def run(self):
g = InfoCluster(affinity='nearest_neighbors', n_neighbors=8)
g.fit(self.pos_list)
self.partition_list = g.partition_list
self.critical_values = g.critical_values
self.get_category = g.get_category
self.g = g
def run(self):
g = InfoCluster(gamma=self._gamma)
g.fit(self.pos_list)
self.partition_list = g.partition_list
self.critical_values = g.critical_values
self.get_category = g.get_category
self.g = g
def predict_with_same_ancestor(self, tree_node, node_index_i, node_index_j,
weight_added):
# rerun the info-clustering algorithm, if tree_node has many children, the speed is very slow
child_node_list = tree_node.get_children()
n_nodes = len(child_node_list)
affinity_matrix = np.zeros([n_nodes, n_nodes])
for ii in range(n_nodes):
for jj in range(ii + 1, n_nodes):
affinity_matrix[ii,
jj] = self.get_weight(child_node_list[ii],
child_node_list[jj])
node_index_ii = -1
node_index_jj = -1
for ii in range(n_nodes):
if (str(node_index_i) == child_node_list[ii].name):
node_index_ii = ii
if (str(node_index_j) == child_node_list[ii].name):
node_index_jj = ii
if (node_index_ii < node_index_jj):
affinity_matrix[node_index_ii, node_index_jj] = weight_added
else:
affinity_matrix[node_index_jj, node_index_ii] = weight_added
new_ic = InfoCluster(affinity='precomputed')
new_ic.fit(affinity_matrix)
is_solution_trivial = len(new_ic.critical_values) == 1
return is_solution_trivial
def test_cv_value(self):
g = nx.Graph() # undirected graph
g.add_edge(0, 1, weight=1)
g.add_edge(1, 2, weight=1)
g.add_edge(0, 2, weight=5)
ic = InfoCluster(
affinity='precomputed') # use precomputed graph structure
ic.fit(g)
self.assertAlmostEqual(ic.tree.cv, 2.0)
self.assertAlmostEqual(ic.tree.get_children()[0].cv, 5.0)
def test_workflow(self):
g = nx.Graph() # undirected graph
g.add_edge(0, 1, weight=1)
g.add_edge(1, 2, weight=1)
g.add_edge(0, 2, weight=5)
info_cluster_instance = InfoCluster(affinity='precomputed')
info_cluster_instance.fit(g)
info_cluster_instance.print_hierarchical_tree()
self.assertEqual(info_cluster_instance.get_tree_depth(), 2)
def test_4_point(self):
g = nx.DiGraph()
g.add_edge(0, 2, weight=2)
g.add_edge(0, 1, weight=5)
g.add_edge(0, 3, weight=1)
g.add_edge(1, 2, weight=2)
g.add_edge(1, 3, weight=1)
info_cluster_instance = InfoCluster(affinity='precomputed')
info_cluster_instance.fit(g)
info_cluster_instance.print_hierarchical_tree()
self.assertEqual(info_cluster_instance.get_tree_depth(), 3)
def test_10_point(self):
point_list = np.array([[0.8, 1.3], [0.9, 1.2], [0.9, -0.5],
[0.7, -0.6], [1.0, -0.6], [-1.3, -0.5],
[-1.0, -0.6], [-1.1, -0.8], [-2.7, -0.8],
[-2.6, -0.9]])
info_cluster_instance = InfoCluster(gamma=1, affinity="rbf")
info_cluster_instance.fit(point_list)
info_cluster_instance.print_hierarchical_tree()
# print(info_cluster_instance.critical_values)
print(info_cluster_instance.partition_list)