[R-pkg-devel] "invalid 'envir' argument" note from R-devel
peter dalgaard
pdalgd at gmail.com
Tue Jul 14 09:40:42 CEST 2015
The danger of minimal examples is that you risk working around only one aspect.
The ones Alexandra had (from the sensR package) didn't (all?) involve data(). A couple of them were like the one below, in which the FORTRANesque style of naming the function the same as the variable with the eventual return value is a bit curious, but hardly illegal.
Then again, the version of sensR that she linked to hadn't had the namespace issues sorted, so there was quite a bit of "no visible global" messages as well.
twoAFC <- function() {
twoAFC <- binomial()
twoAFC$link <- "Link for the 2-AFC test"
twoAFC$linkinv <- function(eta) {
ok <- eta > 0
eta[!ok] <- 0.5
eta[ok] <- pnorm(eta[ok] / sqrt(2))
pmin(pmax(eta, 0.5), 1) ## restrict to [0.5, 1] - just to be sure
}
twoAFC$mu.eta <- function(eta) {
ok <- eta >= 0
eta[!ok] <- 0
if(any(ok)) {
sqrt.2 <- sqrt(1/2)
eta[ok] <- dnorm(eta[ok] * sqrt.2) * sqrt.2
}
pmax(eta, 0) ## gradient cannot be negative.
}
twoAFC$linkfun <- function(mu) {
ok <- mu > 0.5 & mu < 1
mu[mu <= 0.5] <- 0
mu[mu >= 1] <- Inf
mu[ok] <- sqrt(2) * qnorm(mu[ok])
pmax(mu, 0) ## delta cannot be negative
}
twoAFC
}
-pd
(In a hotel room in Aarhus, so not going to spend time on this for a couple of days.)
> On 14 Jul 2015, at 01:01 , Duncan Murdoch <murdoch.duncan at gmail.com> wrote:
>
> It happened in r68597. In my example, I hadn't imported the data()
> function from utils; when I did that, it was fixed.
>
> Not sure why the error is complaining about the envir argument.
>
> Duncan Murdoch
>
> On 13/07/2015 6:47 PM, John Fox wrote:
>> Dear Duncan and Peter,
>>
>> I've just arrived at more or less the same thing:
>>
>> foo <- function() data(package="MASS")
>>
>> bar <- function() data(package="MASS", envir=.GlobalEnv)
>>
>> baz <- function() data(package="MASS", envir=globalenv())
>>
>> all trigger the note when included with the Rcmdr sources:
>>
>> * checking R code for possible problems ... NOTE
>> bar: Error while checking: invalid 'envir' argument
>> baz: Error while checking: invalid 'envir' argument
>> foo: Error while checking: invalid 'envir' argument
>>
>> The envir argument to data() defaults to .GlobaEnv .
>>
>> I hope this helps,
>> John
>>
>>> -----Original Message-----
>>> From: Duncan Murdoch [mailto:murdoch.duncan at gmail.com]
>>> Sent: July-13-15 6:32 PM
>>> To: John Fox; 'peter dalgaard'
>>> Cc: r-package-devel at r-project.org
>>> Subject: Re: [R-pkg-devel] "invalid 'envir' argument" note from R-devel
>>>
>>> On 13/07/2015 5:23 PM, John Fox wrote:
>>>> Dear Peter,
>>>>
>>>>> -----Original Message-----
>>>>> From: peter dalgaard [mailto:pdalgd at gmail.com]
>>>>> Sent: July-13-15 4:52 PM
>>>>> To: John Fox
>>>>> Cc: r-package-devel at r-project.org
>>>>> Subject: Re: [R-pkg-devel] "invalid 'envir' argument" note from R-
>>> devel
>>>>>
>>>>> Yes, there was a similar note from Alexandra Kuznetsova a couple of
>>> days
>>>>
>>>> Sorry, I didn't notice that.
>>>>
>>>>> ago. Look unintentional, but it is not easy to spot what triggers it.
>>> If
>>>>> someone could cook up a minimal example, or - maybe easier given the
>>>>> relatively short timeframe - bisect their way to the exact svn
>>> revision
>>>>> that triggered it, it might help in getting it fixed.
>>>>
>>>> I'm afraid that I'm not set up to build R-devel and I'm about to leave
>>> town
>>>> for three weeks. I'll see if I can produce a simpler example
>>> triggering the
>>>> error, however.
