Continued fractions can be used to find rational approximations to real numbers that are "best" in a strict sense. Here's a PHP function that finds a rational approximation to a given (positive) floating point number with a relative error less than $tolerance
:
<?php
function float2rat($n, $tolerance = 1.e-6) {
$h1=1; $h2=0;
$k1=0; $k2=1;
$b = 1/$n;
do {
$b = 1/$b;
$a = floor($b);
$aux = $h1; $h1 = $a*$h1+$h2; $h2 = $aux;
$aux = $k1; $k1 = $a*$k1+$k2; $k2 = $aux;
$b = $b-$a;
} while (abs($n-$h1/$k1) > $n*$tolerance);
return "$h1/$k1";
}
printf("%s
", float2rat(66.66667)); # 200/3
printf("%s
", float2rat(sqrt(2))); # 1393/985
printf("%s
", float2rat(0.43212)); # 748/1731
I have written more about this algorithm and why it works, and even a JavaScript demo here: https://web.archive.org/web/20180731235708/http://jonisalonen.com/2012/converting-decimal-numbers-to-ratios/
与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…