how to use "..."

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
    require(doBy)
    df <-
summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun)
    return(df)
}
[SNIP EXAMPLE -- THANK YOU FOR IT]

Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
   require(doBy)
   df <- summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))), data=dat, FUN=stat_fun)
   return(df)
}

Example, works fine:
my_data <- structure(list(cat = structure(c(1L, 1L, 1L, 1L, 1L, 2L, 2L,
2L, 2L, 2L), .Label = c("A", "B"), class = "factor"), varnum = c(-0.754816565434373,
-1.94101630973709, -0.102461836059522, -0.519952759645808, -1.73772800855664,
-1.13939178585609, 0.522356715260142, -0.701428514907824, 1.45197576541159,
0.0844567413828095)), .Names = c("cat", "varnum"), row.names = c(NA,
-10L), class = "data.frame")
stat.group(dat=my_data, vec_cat=1, stat_fun=mean)


Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)


I know I can do it using this "..." but I have no idea how to do it. I've tried to search for it, but a search with "..." doesn't yield interesting results!

On Thu, Jan 17, 2013 at 2:36 PM, Ivan Calandra
<ivan.calandra at u-bourgogne.fr> wrote:

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
     require(doBy)
     df <-
summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun)
     return(df)
}
[SNIP EXAMPLE -- THANK YOU FOR IT]

Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)

Thanks for the great working examples!

It's actually not too hard here --  just pass "..." down as if it were
an argument and let summaryBy() do the hard work of actually handling
the dots:

stat.group <- function(dat, vec_cat, stat_fun, ...){
     require(doBy)
     df <- summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun, ...)
     return(df)
}

Also, note that as a matter of style, you can actually clean this up a
little bit: R follows the trend of many functional languages in
automatically returning the value of the last expression evaluated:

stat.group <- function(dat, vec_cat, stat_fun, ...){
     require(doBy)
     summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun, ...)
}

Cheers,
Michael

Ok, it is that simple... Actually I had tried it but messed up so that it
didn't work.
Do you know where I can find some documentation about it?

Regarding return(), I know that it's not necessary, but when the function
gets more complicated, I like to have it because it becomes clearer to me.

Thanks all!
Ivan

--
Ivan CALANDRA
Universit? de Bourgogne
UMR CNRS/uB 6282 Biog?osciences
6 Boulevard Gabriel
21000 Dijon, FRANCE
+33(0)3.80.39.63.06
ivan.calandra at u-bourgogne.fr
http://biogeosciences.u-bourgogne.fr/calandra

Le 17/01/13 15:55, R. Michael Weylandt a ?crit :

On Thu, Jan 17, 2013 at 2:36 PM, Ivan Calandra
<ivan.calandra at u-bourgogne.fr> wrote:

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
     require(doBy)
     df <-

summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun)
     return(df)
}
[SNIP EXAMPLE -- THANK YOU FOR IT]

Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)

Thanks for the great working examples!

It's actually not too hard here --  just pass "..." down as if it were
an argument and let summaryBy() do the hard work of actually handling
the dots:

stat.group <- function(dat, vec_cat, stat_fun, ...){
     require(doBy)
     df <-
summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun, ...)
     return(df)
}

Also, note that as a matter of style, you can actually clean this up a
little bit: R follows the trend of many functional languages in
automatically returning the value of the last expression evaluated:

stat.group <- function(dat, vec_cat, stat_fun, ...){
     require(doBy)

summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun, ...)
}

Cheers,
Michael

Ok, it is that simple... Actually I had tried it but messed up so that it
didn't work.
Do you know where I can find some documentation about it?

Regarding return(), I know that it's not necessary, but when the function
gets more complicated, I like to have it because it becomes clearer to me.

Thanks all!
Ivan

--
Ivan CALANDRA
Universit? de Bourgogne
UMR CNRS/uB 6282 Biog?osciences
6 Boulevard Gabriel
21000 Dijon, FRANCE
+33(0)3.80.39.63.06
ivan.calandra at u-bourgogne.fr
http://biogeosciences.u-bourgogne.fr/calandra

Le 17/01/13 15:55, R. Michael Weylandt a ?crit :

On Thu, Jan 17, 2013 at 2:36 PM, Ivan Calandra
<ivan.calandra at u-bourgogne.fr> wrote:

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
     require(doBy)
     df <-

summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun)
     return(df)
}
[SNIP EXAMPLE -- THANK YOU FOR IT]

Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)

Thanks for the great working examples!

