JavaAlgorithms/DataStructures/Graphs/A_Star.java
2020-06-19 18:10:49 +02:00

160 lines
6.6 KiB
Java

/*
Time Complexity = O(E), where E is equal to the number of edges
*/
package A_Star;
import java.util.*;
public class A_Star {
private static class Graph {
//Graph's structure can be changed only applying changes to this class.
private ArrayList<Edge> [] graph;
//Initialise ArrayLists in Constructor
public Graph(int size) {
this.graph = new ArrayList[size];
for (int i = 0; i < size; i++) {
this.graph[i] = new ArrayList<>();
}
}
private ArrayList<Edge> getNeighbours(int from) { return this.graph[from]; }
//Graph is bidirectional, for just one direction remove second instruction of this method.
private void addEdge (Edge edge) {
this.graph[edge.getFrom()].add(new Edge(edge.getFrom(), edge.getTo(), edge.getWeight()));
this.graph[edge.getTo()].add(new Edge(edge.getTo(), edge.getFrom(), edge.getWeight()));
}
}
private static class Edge {
private int from;
private int to;
private int weight;
public Edge(int from, int to, int weight) {
this.from = from;
this.to = to;
this.weight = weight;
}
public int getFrom() { return from; }
public int getTo() { return to; }
public int getWeight() { return weight; }
}
//class to iterate during the algorithm execution, and also used to return the solution.
private static class PathAndDistance {
private int distance; //distance advanced so far.
private ArrayList<Integer> path; //list of visited nodes in this path.
private int estimated; //heuristic value associated to the last node od the path (current node).
public PathAndDistance(int distance, ArrayList<Integer> path, int estimated) {
this.distance = distance;
this.path = path;
this.estimated = estimated;
}
public int getDistance() { return distance; }
public ArrayList<Integer> getPath() { return path; }
public int getEstimated() { return estimated; }
private void printSolution () {
if (this.path != null)
System.out.println("Optimal path: " + this.path.toString() +
", distance: " + this.distance);
else
System.out.println("There is no path available to connect the points");
}
}
private static void initializeGraph(Graph graph, ArrayList<Integer> data) {
for (int i = 0; i < data.size(); i+=4) {
graph.addEdge(new Edge(data.get(i), data.get(i + 1), data.get(i + 2)));
}
/*
.x. node
(y) cost
- or | or / bidirectional connection
( 98)- .7. -(86)- .4.
|
( 85)- .17. -(142)- .18. -(92)- .8. -(87)- .11.
|
. 1. -------------------- (160)
| \ |
(211) \ .6.
| \ |
. 5. (101)-.13. -(138) (115)
| | | /
( 99) ( 97) | /
| | | /
.12. -(151)- .15. -(80)- .14. | /
| | | | /
( 71) (140) (146)- .2. -(120)
| | |
.19. -( 75)- . 0. .10. -(75)- .3.
| |
(118) ( 70)
| |
.16. -(111)- .9.
*/
}
public static void main(String[] args) {
//heuristic function optimistic values
int[] heuristic = {366, 0, 160, 242, 161, 178, 77, 151, 226,
244, 241, 234, 380, 98, 193, 253, 329, 80, 199, 374};
Graph graph = new Graph(20);
ArrayList<Integer> graphData = new ArrayList<>(Arrays.asList(0, 19, 75, null,
0, 15, 140, null, 0, 16, 118, null, 19, 12, 71, null, 12, 15, 151, null,
16, 9, 111, null, 9, 10, 70, null, 10, 3, 75, null, 3, 2, 120, null,
2, 14, 146, null, 2, 13, 138, null, 2, 6, 115, null, 15, 14, 80, null,
15, 5, 99, null, 14, 13, 97, null, 5, 1, 211, null, 13, 1, 101, null,
6, 1, 160, null, 1, 17, 85, null, 17, 7, 98, null, 7, 4, 86, null,
17, 18, 142, null, 18, 8, 92, null, 8, 11, 87));
initializeGraph(graph, graphData);
PathAndDistance solution = aStar(3, 1, graph, heuristic);
solution.printSolution();
}
public static PathAndDistance aStar(int from, int to, Graph graph, int[] heuristic) {
//nodes are prioritised by the less value of the current distance of their paths, and the estimated value
//given by the heuristic function to reach the destination point from the current point.
PriorityQueue<PathAndDistance> queue = new PriorityQueue<>
(Comparator.comparingInt(a -> (a.getDistance() + a.getEstimated())));
//dummy data to start the algorithm from the beginning point
queue.add(new PathAndDistance(0, new ArrayList<>(Arrays.asList(from)), 0));
boolean solutionFound = false;
PathAndDistance currentData = new PathAndDistance(-1, null, -1);
while (!queue.isEmpty() && !solutionFound) {
currentData = queue.poll(); //first in the queue, best node so keep exploring.
int currentPosition = currentData.getPath().get(currentData.getPath().size() - 1); //current node.
if (currentPosition == to)
solutionFound = true;
else
for (Edge edge : graph.getNeighbours(currentPosition))
if (!currentData.getPath().contains(edge.getTo())) { //Avoid Cycles
ArrayList<Integer> updatedPath = new ArrayList<>(currentData.getPath());
updatedPath.add(edge.getTo()); //Add the new node to the path, update the distance,
// and the heuristic function value associated to that path.
queue.add(new PathAndDistance(currentData.getDistance() + edge.getWeight(),
updatedPath, heuristic[edge.getTo()]));
}
}
return (solutionFound) ? currentData : new PathAndDistance(-1, null, -1);
//Out of while loop, if there is a solution, the current Data stores the optimal path, and its distance
}
}