>>>
>>> Hana Sevcikova posted a simple example. I'll bisect on it.
>>>
>>> Here's her example:
>>>
>>> e <- new.env()
>>> data("mydataset", envir=e)
>>>
>>> I've substituted USArrests for "mydataset".
>>>
>>> Duncan Murdoch
>>>
>>>>
>>>> Thanks for this,
>>>> John
>>>>
>>>>>
>>>>> -pd
>>>>>
>>>>>> On 13 Jul 2015, at 22:31 , John Fox <jfox at mcmaster.ca> wrote:
>>>>>>
>>>>>> Dear list members,
>>>>>>
>>>>>> I'm getting a new note from R-devel that I haven't seen before when
>>>>> checking
>>>>>> the development version of the Rcmdr package:
>>>>>>
>>>>>> * checking R code for possible problems ... NOTE
>>>>>> readDataFromPackage: Error while checking: invalid 'envir' argument
>>>>>>
>>>>>> This note doesn't appear in R 3.2.1.
>>>>>>
>>>>>> My session info:
>>>>>>
>>>>>> -------- snip ----------
>>>>>>
>>>>>> R Under development (unstable) (2015-07-12 r68650)
>>>>>> Platform: x86_64-w64-mingw32/x64 (64-bit)
>>>>>> Running under: Windows 7 x64 (build 7601) Service Pack 1
>>>>>>
>>>>>> locale:
>>>>>> [1] LC_COLLATE=English_Canada.1252 LC_CTYPE=English_Canada.1252
>>>>>> LC_MONETARY=English_Canada.1252
>>>>>> [4] LC_NUMERIC=C LC_TIME=English_Canada.1252
>>>>>>
>>>>>> attached base packages:
>>>>>> [1] stats graphics grDevices utils datasets methods base
>>>>>>
>>>>>> loaded via a namespace (and not attached):
>>>>>> [1] tools_3.3.0
>>>>>>
>>>>>> -------- snip ----------
>>>>>>
>>>>>> I can't even localize the problem in readDataFromPackage(). There
>>> are
>>>>> only
>>>>>> two places in this function where there's a function call with an
>>>>> envir
>>>>>> argument, and I still get the note if I comment these out. As well,
>>>>>> readDataFromPackage() seems to work as intended -- there is no
>>> obvious
>>>>> error
>>>>>> in it.
>>>>>>
>>>>>> FWIW, here's readDataFromPackage(). The complete sources for the
>>>>> development
>>>>>> version of the Rcmdr package are on R-Forge.
>>>>>>
>>>>>> -------- snip ----------
>>>>>>
>>>>>> readDataFromPackage <- function() {
>>>>>> env <- environment()
>>>>>> datasets <- NULL
>>>>>> initializeDialog(title=gettextRcmdr("Read Data From Package"))
>>>>>> dsname <- tclVar("")
>>>>>> package <- NULL
>>>>>> enterFrame <- tkframe(top)
>>>>>> entryDsname <- ttkentry(enterFrame, width="20",
>>>>> textvariable=dsname)
>>>>>> packages <- sort(.packages())
>>>>>> packages <- packages[! packages %in% c("base", "stats")]
>>>>>> packages <- packages[sapply(packages, function(package){
>>>>>> ds <-
>>>>>> data(package=package)$results
>>>>>> if (nrow(ds) == 0)
>>>>>> return(FALSE)
>>>>>> ds <- ds[, "Item"]
>>>>>> valid <- sapply(ds,
>>>>>> is.valid.name)
>>>>>> length(ds[valid]) > 0
>>>>>> })]
>>>>>> packageDatasetFrame <- tkframe(top)
>>>>>> packageFrame <- tkframe(packageDatasetFrame)
>>>>>> max.height <- getRcmdr("variable.list.height")
>>>>>> packageBox <- tklistbox(packageFrame, height=min(max.height,
>>>>>> length(packages)),
>>>>>> exportselection="FALSE",
>>>>>> selectmode="single", background="white")
>>>>>> packageScroll <- ttkscrollbar(packageFrame,
>>>>>> command=function(...) tkyview(packageBox, ...))
>>>>>> tkconfigure(packageBox, yscrollcommand=function(...)
>>>>>> tkset(packageScroll, ...))