It's actually not too hard here --  just pass "..." down as if it were
an argument and let summaryBy() do the hard work of actually handling
the dots:

stat.group <- function(dat, vec_cat, stat_fun, ...){
     require(doBy)
     df <-
summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun, ...)
     return(df)
}

Also, note that as a matter of style, you can actually clean this up a
little bit: R follows the trend of many functional languages in
automatically returning the value of the last expression evaluated:

stat.group <- function(dat, vec_cat, stat_fun, ...){
     require(doBy)

summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun, ...)
}

Cheers,
Michael

The ellipsis object is not listed in the base help pages!

help(`+`) # this works - help on arithmetic operators
help("+") # also works
help(`...`) # fails with Error: '...' used in an incorrect context
help("...") # fails also with No documentation for '...' in specified packages and libraries: you could try '??...'


-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Ivan Calandra
Sent: Friday, 18 January 2013 4:43a
To: R. Michael Weylandt
Cc: R list
Subject: Re: [R] how to use "..."

Do you know where I can find some documentation about it?

The "help" facility is applicable to functions and data sets.  It is not 
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the 
associated
rules of precedence).

    cheers,

        Rolf Turner

On 01/18/2013 09:17 AM, Steve Taylor wrote:

The ellipsis object is not listed in the base help pages!

help(`+`) # this works - help on arithmetic operators
help("+") # also works
help(`...`) # fails with Error: '...' used in an incorrect context
help("...") # fails also with No documentation for '...' in specified packages and libraries: you could try '??...'


-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Ivan Calandra
Sent: Friday, 18 January 2013 4:43a
To: R. Michael Weylandt
Cc: R list
Subject: =?utf-8?B?UmU6IFtSXSBob3cgdG8gdXNlICIuLi4i==?>> 
Do you know where I can find some documentation about it?

Rolf
Perhaps the philosophy of the help system needs to change . . .
John

Rolf
Perhaps the philosophy of the help system needs to change . . .

John

Sent from my iPhone

On Jan 17, 2013, at 7:11 PM, "Rolf Turner <rolf.turner at xtra.co.nz>" <rolf.turner at xtra.co.nz> wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

     cheers,

         Rolf Turner

On 01/18/2013 09:17 AM, Steve Taylor wrote:

The ellipsis object is not listed in the base help pages!

help(`+`) # this works - help on arithmetic operators
help("+") # also works
help(`...`) # fails with Error: '...' used in an incorrect context
help("...") # fails also with No documentation for '...' in specified packages and libraries: you could try '??...'


-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of Ivan Calandra
Sent: Friday, 18 January 2013 4:43a
To: R. Michael Weylandt
Cc: R list
Subject: =?utf-8?B?UmU6IFtSXSBob3cgdG8gdXNlICIuLi4i==?>>
Do you know where I can find some documentation about it?

Confidentiality Statement:
This email message, including any attachments, is for th...{{dropped:6}}

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Rolf Turner wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns to use help() and help.search() or their equivalents.  We don't need a manual to teach that.

Rolf Turner wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

Then why are there help pages ?":::", ?"for", ?"NA_character_", ?".C" and ?"??" but not ?"..." ?!

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns to use

help() and help.search() or their equivalents.  We don't need a manual to teach
that.

There are other parts of the help system that you don't appear to know
about.  Run help.start() for the main index page.  help() is kind of a
shortcut to the section labelled "Packages".  help.search() gets you to
"Search Engine & Keywords".  There are 15 other headings to explore as well.

Rolf Turner wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

Then why are there help pages ?":::", ?"for", ?"NA_character_", ?".C" and ?"??"

but not ?"..." ?!

Those are all functions or other objects, while ... is not.  There are
examples of topic aliases that are not objects, e.g. ?Syntax, but mostly
Rolf is right.  Most discussion of the syntax is in the manuals.

