def test_extreme_example(self):
# 0, 1, 2, 3
# 4, 5, 6, 7
# 8, 9,10,11
#12,13,14,15
self.label = numpy.array(range(16)).reshape((4,4))
(adj_mat, adj_dic) = selective_search._calc_adjacency_matrix(self.label, 16)
assert numpy.array_equal(adj_mat.transpose(), adj_mat)
assert set(numpy.flatnonzero(adj_mat[ 0])) == { 0, 1, 4}
assert set(numpy.flatnonzero(adj_mat[ 1])) == { 0, 1, 2, 5}
assert set(numpy.flatnonzero(adj_mat[ 2])) == { 1, 2, 3, 6}
assert set(numpy.flatnonzero(adj_mat[ 3])) == { 2, 3, 7}
assert set(numpy.flatnonzero(adj_mat[ 4])) == { 0, 4, 5, 8}
assert set(numpy.flatnonzero(adj_mat[ 5])) == { 1, 4, 5, 6, 9}
assert set(numpy.flatnonzero(adj_mat[ 6])) == { 2, 5, 6, 7, 10}
assert set(numpy.flatnonzero(adj_mat[ 7])) == { 3, 6, 7, 11}
assert set(numpy.flatnonzero(adj_mat[ 8])) == { 4, 8, 9, 12}
assert set(numpy.flatnonzero(adj_mat[ 9])) == { 5, 8, 9, 10, 13}
assert set(numpy.flatnonzero(adj_mat[10])) == { 6, 9, 10, 11, 14}
assert set(numpy.flatnonzero(adj_mat[11])) == { 7, 10, 11, 15}
assert set(numpy.flatnonzero(adj_mat[12])) == { 8, 12, 13}
assert set(numpy.flatnonzero(adj_mat[13])) == { 9, 12, 13, 14}
assert set(numpy.flatnonzero(adj_mat[14])) == {10, 13, 14, 15}
assert set(numpy.flatnonzero(adj_mat[15])) == {11, 14, 15}
for (i, adj_labels) in adj_dic.items():
assert set(numpy.flatnonzero(adj_mat[i])) - {i} == adj_labels
def test_only_1_segment(self):
# 0, 0, 0, 0
# 0, 0, 0, 0
# 0, 0, 0, 0
# 0, 0, 0, 0
(adj_mat, adj_dic) = selective_search._calc_adjacency_matrix(self.label, 1)
assert type(adj_mat) == numpy.ndarray
assert adj_mat.shape == (1, 1) and adj_mat.dtype == bool
assert adj_mat[0, 0] == True
assert adj_dic[0] == set()
def test_fully_adjacent(self):
# 1, 1, 0, 0
# 1, 1, 0, 0
# 1, 1, 0, 0
# 1, 1, 0, 0
self.label[:2, :] = 1
expected_mat = numpy.array([[True, True],\
[True, True]])
(adj_mat, adj_dic) = selective_search._calc_adjacency_matrix(self.label, 2)
assert adj_mat.shape == (2, 2) and adj_mat.dtype == bool
assert numpy.array_equal(adj_mat, expected_mat)
assert adj_dic[0] == {1}
assert adj_dic[1] == {0}
def test_edge_case_horizontal(self):
# 0, 0, 0, 0
# 0, 0, 0, 0
# 0, 0, 0, 0
# 1, 1, 2, 2
self.label[-1:, :2] = 1
self.label[-1:, 2:] = 2
expected_mat = numpy.array([[True, True, True],\
[True, True, True],\
[True, True, True]])
(adj_mat, adj_dic) = selective_search._calc_adjacency_matrix(self.label, 3)
assert numpy.array_equal(expected_mat, adj_mat)
assert adj_dic[0] == {1, 2}
assert adj_dic[1] == {0, 2}
assert adj_dic[2] == {0, 1}
def test_edge_case_vertical(self):
# 0, 0, 0, 1
# 0, 0, 0, 1
# 0, 0, 0, 2
# 0, 0, 0, 2
self.label[:2, -1:] = 1
self.label[2:, -1:] = 2
expected_mat = numpy.array([[True, True, True],\
[True, True, True],\
[True, True, True]])
(adj_mat, adj_dic) = selective_search._calc_adjacency_matrix(self.label, 3)
assert numpy.array_equal(expected_mat, adj_mat)
assert adj_dic[0] == {1, 2}
assert adj_dic[1] == {0, 2}
assert adj_dic[2] == {0, 1}
def test_partially_adjacent(self):
# 0, 0, 1, 1
# 0, 0, 1, 1
# 2, 2, 3, 3
# 2, 2, 3, 3
self.label[:2, :2] = 0
self.label[:2, 2:] = 1
self.label[2:, :2] = 2
self.label[2:, 2:] = 3
expected_mat = numpy.array([[True, True, True, False],\
[True, True, False, True],\
[True, False, True, True],\
[False, True, True, True]])
(adj_mat, adj_dic) = selective_search._calc_adjacency_matrix(self.label, 4)
assert numpy.diag(adj_mat).all()
assert numpy.array_equal(adj_mat.transpose(), adj_mat)
assert numpy.array_equal(adj_mat, expected_mat)
assert adj_dic[0] == {1, 2}
assert adj_dic[1] == {0, 3}
assert adj_dic[2] == {0, 3}
assert adj_dic[3] == {1, 2}
