FIX: Fixed edge ratio metric

RienkF · May 31, 2022 · 636aa02 · 636aa02
1 parent 95f26e0
commit 636aa02
Show file tree

Hide file tree

Showing 3 changed files with 20 additions and 18 deletions.
diff --git a/algorithm/edge_ratio.py b/algorithm/edge_ratio.py
@@ -6,19 +6,17 @@
 
 # TODO: Type partition
 def global_edge_ratio(G: networkx.DiGraph, partitions):
-    score_sum = 0
-    for partition in partitions:
-        internal_edges = len(G.subgraph(partition).edges)
-        out_edges = len(list(edge_boundary(G, partition, G.nodes - partition)))
-        denom = internal_edges + out_edges
-        if denom == 0:
-            continue
-        score_sum += internal_edges / denom
-    return score_sum
+    return sum(map(lambda partition: edge_boundary_ratio(G, partition), partitions))
 
 
 def local_edge_ratio(
-    G: networkx.DiGraph, u, neighbour, node_to_community, inner_partition, partition, original_graph: networkx.DiGraph
+    G: networkx.DiGraph,
+    u,
+    neighbour,
+    node_to_community,
+    inner_partition,
+    partition,
+    original_graph: networkx.DiGraph,
 ):
     """
     Calculates the change in score if u is moved to the community of the given neighbour.
@@ -30,7 +28,7 @@ def local_edge_ratio(
     :param partition: Total partition of the complete graph.
     :return: Change in local score
     """
-    nodes_in_u = original_graph.nodes[u].get("nodes", {u})
+    nodes_in_u = G.nodes[u].get("nodes", {u})
 
     old_score_u = edge_boundary_ratio(original_graph, nodes_in_u)
 
@@ -57,7 +55,7 @@ def edge_boundary_ratio(G: networkx.DiGraph, partition):
             else:
                 in_edge_size += 1
     in_edge_size = in_edge_size / 2
-    denom = (edge_boundary_size + in_edge_size)
+    denom = edge_boundary_size + in_edge_size
     if denom == 0:
         return 0
     return in_edge_size / denom
diff --git a/louvain.py b/louvain.py
@@ -70,8 +70,8 @@ def louvain_partitions(
         print("Executing iteration {}".format(counter))
         counter += 1
         yield partition
-        new_community_score = global_community_measure(graph, partition)
-        print(f"Calculated modularity: {new_community_score}")
+        new_community_score = global_community_measure(G, partition)
+        print(f"Calculated global measure score: {new_community_score}")
         if new_community_score - comm_score <= threshold:
             return
         comm_score = new_community_score
@@ -119,11 +119,16 @@ def _one_level(
             best_community_score = 0
             best_com = node_to_community[u]
 
-            # TODO - Compute for each neighbour, the increase in score.
             # We pass the node_to_community dict, as well as the current node, and its neighbours
             for neighbour in G.neighbors(u):
                 new_score = local_community_measure(
-                    G, u, neighbour, node_to_community, inner_partition, partition, original_graph
+                    G,
+                    u,
+                    neighbour,
+                    node_to_community,
+                    inner_partition,
+                    partition,
+                    original_graph,
                 )
                 if new_score > best_community_score:
                     best_community_score = new_score

diff --git a/main.py b/main.py
@@ -37,14 +37,13 @@ def load_network(
 def main():
     # G = load_network()
     # Barabasi-Albert graph
-    G = nx.barabasi_albert_graph(500, 5)
+    G = nx.barabasi_albert_graph(5000, 5)
     # Add edge weight of 1 to each edge
     for u, v, d in G.edges(data=True):
         d["weight"] = 1
     G = nx.to_directed(G)
     print(f"Loaded {len(G.nodes)} nodes and {len(G.edges)} edges")
 
-
     communities = louvain_communities(G, global_edge_ratio, local_edge_ratio)
     print(len(communities))