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns to use

help() and help.search() or their equivalents.  We don't need a manual to teach
that.

There are other parts of the help system that you don't appear to know
about.  Run help.start() for the main index page.  help() is kind of a
shortcut to the section labelled "Packages".  help.search() gets you to
"Search Engine & Keywords".  There are 15 other headings to explore as well.

I know about them even tho I didn't mention them.

Rolf Turner wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

Then why are there help pages ?":::", ?"for", ?"NA_character_", ?".C" and ?"??"

but not ?"..." ?!

Those are all functions or other objects, while ... is not.  There are
examples of topic aliases that are not objects, e.g. ?Syntax, but mostly
Rolf is right.  Most discussion of the syntax is in the manuals.

Nevertheless, help("...") is a reasonable thing for a user to do and expect some help.

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns to use

help() and help.search() or their equivalents.  We don't need a manual to teach
that.

There are other parts of the help system that you don't appear to know
about.  Run help.start() for the main index page.  help() is kind of a
shortcut to the section labelled "Packages".  help.search() gets you to
"Search Engine & Keywords".  There are 15 other headings to explore as well.

I know about them even tho I didn't mention them.

Rolf Turner wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

Then why are there help pages ?":::", ?"for", ?"NA_character_", ?".C" and ?"??"

but not ?"..." ?!

Those are all functions or other objects, while ... is not.  There are
examples of topic aliases that are not objects, e.g. ?Syntax, but mostly
Rolf is right.  Most discussion of the syntax is in the manuals.

Nevertheless, help("...") is a reasonable thing for a user to do and expect some help.

Why are the help pages not right? The ... construct is a fundamental
part of the language syntax. Information about this fundamental
construct should be easily available!

John

John David Sorkin M.D., Ph.D.
Chief, Biostatistics and Informatics
University of Maryland School of Medicine Division of Gerontology
Baltimore VA Medical Center
10 North Greene Street
GRECC (BT/18/GR)
Baltimore, MD 21201-1524
(Phone) 410-605-7119
(Fax) 410-605-7913 (Please call phone number above prior to faxing)>>>
Duncan Murdoch <murdoch.duncan at gmail.com> 1/20/2013 5:06 PM >>>
On 13-01-20 2:56 PM, Steve Taylor wrote:

 >> On 13-01-20 2:28 PM, Steve Taylor wrote:

 >>>> From: Duncan Murdoch
 >>>> Maybe we just need a manual on how to use the existing help

system.  But

 >>>> I suspect people who won't read the existing manuals won't read that
 >>>> one, either.

 >>>
 >>> Duncan, I assume you're being facetious.  Every R user soon learns

to use

 >> help() and help.search() or their equivalents.  We don't need a

manual to teach

 >> that.
 >>
 >> There are other parts of the help system that you don't appear to know
 >> about.  Run help.start() for the main index page.  help() is kind of a
 >> shortcut to the section labelled "Packages".  help.search() gets you to
 >> "Search Engine & Keywords".  There are 15 other headings to explore

as well.

 >
 > I know about them even tho I didn't mention them.
 >

 >>>

 >>>>> Rolf Turner wrote:

 >>>>>> The "help" facility is applicable to functions and data sets.

It is not

 >>>>>> designed
 >>>>>> or intended to give "help" with respect to R syntax (with the

exception of

 >>>>>> the basic syntax of the operators --- unary and binary --- and the
 >>>>>> associated
 >>>>>> rules of precedence).

 >>>
 >>> Then why are there help pages ?":::", ?"for", ?"NA_character_",

?".C" and ?"??"

 >> but not ?"..." ?!
 >>
 >> Those are all functions or other objects, while ... is not.  There are
 >> examples of topic aliases that are not objects, e.g. ?Syntax, but mostly
 >> Rolf is right.  Most discussion of the syntax is in the manuals.

 >
 > Nevertheless, help("...") is a reasonable thing for a user to do and

expect some help.

 >

I agree it should do something.  I think I'll have it lead to ?Reserved,
and add pointers to the manuals from there.  I don't think the help()
pages are right for it.

