我试图对普林斯顿进行算法编程，我遇到了关于内存测试的问题。分配要求我们运行渗透程序N次并找到结果的介质，并且我写了一个percolationtest.java，并且每次都创建一个实例变量，它工作，但是使用了太多的内存，教师建议我使用局部变量，但我不知道如何。有人可以帮助我并给我一些建议，我真的很感激。
public class PercolationStats {
private int N, T, totalSum;
private double []fraction;
private int []count;
public PercolationStats(int N, int T) {
if (N <= 0 || T <= 0)
throw new IllegalArgumentException();
else {
this.N = N;
this.T = T;
count = new int [T];
totalSum = N*N;
fraction = new double[T];
int randomX, randomY;
for (int i = 0; i < T; i++) {
Percolation perc = new Percolation(N);
while (true) {
if (perc.percolates()) {
fraction[i] = (double) count[i]/totalSum;
break;
}
randomX = StdRandom.uniform(1, N+1);
randomY = StdRandom.uniform(1, N+1);
if (perc.isOpen(randomX, randomY)) continue;
else {
perc.open(randomX, randomY);
count[i]++;
}
}
}
}
} // perform T independent experiments on an N-by-N grid
public double mean() {
double totalFraction = 0;
for (int i = 0; i < T; i++) {
totalFraction += fraction[i];
}
return totalFraction/T;
} // sample mean of percolation threshold
public double stddev() {
double u = this.mean();
double sum = 0;
for (int i = 0; i < T; i++) {
sum += (fraction[i] - u) * (fraction[i] - u);
}
return Math.sqrt(sum/(T-1));
} // sample standard deviation of percolation threshold
public double confidenceLo() {
double u = this.mean();
double theta = this.stddev();
double sqrtT = Math.sqrt(T);
return u-1.96*theta/sqrtT;
} // low endpoint of 95% confidence interval
public double confidenceHi() {
double u = this.mean();
double theta = this.stddev();
double sqrtT = Math.sqrt(T);
return u+1.96*theta/sqrtT;
} // high endpoint of 95% confidence interval
public static void main(String[] args) {
int N = 200;
int T = 100;
if (args.length == 1) N = Integer.parseInt(args[0]);
else if (args.length == 2) {
N = Integer.parseInt(args[0]);
T = Integer.parseInt(args[1]); }
PercolationStats a = new PercolationStats(N, T);
System.out.print("mean = ");
System.out.println(a.mean());
System.out.print("stddev = ");
System.out.println(a.stddev());
System.out.print("95% confidence interval = ");
System.out.print(a.confidenceLo());
System.out.print(", ");
System.out.println(a.confidenceHi());
} }
public class Percolation {
private boolean[][] site;
private WeightedQuickUnionUF uf;
private int N;
public Percolation(int N) {
if (N < 1)
throw new IllegalArgumentException();
else {
site = new boolean[N + 2][N + 2];
for (int j = 1; j <= N; j++) {
site[0][j] = true;
site[N + 1][j] = true;
}
uf = new WeightedQuickUnionUF((N + 2) * (N + 2));
for (int i = 1; i <= N; i++) {
uf.union(0, i);
}
this.N = N;
}
}
public void open(int i, int j) {
if (i > N || i < 1 || j > N || j < 1)
throw new IndexOutOfBoundsException();
else {
if (!site[i][j]) {
site[i][j] = true;
if (site[i - 1][j]) {
uf.union((N + 2) * (i - 1) + j, (N + 2) * i + j);
}
if (site[i + 1][j]) {
uf.union((N + 2) * i + j, (N + 2) * (i + 1) + j);
}
if (site[i][j + 1]) {
uf.union((N + 2) * i + (j + 1), (N + 2) * i + j);
}
if (site[i][j - 1]) {
uf.union((N + 2) * i + (j - 1), (N + 2) * i + j);
}
}
}
}
public boolean isOpen(int i, int j) {
if (i > N || i < 1 || j > N || j < 1)
throw new IndexOutOfBoundsException();
else
return site[i][j];
}
public boolean isFull(int i, int j) {
if (i > N || i < 1 || j > N || j < 1)
throw new IndexOutOfBoundsException();
else
return site[i][j] && (i == 1 || uf.connected((N + 2) * i + j, 0));
}
public boolean percolates() {
for (int i = 1; i <= N; i++) {
if (this.isFull(N, i)) {
return true;
}
}
return false;
}
public static void main(String[] args) {
}
}
答案 0 :(得分：0)
添加了meanValue实例变量以保持平均值，并将其替换为您曾经调用过的mean（）方法的多个地方，这些方法一次又一次地计算。同样修改了＆＃34; int [] count＆＃34;作为您在构造函数外部未使用的局部变量。发布你的＆＃34; Percolation＆＃34;和＆＃34; StdRandom＆＃34;用于更多代码优化的类。您可以运行此代码并进行测试，它应该减少运行时间而不是您的。
public class PercolationStats {
private int N, T, totalSum;
private double []fraction;
private double meanValue;
public PercolationStats(int N, int T) {
if (N <= 0 || T <= 0)
throw new IllegalArgumentException();
else {
this.N = N;
this.T = T;
int [] count = new int [T];
totalSum = N*N;
fraction = new double[T];
int randomX, randomY;
for (int i = 0; i < T; i++) {
Percolation perc = new Percolation(N);
while (true) {
if (perc.percolates()) {
fraction[i] = (double) count[i]/totalSum;
break;
}
randomX = StdRandom.uniform(1, N+1);
randomY = StdRandom.uniform(1, N+1);
if (perc.isOpen(randomX, randomY)) continue;
else {
perc.open(randomX, randomY);
count[i]++;
}
}
}
}
}
// perform T independent experiments on an N-by-N grid
public double mean() {
double totalFraction = 0;
for (int i = 0; i < T; i++) {
totalFraction += fraction[i];
}
meanValue = totalFraction/T;
return meanValue;
} // sample mean of percolation threshold
public double stddev() {
double u = meanValue;
double sum = 0;
for (int i = 0; i < T; i++) {
sum += (fraction[i] - u) * (fraction[i] - u);
}
return Math.sqrt(sum/(T-1));
} // sample standard deviation of percolation threshold
public double confidenceLo() {
double u = meanValue;
double theta = this.stddev();
double sqrtT = Math.sqrt(T);
return u-1.96*theta/sqrtT;
} // low endpoint of 95% confidence interval
public double confidenceHi() {
double u = meanValue;
double theta = this.stddev();
double sqrtT = Math.sqrt(T);
return u+1.96*theta/sqrtT;
} // high endpoint of 95% confidence interval
public static void main(String[] args) {
int N = 200;
int T = 100;
if (args.length == 1) N = Integer.parseInt(args[0]);
else if (args.length == 2) {
N = Integer.parseInt(args[0]);
T = Integer.parseInt(args[1]); }
PercolationStats a = new PercolationStats(N, T);
System.out.print("mean = ");
System.out.println(a.mean());
System.out.print("stddev = ");
System.out.println(a.stddev());
System.out.print("95% confidence interval = ");
System.out.print(a.confidenceLo());
System.out.print(", ");
System.out.println(a.confidenceHi());
}
}