Skip to content

Latest commit

 

History

History
119 lines (83 loc) · 4.75 KB

perl.md

File metadata and controls

119 lines (83 loc) · 4.75 KB
title description
Floating-point cheat sheet for Perl
Tips for using floating-point and decimal numbers in Perl

Floating-Point Types

Perl supports platform-native floating-point as scalar values; in practice this usually means IEEE 754 double precision.

Exact Types

Perl can also store decimal numbers as strings, but the builtin arithmetic operators will convert them to integer or floating-point values to perform the operation.

The Math::BigFloat extension provides an arbitrary-precision decimal type:

	use Math::BigFloat ':constant'
	my $f = 0.1 + 0.2; # returns exactly 0.3

The Number::Fraction extension provides a fraction type that overloads the arithmetic operators with symbolic fraction arithmetic:

	use Number::Fraction ':constants';
	my $f = '1/2' - '1/3'; # returns 1/6

The Math::BigRat extension provides similar functionality. Its advantage is compatibility with the Math::BigInt and Math::BigFloat extensions, but it does not seem to support fraction literals.

How to Round

To get a string:

	$result = sprintf("%.2f", 1.2345); # returns 1.23

To format output:

	printf("%.2f", 1.2); # prints 1.20

Note that this implicitly uses round-to-even. The variable $# contains the default format for printing numbers, but its use is considered deprecated.

The Math::Round extension provides various functions for rounding floating-point values:

	use Math::Round qw(:all);
	$result = nearest(.1, 4.567) # prints 4.6
	$result = nearest(.01, 4.567) # prints 4.57

The Math::BigFloat extension also supports various rounding modes:

	use Math::BigFloat;
	my $n = Math::BigFloat->new(123.455);
	my $f1 = $n->round('','-2','common'); # returns 123.46
	my $f2 = $n->round('','-2','zero'); # returns 123.45

Resources

title: 为 Perl 准备的浮点小抄 description: Tips for using floating-point and decimal numbers in Perl

浮点类型

Perl 支持作为标量的平台原生浮点；实际上就是指 IEEE 754 双精度。

准确类型

Perl 可以用字符串的方式存储小数，但内置的算数运算会把它们转换成整数或者浮点数来进行运算。

Math::BigFloat 扩展提供了 arbitrary-precision 小数类型：

	use Math::BigFloat ':constant'
	my $f = 0.1 + 0.2; # 返回 exactly 0.3

Number::Fraction 扩展提供了分数类型，它用分数算数里的符号重载了算数运算符：

	use Number::Fraction ':constants';
	my $f = '1/2' - '1/3'; # 返回 1/6

Math::BigRat 扩展提供了类似的功能。它的优点在于兼容 Math::BigInt 和 Math::BigFloat 扩展，但它好像不支持分数字面量。

如何取整

得到字符串：

	$result = sprintf("%.2f", 1.2345); # 返回 1.23

如何格式化输出：

	printf("%.2f", 1.2); # 打印 1.20

注意，这个实现使用 round-to-even。 变量 $# 存有打印数字的缺省格式，但这不推荐使用。

Math::Round 扩展提供不同的不同的函数来取整浮点值：

	use Math::Round qw(:all);
	$result = nearest(.1, 4.567) # 打印 4.6
	$result = nearest(.01, 4.567) # 打印 4.57

Math::BigFloat 扩展也支持不同的取整模式：

	use Math::BigFloat;
	my $n = Math::BigFloat->new(123.455);
	my $f1 = $n->round('','-2','common'); # 返回 123.46
	my $f2 = $n->round('','-2','zero'); # 返回 123.45

资源