Duncan Murdoch

Why are the help pages not right? The ... construct is a fundamental part of the language syntax. Information about this fundamental construct should be easily available!

John


John David Sorkin M.D., Ph.D.
Chief, Biostatistics and Informatics
University of Maryland School of Medicine Division of Gerontology
Baltimore VA Medical Center
10 North Greene Street
GRECC (BT/18/GR)
Baltimore, MD 21201-1524
(Phone) 410-605-7119
(Fax) 410-605-7913 (Please call phone number above prior to faxing)>>> Duncan Murdoch <murdoch.duncan at gmail.com> 1/20/2013 5:06 PM >>>
On 13-01-20 2:56 PM, Steve Taylor wrote:

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns to use

help() and help.search() or their equivalents.  We don't need a manual to teach
that.

There are other parts of the help system that you don't appear to know
about.  Run help.start() for the main index page.  help() is kind of a
shortcut to the section labelled "Packages".  help.search() gets you to
"Search Engine & Keywords".  There are 15 other headings to explore as well.

I know about them even tho I didn't mention them.

Rolf Turner wrote:

The "help" facility is applicable to functions and data sets.  It is not
designed
or intended to give "help" with respect to R syntax (with the exception of
the basic syntax of the operators --- unary and binary --- and the
associated
rules of precedence).

Then why are there help pages ?":::", ?"for", ?"NA_character_", ?".C" and ?"??"

but not ?"..." ?!

Those are all functions or other objects, while ... is not.  There are
examples of topic aliases that are not objects, e.g. ?Syntax, but mostly
Rolf is right.  Most discussion of the syntax is in the manuals.

Nevertheless, help("...") is a reasonable thing for a user to do and expect some help.

I agree it should do something.  I think I'll have it lead to ?Reserved,
and add pointers to the manuals from there.  I don't think the help()
pages are right for it.

Duncan Murdoch

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help  
system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns  
to use help() and help.search() or their equivalents.  We don't  
need a manual to teach that.

There are other parts of the help system that you don't appear to  
know about.  Run help.start() for the main index page.  help() is  
kind of a shortcut to the section labelled "Packages".   
help.search() gets you to "Search Engine & Keywords".  There are 15  
other headings to explore as well.

On Jan 20, 2013, at 11:46 AM, Duncan Murdoch wrote:

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help
system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns
to use help() and help.search() or their equivalents.  We don't
need a manual to teach that.

There are other parts of the help system that you don't appear to
know about.  Run help.start() for the main index page.  help() is
kind of a shortcut to the section labelled "Packages".
help.search() gets you to "Search Engine & Keywords".  There are 15
other headings to explore as well.

help.search("...")

On my Mac after a very long wait it appears I get all of the functions
in every installed package that have a three-dots argument in their
argument list displayed in the Usage sections. That does not seem very
"help"-ful. I tried limiting with keyword="programming", that did not
appear effective. I wasn't able to figure out from the
help(help.search) page what I need to do to constrict to non-Usage
matches yet expand to items linked from the help.start() page.

If I just search with a browser for "..." in "Introduction to R" I get
what I consider to be useful material.

On 13-01-20 6:23 PM, David Winsemius wrote:

On Jan 20, 2013, at 11:46 AM, Duncan Murdoch wrote:

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help
system.  But
I suspect people who won't read the existing manuals won't read  
that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns
to use help() and help.search() or their equivalents.  We don't
need a manual to teach that.

There are other parts of the help system that you don't appear to
know about.  Run help.start() for the main index page.  help() is
kind of a shortcut to the section labelled "Packages".
help.search() gets you to "Search Engine & Keywords".  There are 15
other headings to explore as well.

help.search("...")

On my Mac after a very long wait it appears I get all of the  
functions
in every installed package that have a three-dots argument in their
argument list displayed in the Usage sections. That does not seem  
very
"help"-ful. I tried limiting with keyword="programming", that did not
appear effective. I wasn't able to figure out from the
help(help.search) page what I need to do to constrict to non-Usage
matches yet expand to items linked from the help.start() page.

I don't think you're reading the same "help.search" help page that  
I'm reading.  help.search("...") is pretty clearly documented to do  
a search for help pages with "alias", "concept" or "title" matching  
the regular expression "...", i.e. they have three characters in one  
of those.  You should have found every single help page.

