Simon O'Shea
683 lines
18 KiB
Java
683 lines
18 KiB
Java
package sampleworld;
|
|
|
|
import java.util.Scanner;
|
|
import java.util.List;
|
|
import java.util.ArrayList;
|
|
import java.util.Arrays;
|
|
import java.util.Comparator;
|
|
import java.util.Stack;
|
|
import java.util.Queue;
|
|
import java.util.PriorityQueue;
|
|
import java.util.LinkedList;
|
|
import java.util.concurrent.TimeUnit;
|
|
|
|
@SuppressWarnings( "javadoc" )
|
|
public class SampleWorld
|
|
{
|
|
//////////////////////////////////////////////////////////
|
|
// GLOBAL VARIABLES THAT REPRESENT STATIC WORLD FEATURES//
|
|
//////////////////////////////////////////////////////////
|
|
|
|
// global variables used to hold world dimensions
|
|
public static int worldCols, worldRows;
|
|
|
|
// used to store coordinates of all obstacles in our world
|
|
public static List<int[]> obstacles = new ArrayList<int[]>();
|
|
|
|
// used to store all unique states that agent has visited (holds agent and sample coords)
|
|
public static List<State> closed = new ArrayList<State>();
|
|
|
|
// used to keep track of number of operations and is used in output
|
|
public static int expansions = 0;
|
|
public static int nodesGenerated = 0;
|
|
|
|
// helper variables just to make indexing arrays easier to read
|
|
public static int row = 0;
|
|
public static int col = 1;
|
|
|
|
public static void main(String[] args)
|
|
{
|
|
|
|
//////////////////////////////
|
|
// COLLECT BASIC WORLD INFO //
|
|
//////////////////////////////
|
|
Scanner input = new Scanner(System.in);
|
|
input.useDelimiter(System.lineSeparator());
|
|
|
|
worldCols = Integer.parseInt(input.next());
|
|
worldRows = Integer.parseInt(input.next());
|
|
|
|
/////////////////////////////
|
|
// CREATE WORLD STRUCTURES //
|
|
/////////////////////////////
|
|
char[][] world = new char[worldRows][worldCols];
|
|
List<int[]> samples = new ArrayList<int[]>();
|
|
int[] agent = new int[] {-1, -1};
|
|
|
|
////////////////////////////////////
|
|
// SAVE INITIAL STATE AS 2D ARRAY //
|
|
////////////////////////////////////
|
|
for(int rows = 0; rows < worldRows; rows++)
|
|
{
|
|
// grab one row at a time
|
|
char[] line = input.next().toCharArray();
|
|
|
|
for(int cols = 0; cols < worldCols; cols++)
|
|
{
|
|
// check each column for special characters
|
|
if(line[cols] == '@')
|
|
{
|
|
agent = new int[]{rows, cols};
|
|
}
|
|
if(line[cols] == '#')
|
|
{
|
|
obstacles.add(new int[]{rows, cols});
|
|
}
|
|
if(line[cols] == '*')
|
|
{
|
|
samples.add(new int[]{rows, cols});
|
|
}
|
|
|
|
// then add character to our local world
|
|
world[rows][cols] = line[cols];
|
|
}
|
|
}
|
|
|
|
input.close();
|
|
|
|
//////////////////////////////////////////////////
|
|
// CONSTRUCT INTIAL STATE NODE AND BEGIN SEARCH //
|
|
//////////////////////////////////////////////////
|
|
|
|
// initial state has no parent and no previous action. '0' is being interpreted as null here
|
|
Node initialState = new Node(null, agent, samples, '0', 0, 0);
|
|
Stack<Node> result = new Stack<Node>();
|
|
|
|
// select search algorithm to use
|
|
if(args[0].equals("dfs"))
|
|
result = dfs(initialState, -1);
|
|
|
|
if(args[0].equals("ucs"))
|
|
result = ucs(initialState);
|
|
|
|
if(args[0].equals("ids"))
|
|
result = ids(initialState);
|
|
|
|
if(args[0].equals("astar"))
|
|
result = astar(initialState, args[1]);
|
|
|
|
int length = result.size();
|
|
|
|
// decide whether to animate
|
|
if(args.length == 2)
|
|
{
|
|
if(args[1].equals("animate"))
|
|
{
|
|
try
|
|
{
|
|
animate(world, result);
|
|
}
|
|
catch (InterruptedException e)
|
|
{
|
|
e.printStackTrace();
|
|
}
|
|
