Re: [R] Problem setting TMPDIR on the fly
Sorted - sort of... I've simply wrapped the invocation of R in a shell script along with the necessary environment variables. However, this is precisely the solution that I was hoping to avoid! I can see why it would be inadvisable to switch TMPDIR mid way through a session, but wouldn't it be useful to able switch it on the way in? As ever thanks for the help, cheers, Ben -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Peter Dalgaard Sent: 30 October 2006 17:08 To: Benjamin Lloyd-Hughes Cc: Peter Dalgaard Subject: Re: [R] Problem setting TMPDIR on the fly Benjamin Lloyd-Hughes [EMAIL PROTECTED] writes: Hi Peter, Thanks for the swift response! The problem is arising because I'm running R in batch mode through cron. As such I'd rather set all such environment variables explicitly and not make any assumptions about what variables maybe getting set on the way in. A typical call looks like this /usr/bin/R CMD BATCH --vanilla foo.R bar.Rout The first few lines of foo.R might run .libPaths(c(/usr/lib/R/library, /eurotempest/software/src/R/packages)) Sys.putenv(TMPDIR = /eurotempest/tmp/R/) Hence the problem. Is there a more sensible way to go? I'd just set it on the command line, as in TMPDIR=/eurotempest/tmp/R/ R -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Peter Dalgaard Sent: 30 October 2006 16:33 To: Benjamin Lloyd-Hughes Cc: r-help@stat.math.ethz.ch Subject: Re: [R] Problem setting TMPDIR on the fly Benjamin Lloyd-Hughes [EMAIL PROTECTED] writes: Hi folks, I'm having a spot of bother with tempdir(). If I set the environment variable TMPDIR in my shell prior to invoking R it works as expected: Sys.getenv(TMPDIR) TMPDIR /eurotempest/tmp/R tempdir() [1] /eurotempest/tmp/R/Rtmp0xY4XD However if I don't set it prior to entry, but attempt to set it from within the current session, the change is not detected/respected: Sys.getenv(TMPDIR) TMPDIR Sys.putenv(TMPDIR = /eurotempest/tmp/R/) Sys.getenv(TMPDIR) TMPDIR /eurotempest/tmp/R/ tempdir() [1] /tmp/Rtmp9chsNX Any ideas? You're not supposed to do that. There is one and only one tempdir per R session. This is used for storing stuff which R expects to be able to find again later, so it would be a bad thing just to switch it out. -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 -- This e-mail and its contents are intended for the use of the addressee(s) and may be confidential/privileged. No-one else may review, copy, disclose or otherwise use it or its contents. If you receive this e-mail in error, please contact the originator and delete it as soon as possible. Benfield Limited is authorised by the Financial Services Authority under the reference number 311884. Registered in England no 1170753. Registered office 55 Bishopsgate. Please refer to Benfield Limited's terms and conditions (www.benfieldgroup.com/terms) for a description of our services, duties and points of contact. Please review these terms and conditions at inception and renewal of all reinsurance and insurance contracts. Please note that our terms and conditions and interests in other companies may change over time and these changes will be reflected on the Benfield Limited website. The latest version of our terms and conditions supersedes all previous versions. == -- O__ Peter Dalgaard Øster Farimagsgade 5, Entr.B c/ /'_ --- Dept. of Biostatistics PO Box 2099, 1014 Cph. K (*) \(*) -- University of Copenhagen Denmark Ph: (+45) 35327918 ~~ - ([EMAIL PROTECTED]) FAX: (+45) 35327907 -- This e-mail and its contents are intended for the use of the addressee(s) and may be confidential/privileged. No-one else may review, copy, disclose or otherwise use it or its contents. If you receive this e-mail in error, please contact the originator and delete it as soon as possible. Benfield Limited is authorised by the Financial Services Authority under the reference number 311884. Registered in England no 1170753. Registered office 55 Bishopsgate. Please refer to Benfield Limited's terms and conditions (www.benfieldgroup.com/terms) for a description of our services, duties and points of contact. Please review
[R] Problem setting TMPDIR on the fly
