BigDecimal.setScale()方法用于商业计算的精度设置问题详解

BigDecimal.setScale()方法用于商业计算的精度设置问题详解

    网上的说法繁杂，看起来诸多不便，并且有的说法也不太准确，在这里 做一下求证。用的较多的4个参数：

BigDecimal.ROUND_UP 这个方法做直接进位处理，但是有个特殊情况需要考虑，如果将要舍弃的位上的值是0，不做进位处理。

BigDecimal.ROUND_DOWN 这个方法做直接舍弃处理

BigDecimal.ROUND_HALF_UP 这个方法就是大名鼎鼎的四舍五入
BigDecimal.ROUND_HALF_DOWN 这个方法准确来说，是五舍六入

import java.math.BigDecimal;

/**
*
* @author yangcq
* Java常用商业计算工具类BigDecimal的精度设置方法setScale()详解
*
*/
public class BigDecimalTest {

public static void main(String[] args) {
bigDecimalTest();
}

private static void bigDecimalTest(){
BigDecimal bigDecimal_01 = new BigDecimal("918.845");
BigDecimal bigDecimal_02 = new BigDecimal("918.844");
BigDecimal bigDecimal_03 = new BigDecimal("918.846");

bigDecimal_01.setScale(3);
bigDecimal_01.setScale(2,BigDecimal.ROUND_UP);
bigDecimal_01.setScale(2,BigDecimal.ROUND_DOWN);
bigDecimal_01.setScale(2,BigDecimal.ROUND_HALF_UP);
bigDecimal_01.setScale(2,BigDecimal.ROUND_HALF_DOWN);

System.out.println("测试数据：" + bigDecimal_01);
System.out.println("BigDecimal.ROUND_UP = " + bigDecimal_01.setScale(2,BigDecimal.ROUND_UP));
System.out.println("BigDecimal.ROUND_DOWN = " + bigDecimal_01.setScale(2,BigDecimal.ROUND_DOWN));
System.out.println("BigDecimal.ROUND_HALF_UP = " + bigDecimal_01.setScale(2,BigDecimal.ROUND_HALF_UP));
System.out.println("BigDecimal.ROUND_HALF_DOWN = " + bigDecimal_01.setScale(2,BigDecimal.ROUND_HALF_DOWN));

System.out.println("测试数据：" + bigDecimal_02);
System.out.println("BigDecimal.ROUND_UP = " + bigDecimal_02.setScale(2,BigDecimal.ROUND_UP));
System.out.println("BigDecimal.ROUND_DOWN = " + bigDecimal_02.setScale(2,BigDecimal.ROUND_DOWN));
System.out.println("BigDecimal.ROUND_HALF_UP = " + bigDecimal_02.setScale(2,BigDecimal.ROUND_HALF_UP));
System.out.println("BigDecimal.ROUND_HALF_DOWN = " + bigDecimal_02.setScale(2,BigDecimal.ROUND_HALF_DOWN));

System.out.println("测试数据：" + bigDecimal_03);
System.out.println("BigDecimal.ROUND_UP = " + bigDecimal_03.setScale(2,BigDecimal.ROUND_UP));
System.out.println("BigDecimal.ROUND_DOWN = " + bigDecimal_03.setScale(2,BigDecimal.ROUND_DOWN));
System.out.println("BigDecimal.ROUND_HALF_UP = " + bigDecimal_03.setScale(2,BigDecimal.ROUND_HALF_UP));
System.out.println("BigDecimal.ROUND_HALF_DOWN = " + bigDecimal_03.setScale(2,BigDecimal.ROUND_HALF_DOWN));

// BigDecimal.ROUND_UP 这个方法做直接进位处理，但是有个特殊情况需要考虑，如果将要舍弃的位上的值是0，不做进位处理。
BigDecimal bigDecimal_04 = new BigDecimal("918.850");
System.out.println("测试数据：" + bigDecimal_04);
System.out.println("BigDecimal.ROUND_UP = " + bigDecimal_04.setScale(2,BigDecimal.ROUND_UP));
System.out.println("BigDecimal.ROUND_DOWN = " + bigDecimal_04.setScale(2,BigDecimal.ROUND_DOWN));
System.out.println("BigDecimal.ROUND_HALF_UP = " + bigDecimal_04.setScale(2,BigDecimal.ROUND_HALF_UP));
System.out.println("BigDecimal.ROUND_HALF_DOWN = " + bigDecimal_04.setScale(2,BigDecimal.ROUND_HALF_DOWN));
}
}

-------------------------------------控制台输出：----------------------------------------------
测试数据：918.845

BigDecimal.ROUND_UP = 918.85

BigDecimal.ROUND_DOWN = 918.84

BigDecimal.ROUND_HALF_UP = 918.85

BigDecimal.ROUND_HALF_DOWN = 918.84

测试数据：918.844

BigDecimal.ROUND_UP = 918.85

BigDecimal.ROUND_DOWN = 918.84

BigDecimal.ROUND_HALF_UP = 918.84

BigDecimal.ROUND_HALF_DOWN = 918.84

测试数据：918.846

BigDecimal.ROUND_UP = 918.85

BigDecimal.ROUND_DOWN = 918.84

BigDecimal.ROUND_HALF_UP = 918.85

BigDecimal.ROUND_HALF_DOWN = 918.85

测试数据：918.850

BigDecimal.ROUND_UP = 918.85

BigDecimal.ROUND_DOWN = 918.85

BigDecimal.ROUND_HALF_UP = 918.85

BigDecimal.ROUND_HALF_DOWN = 918.85