I just got caught by the way in character vectors are sorted.
It seems that on my machine "sort" (and related functions like
"order")
only consider characters related to punctuation (at least here the "+"
and "-") when there is no difference in the remaining characters:
> x1 <- c("-A","+A")
> x2 <- c("+A","-A")
> sort(x1) # sorting is according to "-" and "+"
[1] "-A" "+A"
> sort(x2)
[1] "-A" "+A"
> x3 <- c("-Aa","-Ab")
> x4 <- c("-Aa","+Ab")
> x5 <- c("+Aa","-Ab")
> sort(x3)
[1] "-Aa" "-Ab" # here the "+" and "-"
are ignored
> sort(x4)
[1] "-Aa" "+Ab"
> sort(x5)
[1] "+Aa" "-Ab"
I understand from the help that this depends on how characters are
collated, and that this scheme follows the multi-level comparison in
unicode (http://www.unicode.org/reports/tr10/).
However, what I need is a strict left-to-right comparison of the sort
provided by strcmp or wcscmp in glibc. The particular ordering of
special characters is not so important, but there should be no
"multi-level" aspect to the sorting.
Is there a way to achieve this in R?
Thanks for your help
Pascal
On 08 Nov 2016, at 13:18 , Pascal A. Niklaus <pascal.niklaus at ieu.uzh.ch> wrote:> I just got caught by the way in character vectors are sorted. > > It seems that on my machine "sort" (and related functions like "order") only consider characters related to punctuation (at least here the "+" and "-") when there is no difference in the remaining characters: > > > x1 <- c("-A","+A") > > x2 <- c("+A","-A") > > sort(x1) # sorting is according to "-" and "+" > [1] "-A" "+A" > > sort(x2) > [1] "-A" "+A" > > > x3 <- c("-Aa","-Ab") > > x4 <- c("-Aa","+Ab") > > x5 <- c("+Aa","-Ab") > > sort(x3) > [1] "-Aa" "-Ab" # here the "+" and "-" are ignored > > sort(x4) > [1] "-Aa" "+Ab" > > sort(x5) > [1] "+Aa" "-Ab" > > I understand from the help that this depends on how characters are collated, and that this scheme follows the multi-level comparison in unicode (http://www.unicode.org/reports/tr10/). > > However, what I need is a strict left-to-right comparison of the sort provided by strcmp or wcscmp in glibc. The particular ordering of special characters is not so important, but there should be no "multi-level" aspect to the sorting. > > Is there a way to achieve this in R? >I'd try one of two ways (the above is not happening for me, so I cannot test): (1) Temporarily set the Locale to "C": Sys.setlocale("LC_COLLATE", "C"). That should work as long as you stay in good ol' ASCII. (2) Figure out (Don't look at me!) how to diddle the ICU settings for your system, icuSetCollate() is claimed to be your friend. -pd> Thanks for your help > > Pascal > > ______________________________________________ > R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code.-- Peter Dalgaard, Professor, Center for Statistics, Copenhagen Business School Solbjerg Plads 3, 2000 Frederiksberg, Denmark Phone: (+45)38153501 Office: A 4.23 Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
Hello,
What is your sessionInfo()?
With me it works as expected:
> sort(c("-", "+"))
[1] "-" "+"
> sort(c("+", "-"))
[1] "-" "+"
> x5 <- c("+Aa","-Ab")
> sort(x5)
[1] "-Ab" "+Aa"
> sessionInfo()
R version 3.3.2 (2016-10-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1
locale:
[1] LC_COLLATE=Portuguese_Portugal.1252
LC_CTYPE=Portuguese_Portugal.1252
[3] LC_MONETARY=Portuguese_Portugal.1252 LC_NUMERIC=C
[5] LC_TIME=Portuguese_Portugal.1252
attached base packages:
[1] stats graphics grDevices utils datasets methods base
Hope this helps,
Rui Barradas
Em 08-11-2016 12:18, Pascal A. Niklaus escreveu:> I just got caught by the way in character vectors are sorted.
>
> It seems that on my machine "sort" (and related functions like
"order")
> only consider characters related to punctuation (at least here the
"+"
> and "-") when there is no difference in the remaining characters:
>
> > x1 <- c("-A","+A")
> > x2 <- c("+A","-A")
> > sort(x1) # sorting is according to "-" and "+"
> [1] "-A" "+A"
> > sort(x2)
> [1] "-A" "+A"
>
> > x3 <- c("-Aa","-Ab")
> > x4 <- c("-Aa","+Ab")
> > x5 <- c("+Aa","-Ab")
> > sort(x3)
> [1] "-Aa" "-Ab" # here the "+" and
"-" are ignored
> > sort(x4)
> [1] "-Aa" "+Ab"
> > sort(x5)
> [1] "+Aa" "-Ab"
>
> I understand from the help that this depends on how characters are
> collated, and that this scheme follows the multi-level comparison in
> unicode (http://www.unicode.org/reports/tr10/).
>
> However, what I need is a strict left-to-right comparison of the sort
> provided by strcmp or wcscmp in glibc. The particular ordering of
> special characters is not so important, but there should be no
> "multi-level" aspect to the sorting.
>
> Is there a way to achieve this in R?
>
> Thanks for your help
>
> Pascal
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
Thanks for all suggestions.
With my build (from the CRAN repo) I don't get ICU support, and setting
LC_COLLATE to "C" did not help.
> capabilities("ICU")
ICU
FALSE
> sessionInfo()
R version 3.3.2 (2016-10-31)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 14.04.5 LTS
locale:
[1] LC_CTYPE=en_DK.UTF-8 LC_NUMERIC=C
[3] LC_TIME=en_DK.UTF-8 LC_COLLATE=C
[5] LC_MONETARY=en_DK.UTF-8 LC_MESSAGES=en_DK.UTF-8
[7] LC_PAPER=en_DK.UTF-8 LC_NAME=C
[9] LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_DK.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] stats graphics grDevices utils datasets methods base
loaded via a namespace (and not attached):
[1] compiler_3.3.2 tools_3.3.2
However, using stringi::stri_sort did the trick !