How do I preserve my float number in ruby

Question

So I'm trying some code out to convert numbers into strings. However, I noticed that in certain cases it does not preserve the last two decimal places. For instance I type 1.01 and 1.04 for addition and I get back 2.04. If I type just 1.05 it preserves the number and returns it exactly. I get whats going on things are being rounded. I don't know how to prevent it from being rounded though. Should I just consider sending (1.01+1.04) to self as only one input?

Warning! I haven't tried this yet so don't know if its supported:

 user_input = (1.04+1.01) #entry from user
 user_input = gets.to_f
 user_input.to_test_string

What I have so far:

    class Float
     def to_test_string

      cents = self % 1
      dollars = self - cents
      cents = cents * 100

      text = "#{dollars.to_i.en.numwords} dollars and #{cents.to_i.en.numwords} cents"

      puts text
      text
     end
    end
  puts "Enter two great floating point numbers for adding"
  puts "First number"
  c = gets.to_f
  puts "Second number"
  d = gets.to_f
  e = c+d
  puts e.to_test_string
  puts "Enter a great floating number! Example 10.34"
  a = gets.to_f 
  puts a.to_test_string

Thanks for the help! Post some code up so I can try!

Are `en` and `numwords` Ruby methods, or something from Rails' ActiveSupport? — Andrew Grimm, Nov 07 '10 at 22:11
@Andrew Grimm, in my honest opinion, it must not to be and will never be in the core or stdlib of Ruby. — Nakilon, Nov 07 '10 at 22:20
@Nakilon: I agree. I was only asking because the question was originally tagged only `ruby` and not `ruby-on-rails`. — Andrew Grimm, Nov 07 '10 at 22:38
en and numwords are from the lingusitics gem. Which for some reason I didnt put into my code. — Matt, Nov 07 '10 at 22:43

score 2 · Accepted Answer · edited May 23 '17 at 11:47

2

First of all: never use float for money — Use Float or Decimal for Accounting Application Dollar Amount?

irb> x = 1.01 + 1.04
=> 2.05
irb> y = x % 1
=> 0.04999999999999982
irb> (y * 100).to_i
=> 4

But if want it VERYVERYVERY much:

irb> (y * 100).round.to_i
=> 5

edited May 23 '17 at 11:47

Community

1
1

answered Nov 07 '10 at 22:03

Nakilon

32,203
13
95
132

score 2 · Answer 2 · answered Nov 07 '10 at 22:12

$ python
Python 2.6.2 (release26-maint, Apr 19 2009, 01:56:41)
[GCC 4.3.3] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> 1.04+1.01
2.0499999999999998

Read this: What Every Computer Scientist Should Know About Floating-Point Arithmetic

Also, what Nakilon said.

score 1 · Answer 3 · answered Nov 07 '10 at 22:16

This is not a problem with ruby, nor your code (although you need to get rid of .en.numwords); it is a problem with the binary floating point representation.

You should use a Fixnum or Bignum to represent the currency.

eg.

class Currency
    def initialize str
        unless str =~ /([0-9]+)\.([0-9]{2})/
            raise 'invalid currency string'
        end
        @cents = $1.to_i * 100 + $2.to_i
    end
end

How do I preserve my float number in ruby

3 Answers3

Linked