R help - Nov 2012 - the results of the SORT function differ from Scilab/Matlab for Complex Numbers

Hello useRs,

The results of the SORT function differ from Scilab/Matlab for Complex 
Numbers in my example.
This design is the desirable in R?

Thanks.
Cleber

r <- c(
  1.7507+0.1689i, 1.7507-0.1689i, 1.3886+0.0000i, 1.0458+0.0792i, 
1.0458-0.0792i,
  0.8279+0.1861i, 0.8279-0.1861i, 0.8263+0.3731i, 0.8263-0.3731i, 
0.6548+0.0000i
)

 > cbind(sort(r, d=T))
                 [,1]
  [1,] 1.7507+0.1689i
  [2,] 1.7507-0.1689i
  [3,] 1.3886+0.0000i
  [4,] 1.0458+0.0792i
  [5,] 1.0458-0.0792i
  [6,] 0.8279+0.1861i
  [7,] 0.8279-0.1861i
  [8,] 0.8263+0.3731i
  [9,] 0.8263-0.3731i
[10,] 0.6548+0.0000i
 >

Scilab
-->gsort(r)
  ans  
     1.7507 + 0.1689i
     1.7507 - 0.1689i
     1.3886
     1.0458 + 0.0792i
     1.0458 - 0.0792i
     0.8263 + 0.3731i
     0.8263 - 0.3731i
     0.8279 + 0.1861i
     0.8279 - 0.1861i
     0.6548

 > sessionInfo()
R version 2.15.2 (2012-10-26)
Platform: x86_64-w64-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=Portuguese_Brazil.1252 LC_CTYPE=Portuguese_Brazil.1252
[3] LC_MONETARY=Portuguese_Brazil.1252 LC_NUMERIC=C
[5] LC_TIME=Portuguese_Brazil.1252

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods base

other attached packages:
[1] R.matlab_1.6.1    R.oo_1.10.2       R.methodsS3_1.4.2

loaded via a namespace (and not attached):
[1] MASS_7.3-22  tools_2.15.2
 >

On Fri, Nov 9, 2012 at 10:02 AM, Cleber N.Borges
<klebyn@yahoo.com.br>wrote:
> Hello useRs,
>
> The results of the SORT function differ from Scilab/Matlab for Complex
> Numbers in my example.
> This design is the desirable in R?
>
Well, it's deliberate and documented.

R sorts complex numbers by real part then by imaginary part.  Matlab,
according to its documentation, sorts by modulus then phase.

There isn't a unique way to sort complex numbers, so you're going to get
differences.  Personally, I think the R method is more straightforward,
since you don't need to decide and remember where the branch cut goes on
the phase coordinate.

   -thomas

-- 
Thomas Lumley
Professor of Biostatistics
University of Auckland

	[[alternative HTML version deleted]]