如何加速深度优先搜索方法？
How to speed up Depth First Search method?
我正在尝试对我的图表进行深度优先搜索，但有些东西大大减慢了它的速度，我不确定是什么原因。
这是我的包代码：
import java.util.Iterator;
import java.util.NoSuchElementException;
public class Bag<Item> implements Iterable<Item> {
private Node<Item> first; // beginning of bag
private Node<Item> end;
private int n; // number of elements in bag
public int label;
public int edges;
public static class Node<Item> {
private Item item;
private Node<Item> next;
public int label;
public int edges;
}
public Bag() {
first = null; // empty bag initialized
end = null;
n = 0;
}
public void add(Item item) {
if (n==0) {
Node<Item> head = new Node<Item>(); // if bag is empty
first = head;
end = head;
head.item = item; // new node both first and end of bag
edges++;
n++;
}
else {
Node<Item> oldlast = end; // old last assigned to end of node
Node<Item> last = new Node<Item>();
last.item = item;
oldlast.next = last; // new node added after old last
end = last;
n++; // size increased
edges++;
}
}
public Iterator<Item> iterator() {
return new LinkedIterator(first); // returns an iterator that iterates over the items in this bag in arbitrary order
}
public class LinkedIterator implements Iterator<Item> {
private Node<Item> current;
public LinkedIterator(Node<Item> first) {
current = first; // iterator starts at head of bag
}
public boolean hasNext() { return current != null; }
public void remove() { throw new UnsupportedOperationException(); }
public Item next() {
if (!hasNext()) throw new NoSuchElementException(); // if there is next item, current is moved to next
Item item = current.item;
current = current.next;
return item; // item is returned
}
}
}
这是我的 driver:
import java.util.ArrayList;
import java.util.Random;
public class Driver {
public static ArrayList<Integer> randomNum(int howMany) {
ArrayList<Integer> numbers = new ArrayList<Integer>(howMany);
Random randomGenerator = new Random();
while (numbers.size() < howMany) {
int rand_int = randomGenerator.nextInt(10000);
if (!numbers.contains(rand_int)) {
numbers.add(rand_int);
}
}
return numbers;
}
public static void main(String[] args) {
ArrayList<Integer> num = randomNum(100);
Graph G = new Graph(num);
System.out.println("The length of longest path for this sequence with graph is: " + G.dfsStart(num));
}
}
我从 driver 向我的 dfsStart 方法发送一个随机整数的 ArrayList，它会查看图中每个起始节点的所有不同路径。我的 DepthFirstSearch 方法为每个起始节点调用 getAdjList 以使用我的 Bag adj 找到它的邻居，然后在回溯之前沿着每条路径前进。
这是我的 Graph 代码，包含我的最长路径方法：
import java.util.ArrayList;
import java.util.NoSuchElementException;
public class Graph {
public final int V; // initializing variables and data structures
public Bag<Integer>[] adj;
public int longestPath;
public Graph(ArrayList<Integer> numbers) {
try {
longestPath = 0;
this.V = numbers.size();
adj = (Bag<Integer>[]) new Bag[V]; // bag initialized
for (int v = 0; v < V; v++) {
adj[v] = new Bag<Integer>();
}
for (int i = 0; i < V; i++) {
adj[i].label = numbers.get(i);
int j = (i + 1);
while (j < numbers.size()) {
if (numbers.get(i) < numbers.get(j)) {
addEdge(i, numbers.get(j));
}
j++;
}
}
}
catch (NoSuchElementException e) {
throw new IllegalArgumentException("invalid input format in Graph constructor", e);
}
}
public void addEdge(int index, int num) {
adj[index].add(num);
}
public int getIndex(int num) {