System.out.println("Distance traveled: " + length + " steps");
|
|
System.out.println("Nodes Expanded: " + SampleWorld.expansions);
|
|
System.out.println("Total Nodes Generated: " + SampleWorld.nodesGenerated);
|
|
return;
|
|
}
|
|
}
|
|
else if(args.length == 3)
|
|
{
|
|
if(args[2].equals("animate"))
|
|
{
|
|
try {
|
|
animate(world, result);
|
|
} catch (InterruptedException e) {
|
|
e.printStackTrace();
|
|
}
|
|
System.out.println("Distance traveled: " + length + " steps");
|
|
System.out.println("Nodes Expanded: " + SampleWorld.expansions);
|
|
System.out.println("Total Nodes Generated: " + SampleWorld.nodesGenerated);
|
|
return;
|
|
}
|
|
}
|
|
|
|
// Print the results and searching stats
|
|
if(length == 0)
|
|
{
|
|
System.out.println("No solution found :(");
|
|
}
|
|
else
|
|
{
|
|
for(int i = 0; i < length; i++)
|
|
{
|
|
Node state = result.pop();
|
|
System.out.println(state.lastMove);
|
|
}
|
|
}
|
|
|
|
System.out.println("\nDistance traveled: " + length + " steps");
|
|
System.out.println("Nodes Expanded: " + SampleWorld.expansions);
|
|
System.out.println("Total Nodes Generated: " + SampleWorld.nodesGenerated);
|
|
}
|
|
|
|
// helper function for me. is fun to watch and also is genuinely useful for debugging
|
|
public static void animate(char[][] world, Stack<Node> result) throws InterruptedException
|
|
{
|
|
clearScreen();
|
|
int agentR, agentC;
|
|
|
|
// print out initial state
|
|
for(int r = 0; r < worldRows; r++)
|
|
{
|
|
for(int c = 0; c < worldCols; c++)
|
|
{
|
|
System.out.print(world[r][c]);
|
|
}
|
|
// only make newline after row if there is another row
|
|
if(r != worldCols - 1)
|
|
System.out.println();
|
|
}
|
|
TimeUnit.MILLISECONDS.sleep(2000);
|
|
clearScreen();
|
|
|
|
// read and print each of the agent's moves
|
|
while(!result.isEmpty())
|
|
{
|
|
Node state = result.pop();
|
|
agentR = state.agent[row];
|
|
agentC = state.agent[col];
|
|
|
|
world[agentR][agentC] = '@';
|
|
|
|
|
|
// does inverse of each move to replace agent's old location with _
|
|
if(state.lastMove == 'U')
|
|
{
|
|
world[agentR + 1][agentC] = '_';
|
|
}
|
|
if(state.lastMove == 'D')
|
|
{
|
|
world[agentR - 1][agentC] = '_';
|
|
}
|
|
if(state.lastMove == 'L')
|
|
{
|
|
world[agentR][agentC + 1] = '_';
|
|
}
|
|
if(state.lastMove == 'R')
|
|
{
|
|
world[agentR][agentC - 1] = '_';
|
|
}
|
|
if(state.lastMove == 'S')
|
|
{
|
|
world[agentR][agentC] = 'S';
|
|
}
|
|
|
|
// TODO try a new loop that just erases the line the agent was/is on
|
|
|
|
// print out new world
|
|
for(int r = 0; r < worldRows; r++)
|
|
{
|
|
for(int c = 0; c < worldCols; c++)
|
|
{
|
|
System.out.print(world[r][c]);
|
|
}
|
|
System.out.println();
|
|
}
|
|
|
|
TimeUnit.SECONDS.sleep(1);
|
|
clearScreen();
|
|
}
|
|
|
|
}
|
|
// helper function for animate()
|
|
public static void clearScreen()
|
|
{
|
|
System.out.println("\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n");
|
|
}
|
|
|
|
///////////////////////////////////
|
|
// SUITE OF SEARCHING ALGORITHMS //
|
|
///////////////////////////////////
|
|
public static Stack<Node> astar(Node initialState, String heuristic)
|
|
{
|
|
Stack<Node> solution = new Stack<Node>();
|
|
Queue<Node> open = new PriorityQueue<Node>();
|
|
|
|
open.add(initialState);
|
|
while(true)
|
|
{
|
|
if(open.isEmpty())
|
|
return solution;
|
|
|
|
Node currentNode = open.remove();
|
|
|
|
// check whether we are in goal state
|
|
if(currentNode.samples.size() == 0)
|
|
{
|
|
while(currentNode.parent != null)
|
|
{
|
|
solution.push(currentNode);
|
|
currentNode = currentNode.parent;
|
|
}
|
|
return solution;
|
|
}
|
|
|
|
