Pauline - Fire

lib/dijkstra.rb

@@ -1,5 +1,52 @@
+# Time Complexity: O(v^2) - Since we are performing a variety of BFS on an adjacency matrix,
+# the time complexity to search all edges will be, at worst case, O(v^2), assuming all vertices v
+# are interconnected. This is because for each node visited, the entire associated row on the 
+# adjacency matrix has to be checked to confirm which edges are actual graph edges. 
 
+# Space Complexity: O(v) - A couple arrays are create, size directly equal to the number of vertices v. 
+# Likewise, the Queue and Set created to track nodes to visit and visited nodes are also directly
+# equal to tne number of vertices in the graph. The return hash is simply a wrapper that envelops
+# the shortest_distances and parent_list arrays. 
 def dijkstra(adjacency_matrix, start_node) 
+    shortest_distances = Array.new(adjacency_matrix[0].length){Float::INFINITY}
+    parent_list = Array.new(shortest_distances.length){nil} 
+    tracking_q = Queue.new
+    visited_nodes = Set.new
 
-
+    # start with start node
+    # start node has no previous node, marked as 0
+    visited_nodes.add(start_node)
+    shortest_distances[start_node] = 0
+
+    adjacency_matrix[start_node].each_with_index do |weight, i|
+        if weight > 0 && i != start_node
+            parent_list[i] = start_node
+            shortest_distances[i] = weight
+            tracking_q.push(i)
+        end
+    end
+
+    until (tracking_q.empty?) 
+        cur_node = tracking_q.pop
+        # skip already visited nodes
+        next if visited_nodes.include?(cur_node)
+        # check all adjacent nodes
+        visited_nodes.add(cur_node)
+
+        # check for adjacent nodes
+        adjacency_matrix[cur_node].each_with_index do |weight, i|
+            # ignore those with no weight
+            if weight > 0 && i != cur_node
+                if shortest_distances[cur_node] + weight < shortest_distances[i]
+                    shortest_distances[i] = shortest_distances[cur_node] + weight
+                    parent_list[i] = cur_node
+                end
+                tracking_q.push(i)
+            end
+        end
+    end
+
+    return {start_node: start_node, 
+            parent_list: parent_list, 
+            shortest_distances: shortest_distances}
 end