>>>>>> for (p in packages) tkinsert(packageBox, "end", p)
>>>>>> datasetFrame <- tkframe(packageDatasetFrame)
>>>>>> datasetBox <- tklistbox(datasetFrame, height=max.height,
>>>>>> exportselection="FALSE",
>>>>>> selectmode="single", background="white")
>>>>>> datasetScroll <- ttkscrollbar(datasetFrame,
>>>>>> command=function(...) tkyview(datasetBox, ...))
>>>>>> tkconfigure(datasetBox, yscrollcommand=function(...)
>>>>>> tkset(datasetScroll, ...))
>>>>>> onPackageSelect <- function(){
>>>>>> assign("package",
>>>>>> packages[as.numeric(tkcurselection(packageBox)) + 1], envir=env)
>>>>>> datasets <<- data(package=package)$results[,3]
>>>>>> valid <- sapply(datasets, is.valid.name)
>>>>>> datasets <<- datasets[valid]
>>>>>> tkdelete(datasetBox, "0", "end")
>>>>>> for (dataset in datasets) tkinsert(datasetBox, "end",
>>>>>> dataset)
>>>>>> tkconfigure(datasetBox, height=min(max.height,
>>>>>> length(datasets)))
>>>>>> firstChar <- tolower(substr(datasets, 1, 1))
>>>>>> len <- length(datasets)
>>>>>> onLetter <- function(letter){
>>>>>> letter <- tolower(letter)
>>>>>> current <- 1 +
>>>>>> round(as.numeric(unlist(strsplit(tclvalue(tkyview(datasetBox) ), "
>>>>>> "))[1])*len)
>>>>>> mat <- match(letter, firstChar[-(1:current)])
>>>>>> if (is.na(mat)) return()
>>>>>> tkyview.scroll(datasetBox, mat, "units")
>>>>>> }
>>>>>> onA <- function() onLetter("a")
>>>>>> onB <- function() onLetter("b")
>>>>>> onC <- function() onLetter("c")
>>>>>> onD <- function() onLetter("d")
>>>>>> onE <- function() onLetter("e")
>>>>>> onF <- function() onLetter("f")
>>>>>> onG <- function() onLetter("g")
>>>>>> onH <- function() onLetter("h")
>>>>>> onI <- function() onLetter("i")
>>>>>> onJ <- function() onLetter("j")
>>>>>> onK <- function() onLetter("k")
>>>>>> onL <- function() onLetter("l")
>>>>>> onM <- function() onLetter("m")
>>>>>> onN <- function() onLetter("n")
>>>>>> onO <- function() onLetter("o")
>>>>>> onP <- function() onLetter("p")
>>>>>> onQ <- function() onLetter("q")
>>>>>> onR <- function() onLetter("r")
>>>>>> onS <- function() onLetter("s")
>>>>>> onT <- function() onLetter("t")
>>>>>> onU <- function() onLetter("u")
>>>>>> onV <- function() onLetter("v")
>>>>>> onW <- function() onLetter("w")
>>>>>> onX <- function() onLetter("x")
>>>>>> onY <- function() onLetter("y")
>>>>>> onZ <- function() onLetter("z")
>>>>>> for (letter in c(letters, LETTERS)){
>>>>>> tkbind(datasetBox, paste("<", letter, ">", sep=""),
>>>>>> get(paste("on", toupper(letter),
>>>>>> sep="")))
>>>>>> }
>>>>>> onClick <- function() tkfocus(datasetBox)
>>>>>> tkbind(datasetBox, "<ButtonPress-1>", onClick)
>>>>>> }
>>>>>> onDatasetSelect <- function(){
>>>>>> tclvalue(dsname) <-
>>>>>> datasets[as.numeric(tkcurselection(datasetBox)) + 1]
>>>>>> }
>>>>>> firstChar <- tolower(substr(packages, 1, 1))
