From: Martin DeMello <martindemello@...>
Date: 2006-12-03T04:51:37+09:00
Subject: Re: convert binary array to floating point decimal

On 12/3/06, Nebiru <Nebiru@gmail.com> wrote:
> new to ruby here, looking for a little help
>
> I'm trying to convert some C# code to ruby, however I'm not getting the
> desired results
> The code is supposed to convert a binary string to a floating point
> number
>
> public float
(int length)
> {
>    float result = 0;
>    int num = 0;
>    // form the integer number from the binary string
>    for (int i = 0; i < length; i++)
>       {
>       num = num << 1;
>       num = num + binaryArray[(length - i) - 1];
>       }
>    // scale to the floating point value
>    result = (float)num;
>    result = result / (1 << binaryArray.length/2);
>    return result;
> }
>
> any help will be appreciated

Ruby has a built in method for converting a binary string into an
integer, so your code is simply

def binaryStringToFloat(str)
  str.to_i(2) * 1.0/str.length
end

but if you want to translate your algorithm literally:

def binaryStringToFloat(str)
  num = 0
  # reverse the string, split it into characters, then iterate over
the characters
  str.reverse.split(//).each {|digit|
    num = (num << 1) + digit.to_i  # << has lower precedence than +
  }

  return num * 1.0 / (1 << str.length)
end

also, the pattern

accumulator = start_val
collection.each {|element|
  accumulator = f(accumulator, element)
}

is captured by the inject method, so we can replace the "each" loop with

num = str.reverse.split(//).inject(0) {|sum, digit| (sum << 1) + digit .to_i}

martin