🐛 fix some open issues on algorithm implementation/inputs specification

GiulioRossetti · Jun 8, 2023 · eec7e5a · eec7e5a
1 parent 7531530
commit eec7e5a
Show file tree

Hide file tree

Showing 4 changed files with 10 additions and 5 deletions.
diff --git a/cdlib/algorithms/attribute_clustering.py b/cdlib/algorithms/attribute_clustering.py
@@ -64,14 +64,18 @@ def eva(
     g = convert_graph_formats(g_original, nx.Graph)
     nx.set_node_attributes(g, labels)
 
+    mapping = dict(zip(g, range(0, len(g))))
+    rev_map = {v: k for k, v in mapping.items()}
+    relabel_g = nx.relabel_nodes(g, mapping)
+
     coms, coms_labels = Eva.eva_best_partition(
-        g, weight=weight, resolution=resolution, alpha=alpha
+        relabel_g, weight=weight, resolution=resolution, alpha=alpha
     )
 
     # Reshaping the results
     coms_to_node = defaultdict(list)
     for n, c in coms.items():
-        coms_to_node[c].append(n)
+        coms_to_node[c].append(rev_map[n])
 
     coms_eva = [list(c) for c in coms_to_node.values()]
     return AttrNodeClustering(

diff --git a/cdlib/algorithms/bipartite_clustering.py b/cdlib/algorithms/bipartite_clustering.py
@@ -50,7 +50,7 @@ def bimlpa(g_original: object, theta: float = 0.3, lambd: int = 7) -> BiNodeClus
     Yes        No       No       Yes
     ========== ======== ======== =========
 
-    :param g_original: a networkx/igraph object
+    :param g_original: a networkx/igraph object (instance of igraph.Graph or nx.Graph).
     :param theta: Label weights threshold. Default 0.3.
     :param lambd: The max number of labels. Default 7.
     :return: BiNodeClustering object

diff --git a/cdlib/algorithms/internal/em.py b/cdlib/algorithms/internal/em.py
@@ -28,7 +28,7 @@ def e_step(self, q):
                 q[i].append(x)
                 norm += x
             for g in range(self.k):
-                q[i][g] /= norm
+                q[i][g] /= (norm + 1e-10)  # tolerance adjustment
 
     def m_step(self, q):
         for g in range(self.k):
@@ -47,7 +47,7 @@ def m_step(self, q):
 
     def execute(self):
         # initial parameters
-        X = [1.0 + random.random() for i in range(self.k)]
+        X = [1.0 + random.random() for _ in range(self.k)]
         norm = sum(X)
         self.pi = [x / norm for x in X]
 

diff --git a/cdlib/test/test_attributeclustering.py b/cdlib/test/test_attributeclustering.py
@@ -10,6 +10,7 @@ def test_eva(self):
         l1 = ["one", "two", "three", "four"]
         l2 = ["A", "B", "C"]
         g = nx.barabasi_albert_graph(100, 5)
+
         labels = dict()
 
         for node in g.nodes():