>>>>>> len <- length(packages)
>>>>>> onLetter <- function(letter){
>>>>>> letter <- tolower(letter)
>>>>>> current <- 1 +
>>>>>> round(as.numeric(unlist(strsplit(tclvalue(tkyview(packageBox) ), "
>>>>>> "))[1])*len)
>>>>>> mat <- match(letter, firstChar[-(1:current)])
>>>>>> if (is.na(mat)) return()
>>>>>> tkyview.scroll(packageBox, mat, "units")
>>>>>> }
>>>>>> onA <- function() onLetter("a")
>>>>>> onB <- function() onLetter("b")
>>>>>> onC <- function() onLetter("c")
>>>>>> onD <- function() onLetter("d")
>>>>>> onE <- function() onLetter("e")
>>>>>> onF <- function() onLetter("f")
>>>>>> onG <- function() onLetter("g")
>>>>>> onH <- function() onLetter("h")
>>>>>> onI <- function() onLetter("i")
>>>>>> onJ <- function() onLetter("j")
>>>>>> onK <- function() onLetter("k")
>>>>>> onL <- function() onLetter("l")
>>>>>> onM <- function() onLetter("m")
>>>>>> onN <- function() onLetter("n")
>>>>>> onO <- function() onLetter("o")
>>>>>> onP <- function() onLetter("p")
>>>>>> onQ <- function() onLetter("q")
>>>>>> onR <- function() onLetter("r")
>>>>>> onS <- function() onLetter("s")
>>>>>> onT <- function() onLetter("t")
>>>>>> onU <- function() onLetter("u")
>>>>>> onV <- function() onLetter("v")
>>>>>> onW <- function() onLetter("w")
>>>>>> onX <- function() onLetter("x")
>>>>>> onY <- function() onLetter("y")
>>>>>> onZ <- function() onLetter("z")
>>>>>> for (letter in c(letters, LETTERS)){
>>>>>> tkbind(packageBox, paste("<", letter, ">", sep=""),
>>>>>> get(paste("on", toupper(letter), sep="")))
>>>>>> }
>>>>>> onClick <- function() tkfocus(packageBox)
>>>>>> tkbind(packageBox, "<ButtonPress-1>", onClick)
>>>>>> onOK <- function(){
>>>>>> datasetName <-
>>>>>> datasets[as.numeric(tkcurselection(datasetBox)) + 1]
>>>>>> dsnameValue <- tclvalue(dsname)
>>>>>> if (dsnameValue != "" && is.null(package)){
>>>>>> closeDialog()
>>>>>> if (is.element(dsnameValue, listDataSets())) {
>>>>>> if ("no" ==
>>>>>> tclvalue(checkReplace(dsnameValue, gettextRcmdr("Data set")))){
>>>>>> if (GrabFocus()) tkgrab.release(top)
>>>>>> tkdestroy(top)
>>>>>> readDataFromPackage()
>>>>>> return()
>>>>>> }
>>>>>> }
>>>>>> save.options <- options(warn=2)
>>>>>> check <- try(eval(parse(text=logger(paste("data(",
>>>>>> dsnameValue, ")", sep=""))),
>>>>>> envir=.GlobalEnv),
>>>>>> silent=TRUE)
>>>>>> options(save.options)
>>>>>> if (class(check) == "try-error"){
>>>>>> errorCondition(recall=readDataFromPackage,
>>>>>>
>>>>>> message=sprintf(gettextRcmdr("Data set %s does not exit"),
>>>>> dsnameValue))
>>>>>> return()
>>>>>> }
>>>>>> activeDataSet(dsnameValue)
>>>>>> tkfocus(CommanderWindow())
>>>>>> }
>>>>>> else{
>>>>>> if (is.null(package)) {
>>>>>> errorCondition(recall=readDataFromPackage,
>>>>>> message=gettextRcmdr("You must select a package."))
