perf(motif): use HashSet lookups in getLocalClustering instead of graph.isNeighbor

Replace JUNG's graph.isNeighbor() calls in the local clustering
coefficient computation with O(1) HashSet lookups.  JUNG's isNeighbor
can be O(degree) per call, making the per-vertex cost O(k^3) for
high-degree nodes.  The new approach builds neighbour sets on demand
(or reuses them if called during analyze()), reducing per-vertex cost
to O(k^2).  This matches the pattern already used in findOpenPaths
and findSquares.

Author: sauravbhattacharya001

Gvisual/src/gvisual/GraphMotifFinder.java

@@ -265,28 +265,48 @@ public double getClusteringCoefficient() {
 
     /**
      * Compute per-node local clustering coefficient.
+     *
+     * <p>Uses O(1) HashSet lookups instead of JUNG's {@code isNeighbor()}
+     * which can be O(degree) per call, reducing per-vertex cost from
+     * O(k³) to O(k²).</p>
+     *
      * @return map of node → local clustering coefficient
      */
     public Map<String, Double> getLocalClustering() {
+        // Build neighbour sets if not available (analyze() nulls them out)
+        boolean builtLocally = false;
+        Map<String, Set<String>> nbSets = this.neighborSets;
+        if (nbSets == null) {
+            builtLocally = true;
+            nbSets = new HashMap<>(graph.getVertexCount() * 2);
+            for (String v : graph.getVertices()) {
+                nbSets.put(v, new HashSet<>(graph.getNeighbors(v)));
+            }
+        }
+
         Map<String, Double> result = new LinkedHashMap<>();
         for (String v : graph.getVertices()) {
-            List<String> neighbors = new ArrayList<>(graph.getNeighbors(v));
-            int k = neighbors.size();
+            Collection<String> rawNbrs = graph.getNeighbors(v);
+            int k = rawNbrs.size();
             if (k < 2) {
                 result.put(v, 0.0);
                 continue;
             }
+            List<String> neighbors = new ArrayList<>(rawNbrs);
             int links = 0;
             for (int i = 0; i < neighbors.size(); i++) {
+                Set<String> iNbrs = nbSets.get(neighbors.get(i));
                 for (int j = i + 1; j < neighbors.size(); j++) {
-                    if (graph.isNeighbor(neighbors.get(i), neighbors.get(j))) {
+                    if (iNbrs.contains(neighbors.get(j))) {
                         links++;
                     }
                 }
             }
             double maxLinks = k * (k - 1.0) / 2.0;
             result.put(v, links / maxLinks);
         }
+
+        if (builtLocally) nbSets = null; // let GC reclaim
         return result;
     }