I don't see anything there that suggests it does a search through  
the usage section.  Are you just making stuff up?

If I just search with a browser for "..." in "Introduction to R" I  
get
what I consider to be useful material.

Yes.  Other ways that should work, but don't:

- looking through the index to An Introduction to R.  (The index is  
very weak.)  Same for the R Language Definition.

- doing a Google search for "R ...".  Stupid Google.

Other ways that do work:

- Reading the tables of contents for An Introduction to R, or the R  
Language Definition.

- Reading those manuals.

- ?"..." in R-devel, after I commit the changes I described.

- Searching with grep through the source files named *.Rd or *.texi.  
(For this you need to have the source installed.)

I do think the help system is too fragmented, but I also believe it  
is better than it used to be.  I am sure that people would make  
false claims about it regardless of any improvements, and other  
people would complain about any changes, so it's not a priority for  
me to work on it any more.

Duncan Murdoch

On Jan 20, 2013, at 3:52 PM, Duncan Murdoch wrote:

On 13-01-20 6:23 PM, David Winsemius wrote:

On Jan 20, 2013, at 11:46 AM, Duncan Murdoch wrote:

On 13-01-20 2:28 PM, Steve Taylor wrote:

From: Duncan Murdoch
Maybe we just need a manual on how to use the existing help
system.  But
I suspect people who won't read the existing manuals won't read that
one, either.

Duncan, I assume you're being facetious.  Every R user soon learns
to use help() and help.search() or their equivalents.  We don't
need a manual to teach that.

There are other parts of the help system that you don't appear to
know about.  Run help.start() for the main index page. help() is
kind of a shortcut to the section labelled "Packages".
help.search() gets you to "Search Engine & Keywords". There are 15
other headings to explore as well.

help.search("...")

On my Mac after a very long wait it appears I get all of the functions
in every installed package that have a three-dots argument in their
argument list displayed in the Usage sections. That does not seem very
"help"-ful. I tried limiting with keyword="programming", that did not
appear effective. I wasn't able to figure out from the
help(help.search) page what I need to do to constrict to non-Usage
matches yet expand to items linked from the help.start() page.

I don't think you're reading the same "help.search" help page that 
I'm reading.  help.search("...") is pretty clearly documented to do a 
search for help pages with "alias", "concept" or "title" matching the 
regular expression "...", i.e. they have three characters in one of 
those.  You should have found every single help page.

I don't see anything there that suggests it does a search through the 
usage section.  Are you just making stuff up?

Apparently. I was constructing a theory to explain why my search was 
taking so long and returning so many pages. With your clarification 
above I tried a somewhat more successful search. I clearly missed 
reading the Description paragraph. New strategy:

help.search("\\.\\.\\.")

There are only 15 or so hits (could be more or less depending on the 
composition of the library packages)  and among them is 'dotsMethods'. 
That looks like a useful page for this topic.  The number of hits 
would only be half as large if the author of the "labeling" package 
had not put "..." at the end of 8 of his help page titles.

Sorry for the noise.

-- David.

If I just search with a browser for "..." in "Introduction to R" I get
what I consider to be useful material.

Yes.  Other ways that should work, but don't:

- looking through the index to An Introduction to R.  (The index is 
very weak.)  Same for the R Language Definition.

- doing a Google search for "R ...".  Stupid Google.

Other ways that do work:

- Reading the tables of contents for An Introduction to R, or the R 
Language Definition.

- Reading those manuals.

- ?"..." in R-devel, after I commit the changes I described.

- Searching with grep through the source files named *.Rd or *.texi. 
(For this you need to have the source installed.)

I do think the help system is too fragmented, but I also believe it 
is better than it used to be.  I am sure that people would make false 
claims about it regardless of any improvements, and other people 
would complain about any changes, so it's not a priority for me to 
work on it any more.