>>>>>> return()
>>>>>> }
>>>>>> if (length(datasetName) == 0) {
>>>>>> errorCondition(recall=readDataFromPackage,
>>>>>> message=gettextRcmdr("You must select a data set.") )
>>>>>> return()
>>>>>> }
>>>>>> if (is.element(datasetName, listDataSets())) {
>>>>>> if ("no" ==
>>>>>> tclvalue(checkReplace(datasetName, gettextRcmdr("Data set")))){
>>>>>> if (GrabFocus()) tkgrab.release(top)
>>>>>> tkdestroy(top)
>>>>>> readDataFromPackage()
>>>>>> return()
>>>>>> }
>>>>>> }
>>>>>> closeDialog()
>>>>>> command <- paste("data(", datasetName, ',
>>>>>> package="', package, '")', sep="")
>>>>>> result <- justDoIt(command)
>>>>>> logger(command)
>>>>>> if (class(result)[1] != "try-error")
>>>>>> activeDataSet(datasetName)
>>>>>> tkfocus(CommanderWindow())
>>>>>> }
>>>>>> }
>>>>>> onDataHelp <- function(){
>>>>>> datasetName <- datasets[as.numeric(tkcurselection(datasetBox)) +
>>> 1]
>>>>>> dsnameValue <- tclvalue(dsname)
>>>>>> if (dsnameValue == "") dsnameValue <- datasetName
>>>>>> if (length(dsnameValue) == 0) Message(gettextRcmdr("No data
>>>>>> set selected."), type="warning")
>>>>>> else if (is.null(package)) doItAndPrint(paste('help("',
>>>>>> dsnameValue, '")', sep=""))
>>>>>> else doItAndPrint(paste('help("', dsnameValue, '",
>>>>>> package="', package, '")', sep=""))
>>>>>> }
>>>>>> OKCancelHelp(helpSubject="data")
>>>>>> dataHelpButton <- buttonRcmdr(top, text=gettextRcmdr("Help on
>>>>>> selected data set"), command=onDataHelp)
>>>>>> tkgrid(labelRcmdr(packageDatasetFrame, text=gettextRcmdr("Package
>>>>>> (Double-click to select)"), fg=getRcmdr("title.color"),
>>>>>> font="RcmdrTitleFont"),
>>>>>> labelRcmdr(packageDatasetFrame, text=" "),
>>>>>> labelRcmdr(packageDatasetFrame, text=gettextRcmdr("Data set (Double-
>>>>> click to
>>>>>> select)"),
>>>>>> fg=getRcmdr("title.color"),
>>>>>> font="RcmdrTitleFont"), sticky="w")
>>>>>> tkgrid(packageBox, packageScroll, sticky="nw")
>>>>>> tkgrid(datasetBox, datasetScroll, sticky="nw")
>>>>>> tkgrid(packageFrame, labelRcmdr(packageDatasetFrame, text=" "),
>>>>>> datasetFrame, sticky="nw")
>>>>>> tkgrid(packageDatasetFrame, sticky="w")
>>>>>> tkgrid(labelRcmdr(top, text=gettextRcmdr("OR"), fg="red"),
>>>>>> sticky="w")
>>>>>> tkgrid(labelRcmdr(enterFrame, text=gettextRcmdr("Enter name of data
>>>>>> set: "), fg=getRcmdr("title.color"), font="RcmdrTitleFont"),
>>>>> entryDsname,
>>>>>> sticky="w")
>>>>>> tkgrid(enterFrame, sticky="w")
>>>>>> tkgrid(dataHelpButton, sticky="w")
>>>>>> tkgrid(buttonsFrame, sticky="ew")
>>>>>> tkgrid.configure(packageScroll, sticky="ns")
>>>>>> tkgrid.configure(datasetScroll, sticky="ns")
>>>>>> tkbind(packageBox, "<Double-ButtonPress-1>", onPackageSelect)
>>>>>> tkbind(datasetBox, "<Double-ButtonPress-1>", onDatasetSelect)
>>>>>> dialogSuffix(focus=entryDsname)
>>>>>> }
>>>>>>
>>>>>> -------- snip ----------
>>>>>>
>>>>>> Any insight into the problem would be appreciated.
>>>>>>
>>>>>> Thanks,
>>>>>> John
>>>>>>
>>>>>> -----------------------------------------------
>>>>>> John Fox, Professor
>>>>>> McMaster University
>>>>>> Hamilton, Ontario, Canada
>>>>>> http://socserv.socsci.mcmaster.ca/jfox/
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> ---
>>>>>> This email has been checked for viruses by Avast antivirus software.
>>>>>> https://www.avast.com/antivirus
>>>>>>
>>>>>> ______________________________________________
>>>>>> R-package-devel at r-project.org mailing list
>>>>>> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>>>>>
>>>>> --
>>>>> Peter Dalgaard, Professor,
>>>>> Center for Statistics, Copenhagen Business School
>>>>> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
>>>>> Phone: (+45)38153501
>>>>> Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>>
>>>> ---
>>>> This email has been checked for viruses by Avast antivirus software.
>>>> https://www.avast.com/antivirus
>>>>
>>>> ______________________________________________
>>>> R-package-devel at r-project.org mailing list
>>>> https://stat.ethz.ch/mailman/listinfo/r-package-devel
>>>>
>>
>>
>> ---
>> This email has been checked for viruses by Avast antivirus software.
>> https://www.avast.com/antivirus
>>
>
--
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Email: pd.mes at cbs.dk Priv: PDalgd at gmail.com
More information about the R-package-devel
mailing list