Hi folks, I'm having a spot of bother with tempdir(). If I set the environment variable TMPDIR in my shell prior to invoking R it works as expected: Sys.getenv(TMPDIR) TMPDIR /eurotempest/tmp/R tempdir() [1] /eurotempest/tmp/R/Rtmp0xY4XD However if I don't set it prior to entry, but attempt to set it from within the current session, the change is not detected/respected: Sys.getenv(TMPDIR) TMPDIR Sys.putenv(TMPDIR = /eurotempest/tmp/R/) Sys.getenv(TMPDIR) TMPDIR /eurotempest/tmp/R/ tempdir() [1] /tmp/Rtmp9chsNX Any ideas? Cheers, Ben R.Version() $platform [1] i686-redhat-linux-gnu $arch [1] i686 $os [1] linux-gnu $system [1] i686, linux-gnu $status [1] $major [1] 2 $minor [1] 3.0 $year [1] 2006 $month [1] 04 $day [1] 24 $`svn rev` [1] 37909 $language [1] R $version.string [1] Version 2.3.0 (2006-04-24) -- This e-mail and its contents are intended for the use of the addressee(s) and may be confidential/privileged. No-one else may review, copy, disclose or otherwise use it or its contents. If you receive this e-mail in error, please contact the originator and delete it as soon as possible. Benfield Limited is authorised by the Financial Services Authority under the reference number 311884. Registered in England no 1170753. Registered office 55 Bishopsgate. Please refer to Benfield Limited's terms and conditions (www.benfieldgroup.com/terms) for a description of our services, duties and points of contact. Please review these terms and conditions at inception and renewal of all reinsurance and insurance contracts. Please note that our terms and conditions and interests in other companies may change over time and these changes will be reflected on the Benfield Limited website. The latest version of our terms and conditions supersedes all previous versions. == [[alternative HTML version deleted]] __ R-help@stat.math.ethz.ch 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.
[R] Matrix rotation
Dearest All, Ok so I've had a couple of glasses of wine over lunch today... This is likely to be trivial but I'm struggling to find a more elegant way to obtain the following matrix rotations: M - matrix(c(1,0,0,0), ncol=2) M [,1] [,2] [1,]10 [2,]00 N - abind(M[2,],M[1,],along=2) N [,1] [,2] [1,]01 [2,]00 P - abind(N[2,],N[1,],along=2) P [,1] [,2] [1,]00 [2,]01 Q - abind(P[,2],P[,1],along=2) Q [,1] [,2] [1,]00 [2,]10 And, more generally wish to rotate a n-dimensional data cube about some specified axis. Cheers, Ben __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] Fwd: Matrix rotation
Ok I warned you that I'd been drinking! What I really meant was something to go from: [,1] [,2] [1,]12 [2,]43 to [,1] [,2] [1,]41 [2,]32 to [,1] [,2] [1,]34 [2,]21 to [,1] [,2] [1,]23 [2,]14 Sorry for being a muppet, B Begin forwarded message: From: Benjamin Lloyd-Hughes [EMAIL PROTECTED] Date: 24 November 2005 16:15:58 GMT To: r-help@stat.math.ethz.ch Subject: Matrix rotation Dearest All, Ok so I've had a couple of glasses of wine over lunch today... This is likely to be trivial but I'm struggling to find a more elegant way to obtain the following matrix rotations: M - matrix(c(1,0,0,0), ncol=2) M [,1] [,2] [1,]10 [2,]00 N - abind(M[2,],M[1,],along=2) N [,1] [,2] [1,]01 [2,]00 P - abind(N[2,],N[1,],along=2) P [,1] [,2] [1,]00 [2,]01 Q - abind(P[,2],P[,1],along=2) Q [,1] [,2] [1,]00 [2,]10 And, more generally wish to rotate a n-dimensional data cube about some specified axis. Cheers, Ben [[alternative text/enriched version deleted]] __ R-help@stat.math.ethz.ch mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] whether R can import netcdf file format data ?
I've been using the ncdf package for reading climate data (mostly, NCEP/NCAR and ECMWF reanalysis stuff). The package does what it says on the tin, but can become quite slow when working with large data sets (e.g. a stack of daily grids say 144x73 elements 1970-to date). Does anyone have any tips for speedy retrieval? I'm thinking about typical drilling/slicing type operations. Cheers, Ben -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Prof Brian Ripley Sent: 12 November 2004 15:05 To: Yogesh K. Tiwari Cc: [EMAIL PROTECTED] Subject: Re: [R] whether R can import netcdf file format data ? On Fri, 12 Nov 2004, Yogesh K. Tiwari wrote: Whether 'R' can import Netcdf format data files ?? It can read from and write to them, at least. Any package in 'R' that can help me on this front ?? Do look at the `R Data Import/Export' manual that ships with R (a recent version is needed), or the R FAQ or the list of packages on CRAN. You will find RNetCDF, ncdf and ncvar. -- 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, UKFax: +44 1865 272595 __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
RE: [R] rgdal under windows?
