[R] probleme with R CMD check
Martin Maechler
maechler at stat.math.ethz.ch
Fri Jul 18 11:34:50 CEST 2008
>>>>> "CG" == Christophe Genolini <cgenolin at u-paris10.fr>
>>>>> on Fri, 18 Jul 2008 11:14:25 +0200 writes:
CG> Hi the list,
CG> Back on an old problem. I thaught that was ok, but it is still not
CG> working on my computer.
CG> I write a toy package called packClassic. R CMD build packClassic works
CG> fine.
CG> When I run R CMD CHECK packClassic, I get
CG> --- 8< ---------------
CG> * checking for unstated dependencies in R code ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> --- 8< ---------------
CG> I gess that my configuration is correct since on an other package (a
CG> real one), I get no problem.
CG> The program is called "progClassic.R". Obviouly, is in a directorie
CG> called \packClassic\R\
CG> The doc are in man, a data base is in data and some test are in tests
CG> Here are the codes, the DESCRIPTION and NAMESPACE, then the file
CG> 00install.out and 00check.log :
CG> I also put the package (since R CMD build works fine) on
CG> http:\\christophe.genolini.free.fr/packClassic_1.0.tar.gz
I tried the above, --- after changing "\\" to "//" but no file
was found.
(read on !)
CG> http:\\christophe.genolini.free.fr/packClassic_1.0.zip
CG> Does someone know what is wrong ?
CG> Christophe
CG> ---- 8< --------- packClassic.R ----------------
CG> ######################################################
CG> ### ProgClassic.R ###
CG> ######################################################
CG> userA <- function(x){x+1}
CG> privateA <- function(x){x+2}
CG> .userB <- function(x){x+10}
CG> .privateB <- function(x){x+20}
CG> userC <- function(x){userA(privateA(x))}
CG> privateC <- function(x){.userB(.privateB(x))}
CG> --- 8< ---------- testsPackClassic.R----------------------------
CG> ##########################################
CG> ### Nom du programme : testClassique.R ###
CG> ##########################################
CG> load("../data/dn.Rda")
CG> source("../R/progClassic.R")
CG> userA(1)
CG> privsateA(2)
CG> .userB(3)
CG> .privateB(4)
CG> userC(5)
CG> privateC(6)
CG> --- 8< ------------ dn.Rda ----------------------
CG> dn <- data.frame(age=c(1,2,4,8,12))
CG> save(dn,file=dn.Rda")
CG> --- 8< ----------- DESCRIPTION -------------
CG> Package:packClassic
CG> Type: Package
CG> Title: A (Not So) Short Introduction To S4
CG> Version: 1.0
CG> Date: 2008-07-16
CG> Author: Christophe Genolini
CG> Maintainer: <genolini at u-paris10.fr>
CG> Description: This package is a toy example used to illustrate the
CG> tutorial A (Not So) Short Introduction to S4.
CG> License: GPL (>=2)
CG> LazyLoad: ys
CG> --- 8< --------------- NAMESPACE --------------
CG> export("userA",
CG> ".userB",
CG> "userC"
CG> )
CG> --- 8< -------- man ----------------
CG> dn.Rd
CG> packClassic-internal.Rd
CG> packClassic-package.Rd
CG> userAC.Rd
CG> --- 8< ------------- 00install.out ----------------------
CG> installing R.css in
CG> C:/DOCUME~1/CHRIST~1/MESDOC~1/DOCSAR~1/R/TUTORI~1/PACKCL~1.RCH
CG> ---------- Making package packClassic ------------
CG> adding build stamp to DESCRIPTION
CG> installing NAMESPACE file and metadata
CG> installing R files
CG> installing data files
CG> installing man source files
CG> installing indices
CG> not zipping data
CG> installing help
>>>> Building/Updating help pages for package 'packClassic'
CG> Formats: text html latex example chm
CG> dn text html latex example
CG> packClassic-internal text html latex
CG> packClassic-package text html latex example
CG> userAC text html latex example
CG> adding MD5 sums
CG> * DONE (packClassic)
CG> --- 8< ----------------- 00check.log ------------------
CG> * using log directory 'C:/Documents and Settings/Christophe/Mes
CG> documents/Docs articles polys résumés/R/tutorial S4/packClassic.Rcheck'
I see many directories with spaces (" ") inside.
