Michal Burda
2015-Jan-14 07:42 UTC
[Rd] R CMD check: "..." used in a situation where it does not exist
Dear R developers,
when running R CMD check, the R Under development (unstable)
(2015-01-13 r67453) gives me the following NOTE:
cbind.fsets: possible error in list(...): ... used in a situation
where it does not exist
The file that causes this note contains:
cbind.fsets <- function(..., deparse.level = 1) {
dots <- list(...)
res <- NULL
resVars <- NULL
resSpecs <- NULL
for (i in seq_along(dots)) {
arg <- dots[[i]]
argName <- names(dots)[i]
if (!is.null(arg)) {
if (!is.fsets(arg)) {
stop("Function 'cbind.fsets' cannot bind arguments
that are not valid 'fsets' objects")
}
class(arg) <- setdiff(class(arg), 'fsets')
if (is.null(res)) {
resVars <- vars(arg)
resSpecs <- specs(arg)
res <- arg
} else {
resVarNames <- c(names(resVars), names(vars(arg)))
resVars <- c(resVars, vars(arg))
names(resVars) <- resVarNames
o1 <- matrix(0, nrow=nrow(resSpecs), ncol=ncol(specs(arg)))
o2 <- matrix(0, nrow=nrow(specs(arg)), ncol=ncol(resSpecs))
resSpecs <- rbind(cbind(resSpecs, o1),
cbind(o2, specs(arg)))
colnames(resSpecs) <- names(resVars)
rownames(resSpecs) <- names(resVars)
res <- cbind(res, arg)
}
}
}
return(fsets(res, resVars, resSpecs))
}
I want to avoid any NOTES prior submitting my package to CRAN, but I
cannot find any information what is wrong with my code and how to get
rid of that message during R CMD check process.
Does anybody of you know what is wrong with my code?
Thank you, in advance.
Best regards,
Michal Burda
junior researcher
IT4Innovations NSC
Institute for Research and Applications of Fuzzy Modeling |
University of Ostrava | IT4Innovations National Supercomputing
Center
30. dubna 22 | 701 03 Ostrava | Czech Republic
e-mail: michal.burda at osu.cz | web IRAFM: irafm.osu.cz | web
IT4I: it4i.cz | phone: +420 597 09 1450
Hadley Wickham
2015-Jan-14 12:47 UTC
[Rd] R CMD check: "..." used in a situation where it does not exist
I think this is bug in R CMD check code. I get a similar error:
rule: possible error in paste0(...): ... used in a situation where it
does not exist
for the simple:
rule <- function(..., pad = "-") {
if (nargs() == 0) {
title <- ""
} else {
title <- paste0(...)
}
width <- getOption("width") - nchar(title) - 1
message(title, " ", paste(rep(pad, width, collapse = "")))
}
where ... clearly does exist.
I skimmed the commits for the last couple of days - the most relevant
would seem to be this one by Michael Lawrence:
https://github.com/wch/r-source/commit/d3a42dabccb23462aa1e0a1edad00450f0a6131c
(but I might've missed others - many of the commit messages are not
terribly explanatory)
Hadley
On Wed, Jan 14, 2015 at 1:42 AM, Michal Burda <michal.burda at osu.cz>
wrote:> Dear R developers,
>
> when running R CMD check, the R Under development (unstable)
> (2015-01-13 r67453) gives me the following NOTE:
>
>
> cbind.fsets: possible error in list(...): ... used in a situation
> where it does not exist
>
>
> The file that causes this note contains:
>
>
> cbind.fsets <- function(..., deparse.level = 1) {
> dots <- list(...)
>
> res <- NULL
> resVars <- NULL
> resSpecs <- NULL
>
> for (i in seq_along(dots)) {
> arg <- dots[[i]]
> argName <- names(dots)[i]
>
> if (!is.null(arg)) {
> if (!is.fsets(arg)) {
> stop("Function 'cbind.fsets' cannot bind
arguments
> that are not valid 'fsets' objects")
> }
> class(arg) <- setdiff(class(arg), 'fsets')
> if (is.null(res)) {
> resVars <- vars(arg)
> resSpecs <- specs(arg)
> res <- arg
> } else {
> resVarNames <- c(names(resVars), names(vars(arg)))
> resVars <- c(resVars, vars(arg))
> names(resVars) <- resVarNames
>
> o1 <- matrix(0, nrow=nrow(resSpecs),
ncol=ncol(specs(arg)))
> o2 <- matrix(0, nrow=nrow(specs(arg)),
ncol=ncol(resSpecs))
> resSpecs <- rbind(cbind(resSpecs, o1),
> cbind(o2, specs(arg)))
> colnames(resSpecs) <- names(resVars)
> rownames(resSpecs) <- names(resVars)
> res <- cbind(res, arg)
> }
> }
> }
>
> return(fsets(res, resVars, resSpecs))
> }
>
>
>
> I want to avoid any NOTES prior submitting my package to CRAN, but I
> cannot find any information what is wrong with my code and how to get
> rid of that message during R CMD check process.
>
> Does anybody of you know what is wrong with my code?
>
> Thank you, in advance.
>
>
>
> Best regards,
>
>
> Michal Burda
> junior researcher
>
>
> IT4Innovations NSC
> Institute for Research and Applications of Fuzzy Modeling |
> University of Ostrava | IT4Innovations National Supercomputing
> Center
> 30. dubna 22 | 701 03 Ostrava | Czech Republic
> e-mail: michal.burda at osu.cz | web IRAFM: irafm.osu.cz | web
> IT4I: it4i.cz | phone: +420 597 09 1450
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
--
http://had.co.nz/
Duncan Murdoch
2015-Jan-14 13:30 UTC
[Rd] R CMD check: "..." used in a situation where it does not exist
On 14/01/2015 4:47 AM, Hadley Wickham wrote:> I think this is bug in R CMD check code. I get a similar error: > > rule: possible error in paste0(...): ... used in a situation where it > does not exist > > for the simple: > > rule <- function(..., pad = "-") { > if (nargs() == 0) { > title <- "" > } else { > title <- paste0(...) > } > width <- getOption("width") - nchar(title) - 1 > message(title, " ", paste(rep(pad, width, collapse = ""))) > } > > where ... clearly does exist. > > I skimmed the commits for the last couple of days - the most relevant > would seem to be this one by Michael Lawrence: > https://github.com/wch/r-source/commit/d3a42dabccb23462aa1e0a1edad00450f0a6131c > (but I might've missed others - many of the commit messages are not > terribly explanatory)Yes, it's a temporary bug in R-devel, not a problem in Michal's code. Duncan Murdoch