java - java 收银机的舍入误差

Question

我在这段代码上遇到了困难。当我输入 95.95 作为购买价格和 100 作为支付价格时，我不断收到舍入错误，例如，它返回 4 美元、4 便士，而不是 4 美元和 5 便士。所有其他值都工作正常，但由于某种原因，当小数位为 0.95 或 0.05 时，很难返回正确的便士...

public static void main(String[] args){

        //Read price per pack
        Scanner in = new Scanner(System.in);

        System.out.println("Enter purchase price: ");
        double purchasePrice = in.nextDouble();
        System.out.println("Enter amount paid: ");
        double amountPaid = in.nextDouble();

        //Compute change due

        if (amountPaid < purchasePrice)
        {System.out.printf("Error, you need more money!");
        }

        else {
        int diff = (int)(( amountPaid - purchasePrice)*100);


        int dollars = diff/100;
            diff = diff%100;

        int quarters = diff/25;
            diff = diff%25;

        int dimes = diff/10;
            diff = diff%10;

        int nickels = diff/5;
            diff = diff%5;

        int pennies = diff/1;
            diff = diff%1;

        System.out.println("--------------------");    
        System.out.println("Total Change due: " + (amountPaid - purchasePrice));
        System.out.println("--------------------");
        System.out.println("Dollars: " + dollars);
        System.out.println("Quarters: " + quarters);
        System.out.println("Dimes: " + dimes);
        System.out.println("Nickels: " + nickels);
        System.out.println("Pennies: " + pennies);

        }
    }
}

Answer 1

你的问题是线路

int diff = (int)(( amountPaid - purchasePrice)*100);

将美分数向下舍入到价格 100 倍以下的下一个 int。由于 double 不一定准确地表示十进制数，因此在进行乘法时可能会出现浮点错误，该错误会放大 100 倍。产品可能略高于或低于整数美分，因此四舍五入可能会导致金额向下舍入，几乎是整美分。

您真正想要做的是将美分数的 double 值四舍五入到最接近的 int，而不是下一个 int 下面。您可以通过在四舍五入之前添加 0.5 来做到这一点。

int diff = (int)(( amountPaid - purchasePrice)*100 + 0.5);

但一般来说，如果您想用它进行精确的计算，则永远不应该使用 double 来存储金额。我建议学习使用 BigDecimal 类，这样的问题就会消失。