for (int i = 0; i < adj.length; i++) {
if (adj[i].label == num) {
return i;
}
}
return -1;
}
public Bag<Integer> getAdjList(int source) {
Bag<Integer> adjList = null;
for (Bag<Integer> list : adj) {
if (list.label == source) {
adjList = list;
break;
}
}
return adjList;
}
public int dfsStart(ArrayList<Integer> numbers) {
for (int i=0;i<numbers.size();i++) {
// Print all paths from current node
depthFirstSearch(numbers.get(i),new ArrayList<>(300));
}
return longestPath;
}
public void depthFirstSearch(int src, ArrayList<Integer> current) {
current.add(src);
Bag<Integer> srcAdj = getAdjList(src);
if (srcAdj.size() == 0) {
// Leaf node
// Print this path
longestPath = Math.max(longestPath, current.size());
}
for (int links : srcAdj) {
depthFirstSearch(links, current);
}
current.remove(current.size()-1);
}
}
我相信下面的建议有助于消除错误，但在超过 150 个顶点的图中尝试找到最长路径时，它仍然慢得令人难以置信。
即使对于一个小的密集图，也可以有许多来自 src 节点的唯一路径。我测试了这个输入 [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25] 有 16777216 来自所有节点的唯一路径。所以你可以期待 OOM 更大的输入。一种方法是在找到路径后立即更新 longestPath，而不是将其添加到列表中。
稍后更改。
addtoCount(current.size());
到
longestPath = Math.max(longestPath, current.size());
确保 longestPath 是 global 并在每个测试用例之前初始化为 0。
好吧，我不知道 JAVA 但是对于深度优先搜索这样的简单事情来说，代码量实在是太多了。
在 C++ 中，它是这样完成的：
void cPathFinder::depthFirst(
int v)
{
// initialize visited flag for each node in graph
myPath.clear();
myPath.resize(nodeCount(), 0);
// start recursive search from starting node
depthRecurse(v, visitor);
}
void cPathFinder::depthRecurse(
int v )
{
// remember this node has been visited
myPath[v] = 1;
// look for new adjacent nodes
for (int w : adjacent(v))
if (!myPath[w])
{
// search from new node
depthRecurse(w);
}
}
我正在尝试对我的图表进行深度优先搜索，但有些东西大大减慢了它的速度，我不确定是什么原因。
这是我的包代码：
import java.util.Iterator;
import java.util.NoSuchElementException;
public class Bag<Item> implements Iterable<Item> {
private Node<Item> first; // beginning of bag
private Node<Item> end;
private int n; // number of elements in bag
public int label;
public int edges;
public static class Node<Item> {
private Item item;
private Node<Item> next;
public int label;
public int edges;
}
public Bag() {
first = null; // empty bag initialized
end = null;
n = 0;
}
public void add(Item item) {
if (n==0) {
Node<Item> head = new Node<Item>(); // if bag is empty
first = head;
end = head;
head.item = item; // new node both first and end of bag
edges++;
n++;
}
else {
Node<Item> oldlast = end; // old last assigned to end of node
Node<Item> last = new Node<Item>();
last.item = item;
oldlast.next = last; // new node added after old last
end = last;
n++; // size increased
edges++;
}
}
public Iterator<Item> iterator() {
return new LinkedIterator(first); // returns an iterator that iterates over the items in this bag in arbitrary order
}
public class LinkedIterator implements Iterator<Item> {
private Node<Item> current;
public LinkedIterator(Node<Item> first) {
current = first; // iterator starts at head of bag
}
public boolean hasNext() { return current != null; }
public void remove() { throw new UnsupportedOperationException(); }
public Item next() {
if (!hasNext()) throw new NoSuchElementException(); // if there is next item, current is moved to next
Item item = current.item;
current = current.next;
return item; // item is returned
}
}
}
这是我的 driver:
import java.util.ArrayList;
import java.util.Random;
public class Driver {
public static ArrayList<Integer> randomNum(int howMany) {
ArrayList<Integer> numbers = new ArrayList<Integer>(howMany);
Random randomGenerator = new Random();
while (numbers.size() < howMany) {
int rand_int = randomGenerator.nextInt(10000);
if (!numbers.contains(rand_int)) {
numbers.add(rand_int);
}
}
return numbers;
}
public static void main(String[] args) {
ArrayList<Integer> num = randomNum(100);
Graph G = new Graph(num);
System.out.println("The length of longest path for this sequence with graph is: " + G.dfsStart(num));
}
}
我从 driver 向我的 dfsStart 方法发送一个随机整数的 ArrayList，它会查看图中每个起始节点的所有不同路径。我的 DepthFirstSearch 方法为每个起始节点调用 getAdjList 以使用我的 Bag adj 找到它的邻居，然后在回溯之前沿着每条路径前进。
这是我的 Graph 代码，包含我的最长路径方法：
import java.util.ArrayList;
import java.util.NoSuchElementException;
public class Graph {
public final int V; // initializing variables and data structures
public Bag<Integer>[] adj;
public int longestPath;
public Graph(ArrayList<Integer> numbers) {
try {
longestPath = 0;
this.V = numbers.size();
adj = (Bag<Integer>[]) new Bag[V]; // bag initialized
for (int v = 0; v < V; v++) {
adj[v] = new Bag<Integer>();
}
for (int i = 0; i < V; i++) {
adj[i].label = numbers.get(i);
int j = (i + 1);
while (j < numbers.size()) {
if (numbers.get(i) < numbers.get(j)) {
addEdge(i, numbers.get(j));
}
j++;
}
}
}
catch (NoSuchElementException e) {
throw new IllegalArgumentException("invalid input format in Graph constructor", e);
}
}
public void addEdge(int index, int num) {
adj[index].add(num);
}
public int getIndex(int num) {
for (int i = 0; i < adj.length; i++) {
if (adj[i].label == num) {
return i;
}
}
return -1;
}
public Bag<Integer> getAdjList(int source) {
Bag<Integer> adjList = null;
for (Bag<Integer> list : adj) {
if (list.label == source) {
adjList = list;
break;
}
}
return adjList;
}
public int dfsStart(ArrayList<Integer> numbers) {
for (int i=0;i<numbers.size();i++) {
// Print all paths from current node
depthFirstSearch(numbers.get(i),new ArrayList<>(300));
}
return longestPath;
}
public void depthFirstSearch(int src, ArrayList<Integer> current) {
current.add(src);
Bag<Integer> srcAdj = getAdjList(src);
if (srcAdj.size() == 0) {
// Leaf node
// Print this path
longestPath = Math.max(longestPath, current.size());
}
for (int links : srcAdj) {
depthFirstSearch(links, current);
}
current.remove(current.size()-1);
}
}
我相信下面的建议有助于消除错误，但在超过 150 个顶点的图中尝试找到最长路径时，它仍然慢得令人难以置信。
即使对于一个小的密集图，也可以有许多来自 src 节点的唯一路径。我测试了这个输入 [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25] 有 16777216 来自所有节点的唯一路径。所以你可以期待 OOM 更大的输入。一种方法是在找到路径后立即更新 longestPath，而不是将其添加到列表中。
稍后更改。
addtoCount(current.size());
到
longestPath = Math.max(longestPath, current.size());
确保 longestPath 是 global 并在每个测试用例之前初始化为 0。
好吧，我不知道 JAVA 但是对于深度优先搜索这样的简单事情来说，代码量实在是太多了。
在 C++ 中，它是这样完成的：
void cPathFinder::depthFirst(
int v)
{
// initialize visited flag for each node in graph
myPath.clear();
myPath.resize(nodeCount(), 0);
// start recursive search from starting node
depthRecurse(v, visitor);
}
void cPathFinder::depthRecurse(
int v )
{
// remember this node has been visited
myPath[v] = 1;
// look for new adjacent nodes
for (int w : adjacent(v))
if (!myPath[w])
{
// search from new node
depthRecurse(w);
}
}