Calculation and control method of floating-point precision for JAVA double or float __java

This article describes how to compute the precision of floating-point numbers in double or float in Java, providing a variety of parameters in Java to achieve the different control of precision. Specific examples are as follows:

* * Created on 2005-6-5 * Author Stephen * Email zhoujianqiang at gmail DOT com * CopyRight (C) 2005-2008, All rights Rese RVed. * * Package com.soft4j.utility; Import Java.math.BigDecimal; /** * Some commonly used tool methods related to decimal precision (rounding, etc.). * * Float/double precision values are divided into the following: <br> * Java.math.BigDecimal.ROUND_UP <br> * Java.math.BigDecimal.ROUND_DOWN <br> * Java.math.BigDecimal.ROUND_CEILING <br> * Java.math.BigDecimal.ROUND_FLOOR <br> * java.math.bigdecimal.round_half_up<br> * Java.math.BigDecimal.ROUND_HALF_DOWN <br> * Java.math.BigDecimal.ROUND_HALF_EVEN <br> * * @author Stephen * @version 1.0.0/Public final class Roundtool {/** * Accuracy of double data. * <p> * For example: <br> * Double value = 100.345678; <br> DOUBLE ret = round (VALUE,4,BIGDECIMAL.ROUND_HALF_UP); <br> * ret for 100.3457 <br> * * @param value * Double data. * @param scale * precision digits (number of decimal places to keep). * @param Roundingmode * Precision value method. * @return The data after the precision calculation. */public static double rounduble value, int scale, int roundingmode) {BigDecimal bd = new BigDecimal (value); BD = Bd.setscale (scale, roundingmode); d Ouble d = bd.doublevalue (); BD = NULL; return D; /** * The Main method used for testing. * * @param argc * Run parameters. */public static void Main (string[] argc) {//below to retain 2 decimal places as an example//round_up///As long as a decimal number greater than 0 exists after the 2nd digit, the 2nd digit is +1 System.out.println (ROUND (12.3401,2,BIGDECIMAL.ROUND_UP)); /12.35 System.out.println (Round ( -12.3401,2,bigdecimal.round_up));//-12.35//round_down//round_up opposite// Discard all decimal System.out.println (round (12.349,2,bigdecimal.round_down)) immediately following the 2nd digit,//12.34 System.out.println (- 12.349,2,bigdecimal.round_down))//-12.34//round_ceiling//If the number >0 is the same as the round_up effect if the number <0 is the same as the Round_down function System.out.println (Round (12.3401,2,bigdecimal.round_ceiling));//12.35 System.out.println (Round ( -12.349,2, bigdecimal.round_ceiling))//-12.34//round_floor//If the number >0 is the same as the Round_down effect if the number <0 is the same as the ROUND_UP function System.out.println (Round (12.349,2,bigdecimal.round_floor));//12.34 System.out.println (Round ( -12).3401,2,bigdecimal.round_floor));//-12.35//round_half_up [most commonly used for this method]//If the 3rd digit >=5, then the 2nd digit +1//NOTE: Look at the value of the 3rd digit only, System.out.println (Round (12.345,2,BIGDECIMAL.ROUND_HALF_UP)) that do not consider decimals after the 3rd digit,//12.35 System.out.println ( 12.3449,2,BIGDECIMAL.ROUND_HALF_UP));//12.34 System.out.println (ROUND ( -12.345,2,BIGDECIMAL.ROUND_HALF_UP));//- 12.35 System.out.println (Round ( -12.3449,2,bigdecimal.round_half_up))//-12.34//round_half_down//If the 3rd digit >=5, Do round_up///If the 3rd digit <5, then do Round_down System.out.println (round (12.345,2,bigdecimal.round_half_down));//12.35 System.out.println (Round (12.3449,2,bigdecimal.round_half_down));//12.34 System.out.println (Round ( -12.345,2, Bigdecimal.round_half_down));//-12.35 System.out.println (ROUND ( -12.3449,2,bigdecimal.round_half_down));//-12.34 Round_half_even//If the 3rd digit is an even number, do round_half_down//If the 3rd digit is odd, do round_half_up System.out.println (ROUND, Bigdecimal.round_half_even));//12.35 System.out.println (ROUND (12.345,2,bigdecimal.round_half_even));//12.35}}