工具使用训练与长链路信用分配

import java.util.*;
public class Main {
    static void near(double expected,double actual) {
        if (!Double.isFinite(actual) || Math.abs(expected-actual)>1e-9*Math.max(1,Math.abs(expected)))
            throw new AssertionError(expected+" != "+actual);
    }
    static void check(boolean ok) { if(!ok) throw new AssertionError(); }
    static void bad(Runnable r) {
        try { r.run(); } catch(IllegalArgumentException e) { return; }
        throw new AssertionError("invalid input accepted");
    }
record Result(double original,double shaped,double boundary) {}
    static Result shape(double[] rewards,double[] potential,double gamma) {
        if(rewards==null||potential==null||potential.length!=rewards.length+1||!Double.isFinite(gamma)||gamma<0||gamma>1)
            throw new IllegalArgumentException("shape");
        for(double x:potential) if(!Double.isFinite(x)) throw new IllegalArgumentException("potential");
        double original=0,shaped=0,discount=1;
        for(int t=0;t<rewards.length;t++) {
            if(!Double.isFinite(rewards[t])) throw new IllegalArgumentException("reward");
            double adjusted=rewards[t]+gamma*potential[t+1]-potential[t]; // 相邻势函数必须使用同一个折扣。
            original+=discount*rewards[t];
            shaped+=discount*adjusted; // 先按当前时刻加权，再推进折扣指数。
            discount*=gamma;
            if(!Double.isFinite(original)||!Double.isFinite(shaped)) throw new IllegalArgumentException("overflow");
        }
        double boundary=-potential[0]+discount*potential[rewards.length]; // 截断时保留末端项，不能假装终止。
        if(!Double.isFinite(boundary)) throw new IllegalArgumentException("overflow");
        return new Result(original,shaped,boundary);
    }
    public static void main(String[] args) {
Result a=shape(new double[]{0,0,1},new double[]{2,3,4,0},1);
        near(1,a.original());near(-1,a.shaped());
        Result b=shape(new double[]{0,0,1},new double[]{2,3,4,0},.9);
        near(b.original()+b.boundary(),b.shaped());
        Result c=shape(new double[]{0,0,1},new double[]{2,3,4,5},.9);
        near(-2+Math.pow(.9,3)*5,c.boundary());near(c.original()+c.boundary(),c.shaped());
        near(0,shape(new double[]{},new double[]{7},.9).shaped());
        near(3,shape(new double[]{5,9},new double[]{2,8,0},0).shaped());
        bad(()->shape(new double[]{1},new double[]{0},1));
        bad(()->shape(new double[]{1},new double[]{0,0},1.1));
        bad(()->shape(new double[]{Double.NaN},new double[]{0,0},.9));
        System.out.println("PASS 8 cases");
    }
}