Success! I'm pleased to inform that Roger's recipe worked a treat for compiling rgdal_0.2-7 in conjunction with gdal-1.2.1 and proj-4.4.8. Using MinGW-3.1.0 with MYS-1.0.8 on an xp pro box (SP1) Proj-4 compiled straight out of the box. Gdal needed a couple of minor tweaks wrt how the boolean type is defined. I suspect that gcc is being strict about name spaces. My hack was an explicit typedef in jpeglib.h: typedef unsigned char boolean; and to comment out the definition in jmorecfg.h After that it was plain sailing except for the curious behaviour of the MinGW dlltool.exe which refused to to pick up R.dll unless it was given the full path _irrespective_ of what was placed in $PATH. Big thanks to Roger and Brian for their helpful advice. Complicated? Not if you learn to read! Cheers, Ben -Original Message- From: Roger Bivand [mailto:[EMAIL PROTECTED] Sent: 19 August 2004 19:27 To: Prof Brian Ripley Cc: Benjamin Lloyd-Hughes; [EMAIL PROTECTED] Subject: Re: [R] rgdal under windows? On Thu, 19 Aug 2004, Prof Brian Ripley wrote: On Thu, 19 Aug 2004, Benjamin Lloyd-Hughes wrote: Has anyone had any joy getting the rgdal package to compile under windows? First you need to meet the system requirements: SystemRequirements: GDAL library from http://www.remotesensing.org/gdal/download.html Since that has build instructions which say GDAL can be built on Windows using MS VC++ 6.x and MS Visual Studio .NET (C++) at the DOS command line. you will first need to get a copy of DOS and those compilers. (I suspect they mean the Windows command prompt, but they don't seem to know the difference which is not confidence-inspiring. I also suspect they mean `or' not `and'.) Then you will have to fathom out how to link against a VC++ DLL which since this C++ is very unlikely to work with MinGW. So you probably need to make the package's rgdal.dll under VC++ -- see README.packages. I have tried and failed to make GDAL with MinGW. I've been trying with MinGW using Duncan Murdoch's Rtools but to no avail. Do read README.packages and follow the advice at the top. You don't have the (mis-attributed) tools in your path. But that is probably the least of your potential problems. Exactly. The closest anyone has got so far is Hisaji Ono, who used MSYS (http://www.mingw.org/) to build PROJ.4 and GDAL (GDAL depends on PROJ.4, PROJ.4 needs a PATH to metadata files for projection and transformation), and then hand-pasted the paths to the GDAL headers and library into src/Makevars, running Rcmd INSTALL rgdal at the Windows command prompt as usual. All of this can be repeated, but is not portable, and does not suit the very valuable standard binary package build system for Windows. Roughly: 1. Download everything you need to build source packages under Windows and make sure it works; 2. Download MSYS and make sure it works; 3. Download the GDAL and PROJ.4 source tarballs, and possibly other libraries you want to use with GDAL, and *within MSYS* untar, ./configure with the appropriate arguments, at least make but maybe also make install - now leave MSYS; 4. Download the rgdal source package, and *at the Windows command prompt* untar it, change the name of configure to something else, create src/Makevars manually from src/Makevars.in, insert the correct values of: PKG_CPPFLAGS as -Ipath to the directory with the GDAL headers, PKG_LIBS as -L ... GDAL libraries -lgdal, run Rcmd INSTALL rgdal (better first Rcmd check rgdal), and repeat until all the problems are resolved. 5. The installation should work locally, but some paths are compiled into the resulting *.dll, so it will, most likely, not be portable. Both Hisaji and I were surprised that this actually seemed to work (as of late Autumn last year for the versions of MinGW and MSYS available then; further fragility is introduced by much of GDAL being written in C++); it is not wholly impossible that it could be made available portably as a Windows binary with an installer, but not through the regular R binary package repositories. -- Roger Bivand Economic Geography Section, Department of Economics, Norwegian School of Economics and Business Administration, Breiviksveien 40, N-5045 Bergen, Norway. voice: +47 55 95 93 55; fax +47 55 95 93 93 e-mail: [EMAIL PROTECTED] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
[R] rgdal under windows?
Hi, Has anyone had any joy getting the rgdal package to compile under windows? I've been trying with MinGW using Duncan Murdoch's Rtools but to no avail. After unpacking the tarball I put it here: D:\rw1091\src\library\rgdalls DESCRIPTION Rconfigure inst man INDEXcleanup configure.in install.R src And then tried: D:\rw1091\src\gnuwin32make pkg-rgdal make: *** [pkg-rgdal] Error 255 Any help would be most appreciated! Cheers, Ben [[alternative HTML version deleted]] __ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html