// if not, expand children and continue
|
|
List<Node> children = currentNode.expand(-1);
|
|
|
|
if(heuristic.equals("h0"))
|
|
h0(children);
|
|
else if(heuristic.equals("h1"))
|
|
h1(children);
|
|
else if(heuristic.equals("h2"))
|
|
h2(children);
|
|
|
|
for(Node child : children)
|
|
{
|
|
open.add(child);
|
|
}
|
|
}
|
|
}
|
|
|
|
public static Stack<Node> ids(Node initialState)
|
|
{
|
|
Stack<Node> result = new Stack<Node>();
|
|
int depth = 1;
|
|
|
|
while(true)
|
|
{
|
|
result = dfs(initialState, depth);
|
|
if(result.size() != 0)
|
|
return result;
|
|
depth++;
|
|
SampleWorld.closed.clear();
|
|
}
|
|
}
|
|
|
|
public static Stack<Node> ucs(Node initialState)
|
|
{
|
|
// used to store nodes not visited (generated but not expanded)
|
|
Queue<Node> open = new LinkedList<Node>();
|
|
open.add(initialState);
|
|
|
|
Stack<Node> solution = new Stack<Node>();
|
|
while(true)
|
|
{
|
|
if(open.isEmpty())
|
|
return solution;
|
|
|
|
Node currentNode = open.remove();
|
|
|
|
// check if agent is in goal state
|
|
if(currentNode.samples.size() == 0)
|
|
{
|
|
while(currentNode.parent != null)
|
|
{
|
|
solution.push(currentNode);
|
|
currentNode = currentNode.parent;
|
|
}
|
|
return solution;
|
|
}
|
|
|
|
// otherwise, expand and continue down path
|
|
List<Node> children = currentNode.expand(-1);
|
|
for(Node child : children)
|
|
{
|
|
open.add(child);
|
|
}
|
|
}
|
|
}
|
|
|
|
public static Stack<Node> dfs(Node initialState, int depth)
|
|
{
|
|
// used to store nodes not visited (generated but not expanded)
|
|
Stack<Node> open = new Stack<Node>();
|
|
open.push(initialState);
|
|
|
|
Stack<Node> solution = new Stack<Node>();
|
|
|
|
while(true)
|
|
{
|
|
// if open is empty, we have exhausted all of our options and there is no solution
|
|
if(open.empty())
|
|
return solution;
|
|
|
|
Node currentNode = open.pop();
|
|
|
|
// check if agent is in goal state
|
|
if(currentNode.samples.size() == 0)
|
|
{
|
|
while(currentNode.parent != null)
|
|
{
|
|
solution.push(currentNode);
|
|
currentNode = currentNode.parent;
|
|
}
|
|
return solution;
|
|
}
|
|
|
|
// otherwise, expand and continue down path
|
|
List<Node> children = currentNode.expand(depth);
|
|
for(Node child : children)
|
|
{
|
|
open.push(child);
|
|
}
|
|
}
|
|
}
|
|
|
|
|
|
//////////////////////////////////
|
|
// SUITE OF HEURISTIC FUNCTIONS //
|
|
//////////////////////////////////
|
|
|
|
public static List<Node> h0(List<Node> children)
|
|
{
|
|
Queue<Node> sorted = new PriorityQueue<>(new NodeComparator());
|
|
|
|
for(int i = 0; i < children.size(); i++)
|
|
{
|
|
// estimated moves left: zero
|
|
children.get(i).heuristic = 0;
|
|
sorted.add(children.get(i));
|
|
}
|
|
|
|
return children;
|
|
}
|
|
|
|
public static List<Node> h1(List<Node> children)
|
|
{
|
|
Queue<Node> sorted = new PriorityQueue<>(new NodeComparator());
|
|
|
|
for(int i = 0; i < children.size(); i++)
|
|
{
|
|
// estimated moves left: samples left
|
|
children.get(i).heuristic = children.get(i).samples.size();
|
|
sorted.add(children.get(i));
|
|
}
|
|
return children;
|
|
}
|
|
|
|
public static Queue<Node> h2(List<Node> children)
|
|
{
|
|
Queue<Node> sorted = new PriorityQueue<>(new NodeComparator());
|
|
|
|
if(children.size() == 0)
|
|
return sorted;
|
|
|
|
Node child;
|
|
|
|
for(int i = 0; i < children.size(); i++)
|
|
{
|
|
// estimated moves left:
|
|
// distance to nearest sample + 1 (+1 to account for sample action)
|
|
child = children.get(i);
|
|
child.heuristic = Node.distance(child.agent, child.nearestSample()) + 1;
|
|
|
|
sorted.add(child);
|
|
}
|
|
|
|
return sorted;
|
|
}
|
|
}
|
|
// end class A1
|
|
|