Duncan Murdoch

David Winsemius, MD
Alameda, CA, USA

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
     require(doBy)
     df <-
summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun)
     return(df)
}

Example, works fine:
my_data <- structure(list(cat = structure(c(1L, 1L, 1L, 1L, 1L, 2L, 2L,
2L, 2L, 2L), .Label = c("A", "B"), class = "factor"), varnum =
c(-0.754816565434373,
-1.94101630973709, -0.102461836059522, -0.519952759645808,
-1.73772800855664,
-1.13939178585609, 0.522356715260142, -0.701428514907824, 1.45197576541159,
0.0844567413828095)), .Names = c("cat", "varnum"), row.names = c(NA,
-10L), class = "data.frame")
stat.group(dat=my_data, vec_cat=1, stat_fun=mean)


Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)


I know I can do it using this "..." but I have no idea how to do it.
I've tried to search for it, but a search with "..." doesn't yield
interesting results!


Thank you in advance for your help!
Ivan

f1( x=1, z= 2:4)

f1( x=1, y= 2:4)  ## and
f1( 2:4)
f1( y=2:4) ## this will produce an error in f2

There is now a blog post that attempts to
answer the question in the subject line:

http://www.burns-stat.com/the-three-dots-construct-in-r/

Pat

On 17/01/2013 14:36, Ivan Calandra wrote:

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
     require(doBy)
     df <-

summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))),
data=dat, FUN=stat_fun)
     return(df)
}

Example, works fine:
my_data <- structure(list(cat = structure(c(1L, 1L, 1L, 1L, 1L, 2L, 2L,
2L, 2L, 2L), .Label = c("A", "B"), class = "factor"), varnum =
c(-0.754816565434373,
-1.94101630973709, -0.102461836059522, -0.519952759645808,
-1.73772800855664,
-1.13939178585609, 0.522356715260142, -0.701428514907824,
1.45197576541159,
0.0844567413828095)), .Names = c("cat", "varnum"), row.names = c(NA,
-10L), class = "data.frame")
stat.group(dat=my_data, vec_cat=1, stat_fun=mean)


Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)


I know I can do it using this "..." but I have no idea how to do it.
I've tried to search for it, but a search with "..." doesn't yield
interesting results!


Thank you in advance for your help!
Ivan

--
Patrick Burns
pburns at pburns.seanet.com
twitter: @burnsstat @portfolioprobe
http://www.portfolioprobe.com/blog
http://www.burns-stat.com
(home of:
 'Impatient R'
 'The R Inferno'
 'Tao Te Programming')

There is now a blog post that attempts to
answer the question in the subject line:

http://www.burns-stat.com/the-three-dots-construct-in-r/

Pat

On 17/01/2013 14:36, Ivan Calandra wrote:

Dear users,

I'm trying to learn how to use the "...".

I have written a function (simplified here) that uses doBy::summaryBy():
# 'dat' is a data.frame from which the aggregation is computed
# 'vec_cat' is a integer vector defining which columns of the data.frame
should be use on the right side of the formula
# 'stat_fun' is the function that will be run to aggregate
stat.group <- function(dat, vec_cat, stat_fun){
     require(doBy)
     df <-
summaryBy(as.formula(paste0(".~",paste0(names(dat)[vec_cat],collapse="+"))), 

data=dat, FUN=stat_fun)
     return(df)
}

Example, works fine:
my_data <- structure(list(cat = structure(c(1L, 1L, 1L, 1L, 1L, 2L, 2L,
2L, 2L, 2L), .Label = c("A", "B"), class = "factor"), varnum =
c(-0.754816565434373,
-1.94101630973709, -0.102461836059522, -0.519952759645808,
-1.73772800855664,
-1.13939178585609, 0.522356715260142, -0.701428514907824, 
1.45197576541159,
0.0844567413828095)), .Names = c("cat", "varnum"), row.names = c(NA,
-10L), class = "data.frame")
stat.group(dat=my_data, vec_cat=1, stat_fun=mean)


Now summaryBy() has an "..." argument and I would like to use it.
For example, I would like to be able to add the trim argument to my call
like this:
stat.group(dat=my_data, vec_cat=1, stat_fun=mean, trim=0.2)


I know I can do it using this "..." but I have no idea how to do it.
I've tried to search for it, but a search with "..." doesn't yield
interesting results!


Thank you in advance for your help!
Ivan