On Thu, 3 Jul 2008, Philip James Smith wrote:
He Ted:
My command will look like:
gnu.dat <- read.csv( pipe( "cut -d,
-f9,2669,2676,2684,506,467,423,6,16,622,474,2,403,19,436,400,444,430,39
0,2671,437,428,2686,23,21,2246,2326,2288,2282,2272,2360,2318,2262,476,5
02,605,519,523,527,531,535
/home/philipsmith/file_name.csv" ) )
This code uses pipe() and the unix command "cut."
What I'm trying to do is to read certain fields in a .csv file..... my
original .csv file has THOUSANDS of columns, so I just want to read
certain
ones...The unix command "cut" uses -d to specify that the comma ","
is the
delimiter in the file, and the -f to identify the field numbers. After
the
field numbers "cut" wants the file name.
Evidently, the unix command called by pipe must be within double
quotes...
and that is my problem...
That is not 'evident', not even true in your example. You passed
pipe() a
length-one character vector containing a command, and the quotes were
something the parser used to recognize the element of the character
vector.
In my actual problem, I've got THOUSANDS of csv files that contain
some
common variables and other variables that differ from file to file.
I've
specified the variables I want as a character vector called namz, and
read
the first line of each file using scan(). I'm then able to find out
using
match() the positions of the variables I want in each file. the code
listed
above gives only 1 of THOUSANDS of possible configurations of those
desired
variables.
In order to get this task done, for each file I get the position
numbers in a
variable called "col.pos" and then create the unix command using
paste:
unix.cmd <- paste( "cut -d -f" , col.pos , " " , fn , sep = ''
)
and then you can use
gnu.dat <- read.csv(pipe(unix.cmd))
Much simpler!
Then what I want to do is to create the R command that will read the
file...
something like:
cmd0 <- paste( "gnu.dat <- read.csv( pipe ( \"" ,
unix.cmd ,
"\" ) ) " , sep='')
Then, I'll use
eval( parse (text = cmd0) )
to read a file... and embed this in more code to read all of the
THOUSANDS of
files, each 1 at a time.
However, in the "cmd0" variable, I've used "\"" ... this does not
work... it
yields a character string that looks like...
cmd0
[1] "gnu.dat <- read.csv( pipe ( \"cut -d
-f9,2669,2676,2684,506,467,423,6,16,622,474,2,403,19,436,400,444,430,39
0,2671,437,428,2686,23,21,2246,2326,2288,2282,2272,2360,2318,2262,476,5
02,605,519,523,527,531,535
/home/philipsmith/Bubba/NIS/2007_claf/y2007.csv\" ) ) "
You'll see in that line the slash that precedes the double quote. the
pipe()
command is giving unix a unix command that has a slash in it: unix is
expecting only a double quote and not a slash and gives and error.
So, those are all of the details. I'll be quite grateful if you can
tell me
how to paste a double quote (") in front of and at the end of a string
so
that unix will recognize the string as a valide unix command.
Actually, this is a shell command line, which the shell parses and
passes
to Unix *provided* the parsing is not switched off by quoting.
If you want to quote a command string, that is what shQuote() is for.
Very gratefully,
Phil Smith
Duluth, GA
(Ted Harding) wrote:
Hi Philip,
I think a bit more clarification may be useful yet!
1: How are you sending the command from R to Linux?
2: What is the command intended to be (as seen by Linux)?
And from what source (quasi-command line; script file; ...)
would it be read by Linux?
For example, on my Linux machine, I just did (in a shell
command line):
[EMAIL PROTECTED]:~$ touch "New Mexico"
[EMAIL PROTECTED]:~$ ls -l "New Mexico"
-rw-r--r-- 1 ted ted 0 2008-07-03 10:58 New Mexico
and then, in R, I did:
system("ls -l \"New Mexico\"")
-rw-r--r-- 1 ted ted 0 2008-07-03 10:58 New Mexico
so I don't seem to have had the problem you describe below.
On the other hand, back in Linux, if I do:
[EMAIL PROTECTED]:~$ ls -l \"New Mexico\"
ls: "New: No such file or directory
ls: Mexico": No such file or directory
which looks like the problem you describe -- so clearly the
command did not get through in that form from R using the
command system("ls -l \"New Mexico\"").
Best wishes,
Ted.
On 03-Jul-08 09:18:14, Philip James Smith wrote:
Thanks for your reply, Ted... I am very grateful for it.
Using your notation, what I need is a character string Y that looks
like:
>
> Y
>[1] ""New Mexico""
rather than
> Y
>[1] "\"New Mexico\""
i.e., Y must have the string 'New Mexico' surrounded by double
quotes,
rather than double quotes preceded by slashes.
The reason why I need it that way is that I've over simplified my
request
and the character string is actually a unix command that needs to be
surrounded by double quotes when embeded in the unix (linux)
executes it.
When unix sees that slash before the quotes, it gives an
error.
I'd be grateful if you can provide a solution to this!!
Thank you, Ted!
Gratefully,
Phil Smith
Duluth, GA
(Ted Harding) wrote:
On 03-Jul-08 01:25:55, Philip James Smith wrote:
Hi R Community:
I've got a character string that looks like: New Mexico
How to I create the new character string that looks like: "New
Mexico" That is, it is the original string (New Mexico) with
double
quotes infront and behind it?
Thanks,
Phil Smith
I tried the following. Is that the sort of thing you want to
achieve?
X<-"New Mexico"
Y<-"\"New Mexico\""
X
# [1] "New Mexico"
Y
# [1] "\"New Mexico\""
plot((1:10),xlab=X,ylab=Y)
Best wishes,
Ted.
--------------------------------------------------------------------
E-Mail: (Ted Harding) <[EMAIL PROTECTED]>
Fax-to-email: +44 (0)870 094 0861
Date: 03-Jul-08 Time:
09:22:10
------------------------------ XFMail
------------------------------
______________________________________________
R-help@r-project.org mailing list
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.
--------------------------------------------------------------------
E-Mail: (Ted Harding) <[EMAIL PROTECTED]>
Fax-to-email: +44 (0)870 094 0861
Date: 03-Jul-08 Time: 11:08:36
------------------------------ XFMail ------------------------------
______________________________________________
R-help@r-project.org mailing list
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.
--
Brian D. Ripley, [EMAIL PROTECTED]
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
______________________________________________
R-help@r-project.org mailing list
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.