|
|
|
|
|
|
|
// Node is used to store all of the current state's details
|
|
class Node implements Comparable<Node>
|
|
{
|
|
// helper variables just to make indexing arrays easier to read
|
|
public static int row = 0;
|
|
public static int col = 1;
|
|
|
|
Node parent;
|
|
int[] agent;
|
|
List<int[]> samples;
|
|
char lastMove; // for output/animation
|
|
int distanceTraveled; // for calculating f(n) value
|
|
double heuristic;
|
|
double fn; // distance traveled + heuristic value
|
|
boolean canSample; // for expansion
|
|
|
|
// construct node :)
|
|
public Node(Node parent, int[] agent, List<int[]> samples, char lastMove, int distanceTraveled, double heuristic)
|
|
{
|
|
this.parent = parent;
|
|
this.agent = agent;
|
|
this.samples = new ArrayList<int[]>(samples);
|
|
this.lastMove = lastMove;
|
|
this.distanceTraveled = distanceTraveled;
|
|
this.heuristic = heuristic;
|
|
this.fn = distanceTraveled + heuristic;
|
|
}
|
|
|
|
// determine valid moves and collect children to be sent back to search
|
|
public List<Node> expand(Integer depth)
|
|
{
|
|
List<Node> children = new ArrayList<Node>();
|
|
if(this.distanceTraveled >= depth && depth != -1)
|
|
return children;
|
|
|
|
// document expansion of node and get ready to collect this node's children
|
|
SampleWorld.expansions++;
|
|
int[] onSample = new int[2];
|
|
|
|
// since this state is being currently visited (expanded), put in closed list
|
|
SampleWorld.closed.add(this.getState());
|
|
|
|
////////////////////////////////////
|
|
// BEGIN CHECKING FOR VALID MOVES //
|
|
////////////////////////////////////
|
|
|
|
// store coordinates for all potential moves to be checked
|
|
int[] up = { this.agent[row] - 1, this.agent[col] };
|
|
int[] down = { this.agent[row] + 1, this.agent[col] };
|
|
int[] left = { this.agent[row], this.agent[col] - 1 };
|
|
int[] right = { this.agent[row], this.agent[col] + 1 };
|
|
|
|
// make sure going up doesn't go outside world-bounds or into obstacle
|
|
if(isOpen(up))
|
|
{
|
|
// if move is valid, create that new node/state and document
|
|
Node child = new Node(this, up, this.samples, 'U', this.distanceTraveled+1, this.heuristic);
|
|
SampleWorld.nodesGenerated++;
|
|
|
|
// make sure that we have not already made that move
|
|
if(!child.getState().inClosed())
|
|
children.add(child);
|
|
}
|
|
|
|
// same idea but for the different potential moves
|
|
if(isOpen(down))
|
|
{
|
|
Node child = new Node(this, down, this.samples, 'D', this.distanceTraveled+1, this.heuristic);
|
|
SampleWorld.nodesGenerated++;
|
|
|
|
if(!child.getState().inClosed())
|
|
children.add(child);
|
|
}
|
|
|
|
if(isOpen(left))
|
|
{
|
|
Node child = new Node(this, left, this.samples, 'L', this.distanceTraveled+1, this.heuristic);
|
|
SampleWorld.nodesGenerated++;
|
|
|
|
if(!child.getState().inClosed())
|
|
children.add(child);
|
|
}
|
|
|
|
if(isOpen(right))
|
|
{
|
|
Node child = new Node(this, right, this.samples, 'R', this.distanceTraveled+1, this.heuristic);
|
|
SampleWorld.nodesGenerated++;
|
|
|
|
if(!child.getState().inClosed())
|
|
children.add(child);
|
|
}
|
|
|
|
// CHECK IF CAN SAMPLE
|
|
onSample = canSample();
|
|
if(onSample[0] == 1)
|
|
{
|
|
Node child = new Node(this, this.agent, this.samples, 'S', this.distanceTraveled+1, this.heuristic);
|
|
child.samples.remove(onSample[1]);
|
|
|
|
SampleWorld.nodesGenerated++;
|
|
children.add(child);
|
|
}
|
|
|
|
return children;
|
|
}
|
|
|
|
// helper for expand, verifies whether potential move is legal
|
|
public boolean isOpen(int[] position)
|
|
{
|
|
// check that agent is not trying to move into an obstacle
|
|
