R help - May 2020 - Bug in function arguments autocomplete and ellipsis?

Dear All,

I am wondering whether function arguments autocomplete causes a bug when
additional ellipsis arguments are used.

Example:
a = function(robot) {
    cat(robot, "\n")
 }
a(r=5) prints 5, meaning that r is autocompleted to ?robot?. Not sure if this is
normal behavior, but this does not cause problems.

This would be fine, but somehow causes problems when there is an additional
ellipsis argument that can be used to autocomplete an already existing argument.
In the example below, when we are calling sens.analysis.all, everything starting
with q is an additional argument (?). Now, k is missed completely. And that is
because there is another actual argument that starts with k ? key.legend.axes
(it is assigned to 4 instead). If ?k=4? is changed to ?kk=4? the problem
disappears.

sens.analysis.all <- function(func, outgrid, parvec, parmin, parmax,
                              length.pgrid, outvname, zlim, plabs,
                              gridlab, mylog, outlog=FALSE, ytick=NULL,
                              xline=NULL, yline=NULL, mypal=topo.colors,
                              key.legend.axes=NULL, plot.guidance=FALSE, ... ) {

   cat(..1, "\n")
   cat(..2, "\n")
   cat(..3, "\n")
 
   out=1
   out
} 

out = sens.analysis.all(numer.wait.times.wrps, NA,
     c(5.4, 0.008, 1.5), c(4.9, 0.0079, 2.0), c(5.5, 0.0090, 3.5),
       length.pgrid=10, outvname="cvs", zlim=c(0.3, 2),
       plabs=c("mu", "lambda", "b"),
gridlab="Soil Moisture [mm]",
       mylog=FALSE, outlog=FALSE, yline=2, plot.guidance=FALSE, q=3.1, k=4,
y.c=670,
       realgrid=seq(630, 670, by=4), myseed=0,
       nt=1000000, burnin=300000, bin.cutoff=500,
       bdW.prelim=prec.struct.Mal)

On 07/05/2020 3:46 a.m., Roman Olson wrote:
> Dear All,
> 
> I am wondering whether function arguments autocomplete causes a bug when
additional ellipsis arguments are used.
> 
> Example:
> a = function(robot) {
>      cat(robot, "\n")
>   }
> a(r=5) prints 5, meaning that r is autocompleted to ?robot?. Not sure if
this is normal behavior, but this does not cause problems.
> 
> This would be fine, but somehow causes problems when there is an additional
ellipsis argument that can be used to autocomplete an already existing argument.
In the example below, when we are calling sens.analysis.all, everything starting
with q is an additional argument (?). Now, k is missed completely. And that is
because there is another actual argument that starts with k ? key.legend.axes
(it is assigned to 4 instead). If ?k=4? is changed to ?kk=4? the problem
disappears.
> 
> sens.analysis.all <- function(func, outgrid, parvec, parmin, parmax,
>                                length.pgrid, outvname, zlim, plabs,
>                                gridlab, mylog, outlog=FALSE, ytick=NULL,
>                                xline=NULL, yline=NULL, mypal=topo.colors,
>                                key.legend.axes=NULL, plot.guidance=FALSE,
... ) {
> 
>     cat(..1, "\n")
>     cat(..2, "\n")
>     cat(..3, "\n")
>   
>     out=1
>     out
> }
> 
> out = sens.analysis.all(numer.wait.times.wrps, NA,
>       c(5.4, 0.008, 1.5), c(4.9, 0.0079, 2.0), c(5.5, 0.0090, 3.5),
>         length.pgrid=10, outvname="cvs", zlim=c(0.3, 2),
>         plabs=c("mu", "lambda", "b"),
gridlab="Soil Moisture [mm]",
>         mylog=FALSE, outlog=FALSE, yline=2, plot.guidance=FALSE, q=3.1,
k=4, y.c=670,
>         realgrid=seq(630, 670, by=4), myseed=0,
>         nt=1000000, burnin=300000, bin.cutoff=500,
>         bdW.prelim=prec.struct.Mal)
I'm afraid I don't understand what you find surprising here.  In your 
first example, the "r" gets attached to "robot" because of
partial
matching.  In the second example, the "k" gets attached to 
"key.legend.axes" for the same reason.  This is expected behaviour. 
If
you don't want to allow  partial matching to arguments, they need to be 
placed *after* the ellipsis.  Those arguments will need to be spelled 
out in full.

Duncan Murdoch

It would help if you consulted the docs, in this case, **The R
Language Definition** and, in particular, 4.3.2 on argument matching.
I won't repeat what it is there, but I believe it will suffice to
dispel your confusion.


Bert Gunter

"The trouble with having an open mind is that people keep coming along
and sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )

On Fri, May 8, 2020 at 7:49 AM Roman Olson <romanolson at yonsei.ac.kr>
wrote:
>
> Dear All,
>
> I am wondering whether function arguments autocomplete causes a bug when
additional ellipsis arguments are used.
>
> Example:
> a = function(robot) {
>     cat(robot, "\n")
>  }
> a(r=5) prints 5, meaning that r is autocompleted to ?robot?. Not sure if
this is normal behavior, but this does not cause problems.
>
> This would be fine, but somehow causes problems when there is an additional
ellipsis argument that can be used to autocomplete an already existing argument.
In the example below, when we are calling sens.analysis.all, everything starting
with q is an additional argument (?). Now, k is missed completely. And that is
because there is another actual argument that starts with k ? key.legend.axes
(it is assigned to 4 instead). If ?k=4? is changed to ?kk=4? the problem
disappears.
>
> sens.analysis.all <- function(func, outgrid, parvec, parmin, parmax,
>                               length.pgrid, outvname, zlim, plabs,
>                               gridlab, mylog, outlog=FALSE, ytick=NULL,
>                               xline=NULL, yline=NULL, mypal=topo.colors,
>                               key.legend.axes=NULL, plot.guidance=FALSE,
... ) {
>
>    cat(..1, "\n")
>    cat(..2, "\n")
>    cat(..3, "\n")
>
>    out=1
>    out
> }
>
> out = sens.analysis.all(numer.wait.times.wrps, NA,
>      c(5.4, 0.008, 1.5), c(4.9, 0.0079, 2.0), c(5.5, 0.0090, 3.5),
>        length.pgrid=10, outvname="cvs", zlim=c(0.3, 2),
>        plabs=c("mu", "lambda", "b"),
gridlab="Soil Moisture [mm]",
>        mylog=FALSE, outlog=FALSE, yline=2, plot.guidance=FALSE, q=3.1, k=4,
y.c=670,
>        realgrid=seq(630, 670, by=4), myseed=0,
>        nt=1000000, burnin=300000, bin.cutoff=500,
>        bdW.prelim=prec.struct.Mal)
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> 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.

Dear All,

Thanks for clarifying this. Now I know that this is expected behavior, and will
try to place ? before the rest of the arguments.

Stay home and stay safe!
-Roman
> 2020. 5. 9. ?? 1:24, Bert Gunter <bgunter.4567 at gmail.com> ??:
> 
> The R
> Language Definition

	[[alternative HTML version deleted]]