详解iOS之关于double/float数据计算精度问题
1、我们的app有一个判断,当用户输入的值,小于等于剩余的余额时,给用户弹窗,代码不往下执行
当用户输入 0.01 时,po 一下的结果是:0.10000.... ,当用户的余额是0.01时,po一下网络请求的返回结果是:0.009999...
当用户输入 66.05时,po 一下的结果是:66.049999...,当用户的余额是66.05时,po一下网络请求的返回结果是:66.049999...
所以,当数据较大时,不会有影响,因为用户输入的数据和网络返回数据都被变成了不精确的小值了;但是,当数据较小时,用户输入的数据没有被转,网络返回的数据被转成了不精确的小值,导致用户输入永远大于网络返回的数据,代码无法往下执行;
解决:使用 NSDecimalNumber 将float、double数据转化成NSDecimalNumber 类型的对象进行 +、-、*、/计算,再取值。
- (double)DecimalNumber:(double)num1 num2:(double)num2 {
NSDecimalNumber *n1 = [NSDecimalNumber decimalNumberWithString:[NSString stringWithFormat:@"%f",num1]];
NSDecimalNumber *n2 = [NSDecimalNumber decimalNumberWithString:[NSString stringWithFormat:@"%f",num2]];
NSDecimalNumber *n3 = [n1 decimalNumberBySubtracting:n2];
return n3.doubleValue;
}
// 调用:
double result = [self DecimalNumber: 2.01 num2: 2]; // 结果为 0.01
其中:
decimalNumberBySubtracting: n1 - n2,返回 n1 - n2 的结果值
decimalNumberByMultiplyingBy:n1 * n2,
decimalNumberByDividingBy:n1 / n2,
decimalNumberByAdding:n1 + n2,
compare:比较 n1 和 n2,返回比较结果;
2、进行数据的四舍五入获得结果
- (NSString*)Rounding:(float)number afterPoint:(NSInteger)position
{
NSDecimalNumberHandler *handler = [NSDecimalNumberHandler decimalNumberHandlerWithRoundingMode: NSRoundPlain scale: position raiseOnExactness: NO raiseOnOverflow: NO raiseOnUnderflow:NO raiseOnDivideByZero: NO];
NSDecimalNumber *floatDecimal = [[NSDecimalNumber alloc] initWithFloat: number];
NSDecimalNumber *resultNumber = [floatDecimal decimalNumberByRoundingAccordingToBehavior:handler];
return [NSString stringWithFormat:@"%@",resultNumber];
}
// 调用:
NSString *result = [self Rounding:8.00092 afterPoint:3]; // 结果为 8.001
其中,参数number是需要四舍五入的数据,position是小数点后保留的位数;
高精度要求
使用系统提供的NSDecimalNumber API进行计算,最后转换为字符串输出显示。NSDecimalNumber转换方法:
#import "NSDecimalNumber+Y_Add.h"
@implementation NSDecimalNumber (Y_Add)
+ (NSDecimalNumber *)y_decimalNumberWithFloat:(float)value{
return [self y_decimalNumberWithFloat:value scale:2];
}
+ (NSDecimalNumber *)y_decimalNumberWithFloat:(float)value scale:(short)scale{
return [self y_decimalNumberWithFloat:value roundingMode:NSRoundBankers scale:scale];
}
+ (NSDecimalNumber *)y_decimalNumberWithFloat:(float)value roundingMode:(NSRoundingMode)roundingMode scale:(short)scale{
return [[[NSDecimalNumber alloc] initWithFloat:value] y_decimalNumberHandlerWithRoundingMode:roundingMode scale:scale];
}
+ (NSDecimalNumber *)y_decimalNumberWithDouble:(double)value{
return [self y_decimalNumberWithDouble:value scale:2];
}
+ (NSDecimalNumber *)y_decimalNumberWithDouble:(double)value scale:(short)scale{
return [self y_decimalNumberWithDouble:value roundingMode:NSRoundBankers scale:scale];
}
+ (NSDecimalNumber *)y_decimalNumberWithDouble:(double)value roundingMode:(NSRoundingMode)roundingMode scale:(short)scale{
return [[[NSDecimalNumber alloc] initWithFloat:value] y_decimalNumberHandlerWithRoundingMode:roundingMode scale:scale];
}
/**
* <#Description#>
*
* @return <#return value description#>
*/
- (NSDecimalNumber *)y_decimalNumberHandler{
return [self y_decimalNumberHandlerWithRoundingMode:NSRoundBankers scale:2];
}
- (NSDecimalNumber *)y_decimalNumberHandlerWithRoundingMode:(NSRoundingMode)roundingMode scale:(short)scale{
NSDecimalNumberHandler *handler = [NSDecimalNumberHandler decimalNumberHandlerWithRoundingMode:roundingMode
scale:scale
raiseOnExactness:NO
raiseOnOverflow:YES
raiseOnUnderflow:YES
raiseOnDivideByZero:YES];
return [self decimalNumberByRoundingAccordingToBehavior:handler];
}
@end
普通精度要求
使用系统提供的数学运算进行计算,最后转换为字符串输出显示。保留小数点后N位的方法:
/**
* formatterNumber .00 小数点后两位
*
* @param number <#number description#>
*
* @return <#return value description#>
*/
+ (NSString *)y_formatterNumber:(NSNumber *)number{
return [self y_formatterNumber:number fractionDigits:2];
}
+ (NSString *)y_formatterNumber:(NSNumber *)number fractionDigits:(NSUInteger)fractionDigits{
NSNumberFormatter *numberFormatter = [[NSNumberFormatter alloc] init];
[numberFormatter setMaximumFractionDigits:fractionDigits];
[numberFormatter setMinimumFractionDigits:fractionDigits];
return [numberFormatter stringFromNumber:number];
}
以上是 详解iOS之关于double/float数据计算精度问题 的全部内容, 来源链接: utcz.com/z/342376.html