_point_indices=((0, 0, 1, 1),), num_grid_points=4)
expected = np.array([
[[[0.97636177, 0.97533402, 0.97636177, 0.97944502],
[0.97533402, 0.97430627, 0.97533402, 0.97841727],
[0.97636177, 0.97533402, 0.97636177, 0.97944502],
[0.97944502, 0.97841727, 0.97944502, 0.98252826]]]
])
ranges = (5, 5)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_single_point_masked_other_point(self):
"""Test behaviour with a non-zero point next to a masked point.
The behaviour here is not right, as the mask is ignored.
"""
cube = set_up_cube()
expected = np.ones_like(cube.data)
mask = np.zeros_like(cube.data)
mask[0][0][6][7] = 1
cube.data = np.ma.masked_array(cube.data, mask=mask)
for time_index in range(len(expected)):
for index, slice_ in enumerate(SINGLE_POINT_RANGE_3_CENTROID):
expected[0][time_index][5 + index][5:10] = slice_
ranges = (3, 3)
result = (
CircularNeighbourhood(
unweighted_mode=False).apply_circular_kernel(cube, ranges))
self.assertArrayAlmostEqual(result.data, expected)
def test_basic(self):
"""Test that the __repr__ returns the expected string."""
result = str(CircularNeighbourhood())
msg = '<CircularNeighbourhood: unweighted_mode: False>'
self.assertEqual(str(result), msg)