[R-SIG-Mac] Unexpected behavior of R round function on Mac

[Duncan Murdoch]

On 11/04/2014, 5:04 AM, peter dalgaard wrote:
> Google for "round to even", and read the Details section of ?round.

The behaviour on the Linux laptop does sound problematic, if it is 
really giving round(4.5) = 5.  We do have the warning there that it 
depends on OS services, but this could cause other problems.

Duncan Murdoch


>
>
> On 11 Apr 2014, at 10:42 , Per Nielsen <evilper at gmail.com> wrote:
>
>> Hi all
>>
>> I have encountered some strange/wrong behavior of the round function in R.
>>
>> This is the behavior on my mac laptop:
>>
>>> R.version
>>                _
>> platform       x86_64-apple-darwin13.0.2
>> arch           x86_64
>> os             darwin13.0.2
>> system         x86_64, darwin13.0.2
>> status
>> major          3
>> minor          0.2
>> year           2013
>> month          09
>> day            25
>> svn rev        63987
>> language       R
>> version.string R version 3.0.2 (2013-09-25)
>> nickname       Frisbee Sailing
>>> round(1.5)
>> [1] 2
>>> round(4.5)
>> [1] 4
>>> floor(1.5 + 0.5)
>> [1] 2
>>> floor(4.5 + 0.5)
>> [1] 5
>>
>> This is the behavior on Ubuntu 12.04 running in a virtual machine on the
>> same mac:
>>
>>> R.version
>>                _
>> platform       x86_64-pc-linux-gnu
>> arch           x86_64
>> os             linux-gnu
>> system         x86_64, linux-gnu
>> status         beta
>> major          3
>> minor          1.0
>> year           2014
>> month          03
>> day            28
>> svn rev        65330
>> language       R
>> version.string R version 3.1.0 beta (2014-03-28 r65330)
>> nickname       Spring Dance
>>> round(1.5)
>> [1] 2
>>> round(4.5)
>> [1] 4
>>> floor(1.5 + 0.5)
>> [1] 2
>>> floor(4.5 + 0.5)
>> [1] 5
>>
>> I would expect round(x.5) = x+1, which only happens sometimes. My colleague
>> using R version 3.1.0 on Linux on a Lenovo laptop gets the expected result
>> in all cases.
>>
>> Is this a bug or somehow expected behavior? It seems to be related to the
>> mac hardware. I can ad hoc fix it using floor, but that is hardly optimal.
>>
>> Thanks,
>> Per
>>
>> 	[[alternative HTML version deleted]]
>>
>> _______________________________________________
>> R-SIG-Mac mailing list
>> R-SIG-Mac at r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>