for(int i = 0; i < SampleWorld.obstacles.size(); i++)
|
|
{
|
|
if(Arrays.compare(SampleWorld.obstacles.get(i), position) == 0)
|
|
return false;
|
|
}
|
|
|
|
// check that agent is not stepping out of the world
|
|
if(!((position[row] >= 0) && (position[row] <= SampleWorld.worldRows - 1) && (position[col] >= 0) && (position[col] <= SampleWorld.worldCols - 1)))
|
|
return false;
|
|
|
|
return true;
|
|
}
|
|
|
|
// returns the coordinates of the sample closest to the agent's current location
|
|
public int[] nearestSample()
|
|
{
|
|
if(samples.size() == 0)
|
|
return agent;
|
|
|
|
int[] s = samples.get(0);
|
|
double lowest = distance(agent, samples.get(0));
|
|
double distance;
|
|
|
|
for(int i = 1; i < samples.size(); i++)
|
|
{
|
|
distance = distance(agent, samples.get(i));
|
|
if(distance < lowest)
|
|
{
|
|
lowest = distance;
|
|
s = samples.get(i);
|
|
}
|
|
}
|
|
return s;
|
|
}
|
|
// helper function for nearestSample()
|
|
public static double distance(int[] a, int[] b)
|
|
{ // _________________________
|
|
// distance between 2D points = √(y2 - y1)^2 + (x2 - x1)^2
|
|
double total = (((b[row] - a[row])*(b[row] - a[row])) + ((b[col] - a[col])*(b[col] - a[col])));
|
|
total = Math.sqrt(total);
|
|
|
|
return total;
|
|
}
|
|
|
|
// helper function for h2
|
|
/////////////////////////////
|
|
// returns two pieces of information if true:
|
|
// [0] is 1 to indicate agent can sample
|
|
// [1] is the index of the valid sample in the list
|
|
public int[] canSample()
|
|
{
|
|
// default to false
|
|
int[] result = new int[2];
|
|
result[0] = 0;
|
|
result[1] = -1;
|
|
|
|
for(int i = 0; i < this.samples.size(); i++)
|
|
{
|
|
if((this.agent[row] == samples.get(i)[row]) && (this.agent[col] == samples.get(i)[col]))
|
|
{
|
|
result[0] = 1;
|
|
result[1] = i;
|
|
return result;
|
|
}
|
|
}
|
|
return result;
|
|
}
|
|
|
|
// returns only the dynamic information directly related to the Node's state.
|
|
// we need a way to extract just this information for the sake of our closed list
|
|
public State getState()
|
|
{
|
|
return new State(this.agent, this.samples);
|
|
}
|
|
|
|
// we use to override comparisons for priorityQueues so that our heuristic calculations are actually used
|
|
@Override
|
|
public int compareTo(Node other)
|
|
{
|
|
if(this.fn > other.fn)
|
|
return 1;
|
|
else
|
|
return -1;
|
|
}
|
|
|
|
}
|
|
// helper class for Node
|
|
class NodeComparator implements Comparator<Node>
|
|
{
|
|
// used to sort nodes based on sum of distance traveled + heuristic estimation of work left
|
|
@Override
|
|
public int compare(Node a, Node b)
|
|
{
|
|
if(a.fn > b.fn)
|
|
return 1;
|
|
else
|
|
return -1;
|
|
}
|
|
}
|
|
// helper class for Node
|
|
class State
|
|
{
|
|
int[] agent;
|
|
int[][] samples;
|
|
|
|
public State(int[] agent, List<int[]> samples)
|
|
{
|
|
this.agent = agent;
|
|
this.samples = new int[samples.size()][];
|
|
|
|
// convert List<int[]> into 2D array of ints [][]
|
|
for(int i = 0; i < samples.size(); i++)
|
|
{
|
|
this.samples[i] = samples.get(i).clone();
|
|
}
|
|
}
|
|
|
|
public boolean inClosed()
|
|
{
|
|
for(int i = 0; i < SampleWorld.closed.size(); i++)
|
|
{
|
|
if(this.equals(SampleWorld.closed.get(i)))
|
|
{
|
|
return true;
|
|
}
|
|
}
|
|
return false;
|
|
}
|
|
|
|
public boolean equals(State other)
|
|
{
|
|
if(this.samples.length != other.samples.length)
|
|
return false;
|
|
|
|
if(Arrays.compare(this.agent, other.agent) != 0)
|
|
return false;
|
|
|
|
for(int i = 0; i < this.samples.length; i++)
|
|
{
|
|
if(Arrays.compare(this.samples[i], other.samples[i]) != 0)
|
|
return false;
|
|
}
|
|
|
|
return true;
|
|
}
|
|
} |