我的随机数相关函数

不念不忘少年蓝@ 2021-07-24 13:39 322阅读 0赞

**一 代码**

package Myrandom;
    
    import java.util.Random;
    
    /**
     * Copyright (C), 2020-2020, XXX有限公司
     * FileName: Myrandom
     * Author:   cakin
     * Date:     2020/1/6
     * Description: 随机数相关函数
     */
    public class Myrandom {
        private static Random random;    // pseudo-random number generator
        private static long seed;
    
        static {
            // this is how the seed was set in Java 1.4
            seed = System.currentTimeMillis();
            random = new Random(seed);
        }
    
        private static void validateNotNull(Object x) {
            if (x == null) {
                throw new IllegalArgumentException("argument is null");
            }
        }
    
        public static double uniform() {
            return random.nextDouble();
        }
    
        // 随机返回[a,b)之间的一个double值
        public static double uniform( double a, double b ) {
            if (!(a < b)) {
                throw new IllegalArgumentException("invalid range: [" + a + ", " + b + ")");
            }
            return a + uniform() * (b - a);
        }
    
        // 随机返回[0..N)之间的一个int值
        public static int uniform(int n) {
            if (n <= 0) throw new IllegalArgumentException("argument must be positive: " + n);
            return random.nextInt(n);
        }
    
        // 返回[a,b)之间一个int值
        public static int uniform(int a, int b) {
            if ((b <= a) || ((long) b - a >= Integer.MAX_VALUE)) {
                throw new IllegalArgumentException("invalid range: [" + a + ", " + b + ")");
            }
            return a + uniform(b - a);
        }
    
        // 随机将double数组的元素排序
        public static void shuffle(double[] a) {
            validateNotNull(a);
            int n = a.length;
            for (int i = 0; i < n; i++) {
                int r = i + uniform(n-i);     // between i and n-1
                double temp = a[i];
                a[i] = a[r];
                a[r] = temp;
            }
        }
    
        // 概率分布函数
        public static int discrete(double[] probabilities) {
            if (probabilities == null) throw new IllegalArgumentException("argument array is null");
            double EPSILON = 1.0E-14;
            double sum = 0.0;
            for (int i = 0; i < probabilities.length; i++) {
                if (!(probabilities[i] >= 0.0))
                    throw new IllegalArgumentException("array entry " + i + " must be nonnegative: " + probabilities[i]);
                sum += probabilities[i];
            }
            if (sum > 1.0 + EPSILON || sum < 1.0 - EPSILON)
                throw new IllegalArgumentException("sum of array entries does not approximately equal 1.0: " + sum);
    
            //参考 https://www.jianshu.com/p/d51f8267e6b9
            while (true) {
                double r = uniform();
                sum = 0.0;
                for (int i = 0; i < probabilities.length; i++) {
                    sum = sum + probabilities[i];
                    if (sum > r) return i;
                }
            }
        }
    
        // 伯努利分布
        public static boolean bernoulli(double p) {
            if (!(p >= 0.0 && p <= 1.0))
                throw new IllegalArgumentException("probability p must be between 0.0 and 1.0: " + p);
            return uniform() < p;
        }
    
        // 正态分布，期望值为0，标准差为1
        // https://baijiahao.baidu.com/s?id=1630253104547104560&wfr=spider&for=pc
        public static double gaussian() {
            // use the polar form of the Box-Muller transform
            double r, x, y;
            do {
                x = uniform(-1.0, 1.0);
                y = uniform(-1.0, 1.0);
                r = x*x + y*y;
            } while (r >= 1 || r == 0);
            return x * Math.sqrt(-2 * Math.log(r) / r);
    
            // Remark:  y * Math.sqrt(-2 * Math.log(r) / r)
            // is an independent random gaussian
        }
    
        // 正态分布，期望值为mu，标准差为sigma
        public static double gaussian(double mu, double sigma) {
            return mu + sigma * gaussian();
        }
    
        public static void main( String[] args ) {
            System.out.println("uniform():"+uniform());
            System.out.println("uniform( double 1.0, double 4.0 ):"+uniform(1.0,4.0));
            System.out.println("uniform(int 5):"+uniform(5));
            System.out.println("uniform(int 1, int 7):"+uniform(1,7));
            double[] a = {1.8,3.4,5.6,4.3,8.7,2.3};
            shuffle(a);
            for(double m:a){
                System.out.print(m+" ");
            }
            System.out.println();
    
            double[] probabilities={0.3,0.3,0.4};
    
            System.out.println("discrete:"+discrete(probabilities));
            System.out.println("bernoulli():"+bernoulli(0.6));
            System.out.println("gaussian():"+gaussian());
            System.out.println("gaussian(double mu, double sigma):"+gaussian(0,1));
        }
    }

**二 测试结果**

uniform():0.8752046669179311
    uniform( double 1.0, double 4.0 ):2.282136762532514
    uniform(int 5):3
    uniform(int 1, int 7):6
    4.3 3.4 2.3 1.8 5.6 8.7 
    discrete:0
    bernoulli():false
    gaussian():-1.4711750721748598
    gaussian(double mu, double sigma):-1.608130645331619