This used to cause all kind of problems in the past,
and has been explicitly discouraged from in "R for windows"
instructions.
But really I have no idea if that's the cause of your problem.
We *could* try your source package *if* it was available from a
URL.
Bonnes salutations,
Martin
CG> * using R version 2.7.1 (2008-06-23)
CG> * using session charset: ISO8859-1
CG> * checking for file 'packClassic/DESCRIPTION' ... OK
CG> * checking extension type ... Package
CG> * this is package 'packClassic' version '1.0'
CG> * checking package name space information ... OK
CG> * checking package dependencies ... OK
CG> * checking if this is a source package ... OK
CG> * checking whether package 'packClassic' can be installed ... OK
CG> * checking package directory ... OK
CG> * checking for portable file names ... OK
CG> * checking DESCRIPTION meta-information ... OK
CG> * checking top-level files ... OK
CG> * checking index information ... OK
CG> * checking package subdirectories ... WARNING
CG> Subdirectory 'data' contains no data sets.
CG> * checking R files for non-ASCII characters ... OK
CG> * checking R files for syntax errors ... OK
CG> * checking whether the package can be loaded ... OK
CG> * checking whether the package can be loaded with stated dependencies ... OK
CG> * checking whether the name space can be loaded with stated dependencies
CG> ... OK
CG> * checking for unstated dependencies in R code ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> See the information on DESCRIPTION files in the chapter 'Creating R
CG> packages' of the 'Writing R Extensions' manual.
CG> * checking S3 generic/method consistency ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> See section 'Generic functions and methods' of the 'Writing R Extensions'
CG> manual.
CG> * checking replacement functions ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> In R, the argument of a replacement function which corresponds to the right
CG> hand side must be named 'value'.
CG> * checking foreign function calls ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> See the chapter 'System and foreign language interfaces' of the 'Writing R
CG> Extensions' manual.
CG> * checking R code for possible problems ... NOTE
CG> Erreur : aucun package nommé 'packClassic' n'existe
CG> Séquence d'appel :
CG> 2: stop(txt, domain = NA)
CG> 1: library(package, lib.loc = lib.loc, character.only = TRUE, verbose =
CG> FALSE)
CG> Exécution arrêtée
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> * checking Rd files ... OK
CG> * checking Rd cross-references ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> Rd_aliases -> .find.package
CG> Exécution arrêtée
CG> * checking for missing documentation entries ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> All user-level objects in a package should have documentation entries.
CG> See the chapter 'Writing R documentation files' in manual 'Writing R
CG> Extensions'.
CG> * checking for code/documentation mismatches ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> * checking Rd \usage sections ... WARNING
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> Functions with \usage entries need to have the appropriate \alias entries,
CG> and all their arguments documented.
CG> The \usage entries must correspond to syntactically valid R code.
CG> See the chapter 'Writing R documentation files' in manual 'Writing R
CG> Extensions'.
CG> Erreur dans .find.package(package, lib.loc) :
CG> aucun package nommé 'packClassic' n'existe
CG> Calls: <Anonymous> -> .find.package
CG> Exécution arrêtée
CG> The \usage entries for S3 methods should use the \method markup and not
CG> their full name.
CG> See the chapter 'Writing R documentation files' in manual 'Writing R
CG> Extensions'.
CG> * checking data for non-ASCII characters ... OK
CG> * creating packClassic-Ex.R ... OK
CG> * checking examples ... ERROR
CG> Running examples in 'packClassic-Ex.R' failed.
