Project

General

Profile

Actions

Backport #3803

closed

BigDecimal::ROUND_HALF_DOWN/ROUND_HALF_EVEN behave incorrectly (disagree with JRuby and the rest of the world)

Added by matthw (Matthew Willson) almost 12 years ago. Updated over 10 years ago.

Status:
Closed
Priority:
Normal
[ruby-core:32136]

Description

=begin
The incorrect behaviour is that all fractional values between 0.5 (inclusive) and 0.6 (non-inclusive) are subject to the rounding policy for 'half', whereas it should only be applied for fractional values exactly equal to 0.5.

This means that, for example, 0.59 is rounded down to 0 by ROUND_HALF_DOWN and ROUND_HALF_EVEN:

 >> BigDecimal.new("0.59").round(0, BigDecimal::ROUND_HALF_DOWN).to_s('F')
 => "0.0"
 >> BigDecimal.new("0.56").round(0, BigDecimal::ROUND_HALF_EVEN).to_s('F')
 => "0.0"

The behaviour is actually specified this way in the RDoc:

ROUND_HALF_DOWN: round up if the appropriate digit >= 6, otherwise truncate

But this is most definitely not what most people would expect from 'round half down' (see eg: http://en.wikipedia.org/wiki/Rounding#Round_half_down ). I would expect it to only round down if the fractional part is less than or exactly equal to 0.5. It needs to take into account more than just the first decimal digit to do this; while the first decimal digit of 0.59 may be 5, it is not equal to 0.5 and hence should not be subject to the 'half down' rounding policy. (Note that ROUND_HALF_UP still works correctly, since the correct behaviour here can indeed be implemented by only inspecting the first decimal digit; but this is not the case for ROUND_HALF_DOWN or in general for ROUND_HALF_EVEN)

Rather misleadingly, ROUND_HALF_EVEN is specified as being "banker's rounding":

ROUND_HALF_EVEN: round towards the even neighbor (Banker’s rounding)

But since it incorrectly applies the 'half towards even' policy not just to 0.5, but to all values in the [0.5, 0.6) interval, it is not "banker's rounding" as this is usually defined, and is surely a biased rounding policy, which is exactly what banker's rounding is seeking to avoid (see eg http://en.wikipedia.org/wiki/Rounding#Round_half_to_even )

I note (as just one example) that Java's BigDecimal specifies and implements ROUND_HALF_DOWN in the correct way: http://download-llnw.oracle.com/javase/6/docs/api/java/math/BigDecimal.html#ROUND_HALF_DOWN

Rounding mode to round towards "nearest neighbor" unless both neighbors are equidistant, in which case round down.

And that JRuby's BigDecimal appears to be based on this, hence is correct and hence disagrees with MRI when (for example) applying ROUND_HALF_DOWN to 0.59:

 irb(main):004:0> RUBY_PLATFORM
 => "java"
 irb(main):005:0> BigDecimal.new("0.59").round(0, BigDecimal::ROUND_HALF_DOWN).to_s('F')
 => "1.0"

Backport to 1.8.7 would be nice if this gets fixed.
=end


Files

ruby_bigdecimal_round_half.patch (3.88 KB) ruby_bigdecimal_round_half.patch Minimal patch with test and docfix matthw (Matthew Willson), 09/15/2010 10:56 AM
ruby_bigdecimal_round_half_with_comments.patch (8.35 KB) ruby_bigdecimal_round_half_with_comments.patch Patch with some extra comments which helped me figure out what's going on matthw (Matthew Willson), 09/15/2010 10:56 AM

Related issues 1 (0 open1 closed)

Has duplicate Ruby master - Bug #4567: BigDecimal::ROUND_HALF_DOWNClosedmrkn (Kenta Murata)04/11/2011Actions
Actions #1

Updated by shyouhei (Shyouhei Urabe) almost 12 years ago

  • Status changed from Open to Assigned
  • Assignee set to mrkn (Kenta Murata)

=begin

=end

Actions #2

Updated by mrkn (Kenta Murata) almost 12 years ago

  • Category set to ext
  • Target version set to 1.9.3

=begin
This is bugs of BigDecimal, which have not been noticed for a long time.
Thank you very much for your report.

While I will correct these bugs in a few days,
if you have appropriate patches, I want to review and incorporate them.

=end

Actions #4

Updated by matthw (Matthew Willson) almost 12 years ago

=begin
Thanks Kenta, yeah that would be great if you get a moment to review the attached patch. I included a regression test and a fix to the rdoc comments. It fixes the issue for me.

The version with comments includes some comments I added to help me figure out what's going on there, the other patch is more minimal.
=end

Actions #5

Updated by matthw (Matthew Willson) almost 12 years ago

=begin
(I should say the patch is against SVN trunk)
=end

Actions #6

Updated by mrkn (Kenta Murata) almost 12 years ago

=begin
Thank you for attaching your patches.
I'll review them tonight.

=end

Actions #7

Updated by mrkn (Kenta Murata) almost 12 years ago

  • Status changed from Assigned to Closed
  • % Done changed from 0 to 100

=begin
This issue was solved with changeset r29293.
Matthew, thank you for reporting this issue.
Your contribution to Ruby is greatly appreciated.
May Ruby be with you.

=end

Actions #8

Updated by semmons99 (Shane Emmons) over 11 years ago

=begin
Why isn't this being back-ported into 1.8.7 and 1.9.2? This is a rather critical bug to anyone working with financial transactions. It seems like this is an obvious bug that should be fixed in all currently maintained branches.
=end

Actions #9

Updated by mrkn (Kenta Murata) over 11 years ago

  • Tracker changed from Bug to Backport
  • Status changed from Closed to Assigned
  • Assignee changed from mrkn (Kenta Murata) to yugui (Yuki Sonoda)

=begin

=end

Updated by mrkn (Kenta Murata) over 11 years ago

  • Target version changed from 1.9.3 to 1.9.2

=begin

=end

Actions #11

Updated by mrkn (Kenta Murata) almost 11 years ago

  • Status changed from Assigned to Closed

This issue was solved with changeset r29293.
Matthew, thank you for reporting this issue.
Your contribution to Ruby is greatly appreciated.
May Ruby be with you.


  • ext/bigdecimal/bigdecimal.c: fix rounding algorithms for half-down
    and half-even. This change is based on the patch created by Matthew
    Willson, the reporter of this bug. [Bug #3803] [ruby-core:32136]
  • test/bigdecimal/test_bigdecimal.rb: add tests for above changes.
Actions #12

Updated by naruse (Yui NARUSE) over 10 years ago

  • Target version deleted (1.9.2)
Actions

Also available in: Atom PDF