Optimization of the OptiSim algorithm (For review)

selector/methods/distance.py

@@ -365,6 +365,16 @@ def algorithm(self, x, max_size) -> Union[List, Iterable]:
         # bv will serve as a mask to discard points within radius r of previously selected points
         bv = np.zeros(n_samples)
         candidates = list(range(n_samples))
+
+        # Initialize min_dists array to store minimum distances from points and selected points
+        # Initially consider all minimum distances as Infinity
+        min_dists = np.full(n_samples, np.inf)
+
+        # Calculate distances of points from all initially selected points
+        for idx in selected:
+            dists = np.linalg.norm(x - x[idx], axis=1)
+            min_dists = np.minimum(min_dists, dists)
+
         # determine which points are within radius r of initial point
         # note: workers=-1 uses all available processors/CPUs
         index_remove = tree.query_ball_point(
@@ -384,20 +394,19 @@ def algorithm(self, x, max_size) -> Union[List, Iterable]:
             except ValueError:
                 sublist = candidates.compressed()
 
-            # create a new kd-tree for nearest neighbor lookup with candidates
-            new_tree = spatial.KDTree(x[selected])
-            # query the kd-tree for nearest neighbors to selected samples
-            # note: workers=-1 uses all available processors/CPUs
-            search, _ = new_tree.query(x[sublist], eps=self.eps, p=self.p, workers=-1)
-            # identify the nearest neighbor with the largest distance from previously selected samples
-            best_idx = sublist[np.argmax(search)]
+            # Select and Append the candidate farthest from its nearest selected point
+            best_idx = sublist[np.argmax(min_dists[sublist])]
             selected.append(best_idx)
 
             count += 1
             if count > max_size:
                 # do this if you have reached the maximum number of points selected
                 return selected
 
+            # Update min_dists array: calculate distances from newly selected point
+            new_dists = np.linalg.norm(x - x[best_idx], axis=1)
+            min_dists = np.minimum(min_dists, new_dists)
+
             # eliminate all samples within radius r of the selected sample
             index_remove = tree.query_ball_point(
                 x[best_idx], self.r, eps=self.eps, p=self.p, workers=-1