CG> --- 8< -----------------------
c> I find the problem: the name of my base directory (the
c> one that contain R, man and test) was not the same than
c> the name of the package...
>>
>> Aha!
>>
c> I thaught this was allowed, I guess I was wrong.
>>
>> It has been documented to be allowed; and it has been, but
>> in some circumstances, R CMD check gave wrong warnings, as you
>> found.
>>
>> I had found and fixed this buglet on May 22, both for R-devel
>> and R-patched (which will become R 2.7.1 ...)
>>
>> Could you try 'R-patched' and confirm that the problem is
>> solved?
>>
>> Regards,
>> Martin
>>
>>
c> Christophe
>>
>> >>>>>>> "c" == cgenolin <cgenolin at u-paris10.fr> on Tue, 03
>> >>>>>>> Jun 2008 09:55:15 +0200 writes:
>> >>
>> >> >> provide commented, minimal, self-contained,
>> >> reproducible code. >> Please do -- both --
>> >>
c> Sorry... Usualy, I do...
>> >>
c> --- 8< ------------ userA <- function(x)x+1 privateA <-
c> function(x)x+2 .userB <- function(x)x+10 .privateB <-
c> function(x)x+20 userC <- function(x)userA(privateA(x))
c> privateC <- function(x).userB(.privateB(x)) userD <-
c> function(x)userA(.userB(x)) privateD <-
c> function(x)privateA(.privateB(x)) --- 8< ------------
>> >>
>> >> If I execute these in an empty R session, and do
>> >> package.skeleton("EXpkg") system("R CMD check EXpkg")
>> >>
>> >> I can NOT see the problem you describe below. Probably
>> >> you'd rather should upload the <pkg>.tar.gz to a public
>> >> place, to make your problem reproducible.
>> >>
>> >> Martin
>> >>
>> >> >>>>>>> "c" == cgenolin <cgenolin at u-paris10.fr> >>>>>>>
>> >> on Tue, 03 Jun 2008 00:32:13 +0200 writes:
>> >> >>
c> Hi the list. I got a probleme with R CMD check. I run a
c> package.skeleton on a very simple code (four toy
c> function, no probleme with the R CMD build). But on the R
c> CMD check packClassique, I get :
>> >> >>
c> * checking for unstated dependencies in R code
c> ... WARNING Erreur dans .find.package(package, lib.loc) :
c> aucun package nommé 'packClassique' n'est trouvé Calls:
c> <Anonymous> -> .find.package Excecution hated
c> [French=Exécution arrêtée]
>> >> >>
c> See the information on DESCRIPTION files in the chapter
c> 'Creating R packages' of the 'Writing R Extensions'
c> manual. * checking S3 generic/method consistency
c> ... WARNING Erreur dans .find.package(package, lib.loc) :
c> aucun package nommé 'packClassique' n'est trouvé Calls:
c> <Anonymous> -> .find.package Exécution arrêtée
>> >> >>
c> And so on.
>> >> >>
c> Any idea on what is wrong ?
>> >> >>
>> >> >> No.
>> >> >>
c> Christophe
>> >> >>
c> ______________________________________________
c> R-help at r-project.org mailing list
c> https://stat.ethz.ch/mailman/listinfo/r-help
>> >> >>
c> PLEASE do read the posting guide
>> >> >> http://www.R-project.org/posting-guide.html
c> and provide commented, minimal, self-contained,
c> reproducible code.
>> >> >>
>> >> >> Please do -- both --
>> >> >>
>> >> >> Martin
>> >> >>
>> >>
>> >>
>> >>
c> ----------------------------------------------------------------
c> Ce message a ete envoye par IMP, grace a l'Universite
c> Paris 10 Nanterre
>> >>
>> >>
>> >>
>>
>> ______________________________________________
c> R-help at r-project.org mailing list
c> https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do
c> read the posting guide
c> http://www.R-project.org/posting-guide.html and provide
c> commented, minimal, self-contained, reproducible code.
>>
>>
More information about the R-help
mailing list