R help - Aug 2004 - gsub, backslash and xtable

R Version 1.9.1  (2004-06-21)
Mac OS X.3.5 Dual 2GHz PowerPC G5
GUI = "AQUA"


I have a data.frame comprising percentiles with the column headings 
containing % characters, e.g.
 > (pp <- colnames(temp2))
[1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
I use xtable to convert the data.frame to Latex but I want to protect 
these % signs from Latex using a backslash in the normal way before 
calling xtable.
I have tried using gsub as follows;
 > gsub("\%","\\%",pp)
[1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
also
 > gsub("%","\134%",pp) #octal for backslash
[1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
Both of which fail to provide what I need.

I verified  my 'regexps' using awk under Darwin thus;
$ cat fred
"5%"  "10%" "25%" "50%" "75%"
"90%" "95%"
$ awk '{gsub(/%/,"\\%"); print}' <fred
"5\%"  "10\%" "25\%" "50\%"
"75\%" "90\%" "95\%"
and
$ awk '{gsub(/%/,"\134%"); print}' <fred
"5\%"  "10\%" "25\%" "50\%"
"75\%" "90\%" "95\%"

As a possble 'work around', I noticed that,
 > chartr("z","\",gsub("%","z%",pp))
Error: syntax error
 > chartr("z","\\",gsub("%","z%",pp))
[1] "5\\%"  "10\\%" "25\\%" "50\\%"
"75\\%" "90\\%" "95\\%"
 >
chartr("z","\134",gsub("%","z%",pp))
[1] "5\\%"  "10\\%" "25\\%" "50\\%"
"75\\%" "90\\%" "95\\%"

As the xtable is then 'catted' to a file and read back (vide infra) I 
actually end up with what I want using the latter example.
However I am very much left with the feeling that R is in control of me 
rather than vise versa.



Secondly, as I am building up a character vector of sentences, tables 
and figures, I wanted to convert my xtable output to a character vector 
with newline
separators. I have only able to accomplish this by printing to a 
temporary file thus,

theTx <- "\\documentclass[A4paper,10pt]{article}"
.
.
theTx <- paste(theTx, paste_xtable(temp2,"Percentiles for scores"),
sep
= "")
.
theTx <- paste(theTx,"\n"
,"\\end{document}","\n", sep = "")
.
cat(theTx) #into a file for Latex

############## with my past_xtable function being ##########
paste_xtable <- function(a_table, cap) {
sink(file = "levzz", append = FALSE, type = "output")
print(xtable(a_table,caption=cap))
sink()
#read it  back
temp <- readLines("levzz", n=-1) #note / get doubled automaticaly
unlink("levzz") #delete file
a <- "\n"
for (i in 1:length(temp)) {
	a <- paste(a,temp[i],"\n",sep = "")
	}
return(a)
}

I should be most grateful for a  more elegant solutions to both these 
issues or a pointer to the documentation.

Paul

"P. B. Pynsent" <p.b.pynsent at bham.ac.uk> writes:
> I have a data.frame comprising percentiles with the column headings
> containing % characters, e.g.
>  > (pp <- colnames(temp2))
> [1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
> I use xtable to convert the data.frame to Latex but I want to protect
> these % signs from Latex using a backslash in the normal way before
> calling xtable.
> I have tried using gsub as follows;
>  > gsub("\%","\\%",pp)
> [1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
> also
...
> However I am very much left with the feeling that R is in control of
> me rather than vise versa.

The generic rule for backslashes is that you need twice as many as you
thought:
> p
[1] "25%"
> gsub("%","\\\\%",p)
[1] "25\\%"
> cat(gsub("%","\\\\%",p),"\n")
25\%

The thing that people usually forget is that you have two levels of
escaping, one in R's string parser and another one in the regexp
machinery. 

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907

On Fri, 27 Aug 2004, P. B. Pynsent wrote:
> R Version 1.9.1  (2004-06-21)
> Mac OS X.3.5 Dual 2GHz PowerPC G5
> GUI = "AQUA"
> 
> 
> I have a data.frame comprising percentiles with the column headings 
> containing % characters, e.g.
>  > (pp <- colnames(temp2))
> [1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
> I use xtable to convert the data.frame to Latex but I want to protect 
> these % signs from Latex using a backslash in the normal way before 
> calling xtable.
> I have tried using gsub as follows;
>  > gsub("\%","\\%",pp)
> [1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
> also
>  > gsub("%","\134%",pp) #octal for backslash
> [1] "5%"  "10%" "25%" "50%"
"75%" "90%" "95%"
> Both of which fail to provide what I need.
Remember you need to double \ in R character strings (in the FAQ, for 
example, and in ?regex):
>  gsub("%","\\\\%",pp)
[1] "5\\%"  "10\\%" "25\\%" "50\\%"
"75\\%" "90\\%" "95\\%"
> cat(gsub("%","\\\\%",pp), "\n")
5\% 10\% 25\% 50\% 75\% 90\% 95\% 



-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

Peter Dalgaard wrote:
> ``The generic rule for backslashes is that you need twice as many
>   as you thought''
	And you have to apply that rule recursively! :-)

				cheers,

					Rolf Turner
					rolf at math.unb.ca