[Rd] bug in modulus operator %% (PR#7852)

[Kjetil Brinchmann Halvorsen]

Prof Brian Ripley wrote:

> I've now found a Windows system that does this.  This is also Windows 
> XP, fully patched, and with the same rw2010.  So it may be 
> chip-specific: the one that works is a P4 and the one that does not is 
> a latest Pentium M.
>
> I am not sure that the guarantee on the help page has been supported 
> for a while.  I've altered the code so it is more likely to be.
>
> BTW,
>
>> options(digits=20)
>> 1 %% 0.001
>
> [1] 0.0009999999999999792
>
> shows why the original is not a bug in R.
>
Why is that not a bug in R? On my machine (windows XP, rw2010 from CRAN) 
I get:

test <- function(x, y) (x %% y) + y * ( x %/% y ) # should be x
 > test(1, 0.001)
[1] 1.001
 > test(1, 0.1)
[1] 1.1
 > test(1, 1)
[1] 1
 > test(1, 0.01)
[1] 1.01

and this differences (well, not the third one) cannot be said to be 
rounding
error.

Kjetil

>
> On Thu, 12 May 2005 ripley at stats.ox.ac.uk wrote:
>
>> On Wed, 11 May 2005 Robert.McGehee at geodecapital.com wrote:
>>
>>> Yes, you are correct. I had only checked one of my platforms. Linux
>>> works as you suggest. But for me on Windows,
>>>
>>>> x <- 1
>>>> y <- 0.2
>>>> x %/% y
>>>
>>> [1] 5  ## I get a 4 in Linux
>>
>>
>> I get 5 on Windows, but
>>
>>> (x %% y) + y * (x %/% y)
>>
>> [1] 1
>>
>> so is there a problem particular to your Windows runtime?
>>
>>
>>>
>>> version
>>>         _             =20
>>> platform i386-pc-mingw32
>>> arch     i386          =20
>>> os       mingw32       =20
>>> system   i386, mingw32 =20
>>> status                 =20
>>> major    2             =20
>>> minor    1.0           =20
>>> year     2005          =20
>>> month    04            =20
>>> day      18            =20
>>> language R             =20
>>>
>>>
>>> -----Original Message-----
>>> From: Peter Dalgaard [mailto:p.dalgaard at biostat.ku.dk]=20
>>> Sent: Wednesday, May 11, 2005 4:14 PM
>>> To: McGehee, Robert
>>> Cc: ted.harding at nessie.mcc.ac.uk; Peter Dalgaard; R-bugs at biostat.ku.dk;
>>> kjetil at acelerate.com; r-devel at stat.math.ethz.ch
>>> Subject: Re: [Rd] bug in modulus operator %% (PR#7852)
>>>
>>>
>>> "McGehee, Robert" <Robert.McGehee at geodecapital.com> writes:
>>>
>>>> Yes, but from ?"%%":
>>>> "It is guaranteed that 'x =3D=3D (x %% y) + y * (x %/% y)' (up to =
>>>
>>> rounding
>>>
>>>> error) ..."
>>>> =20
>>>> (R 2.1.0)
>>>>
>>>>> x <- 1
>>>>> y <- 0.2
>>>>> x %% y
>>>>
>>>> [1] 0.2
>>>>
>>>>> (x %% y) + y * (x %/% y)
>>>>
>>>> [1] 1.2
>>>> =20
>>>> Certainly 1 does not equal 1.2 as the documentation would suggest, and
>>>> these seem like large enough numbers to not be effected by rounding
>>>> errors or lack of precision.
>>>
>>>
>>> Now that looks a bit odd, but it isn't universal:
>>>
>>>> x <- 1
>>>> y <- 0.2
>>>> x %% y
>>>
>>> [1] 0.2
>>>
>>>> x %/% y
>>>
>>> [1] 4
>>>
>>>> (x %% y) + y * (x %/% y)
>>>
>>> [1] 1
>>>
>>> So what platform was that happening on?
>>
>
>


-- 

Kjetil Halvorsen.

Peace is the most effective weapon of mass construction.
               --  Mahdi Elmandjra




-- 
No virus found in this outgoing message.
Checked by AVG Anti-Virus.