On 01/13/2011 09:49 AM, steven mosher wrote: > > > On Thu, Jan 13, 2011 at 6:46 AM, Martin Morgan <mtmor...@fhcrc.org > <mailto:mtmor...@fhcrc.org>> wrote: > > On 01/12/2011 10:54 PM, steven mosher wrote: > > I have J Chambers wonderful text ( Software for data Analysis) and > I've been > > trying > > my hand at some very routine S4 OOP development. > > > > One of the things I was trying to do was to create some very basic S4 > > classes. The first > > was simply a class that had a data.frame as part of its > representation. > > > > setClass("df",representation(dirframe="data.frame")) > > > > The object basically contains a data.frame that represents a file > directory > > listing > > with a column named filename, size, time, etc. > > > > And then I have methods for doing various things with this object. > > > > I then tried to tackle the problem of coercing this S4 object to a > > data.frame. Again just a learning exercise. > > > > The goal would be able to make a call like this > > > > testFrame <- as.data.frame(x) > > > > where x, was an object of class "df" > > > > If I try to define "as.data.frame" as a S4 method, then I can make > it work, > > but I then destroy the S3 functionality > > of as.data.frame, so that if I were to try to coerce a matrix to a > > data.frame it would work. > > Hi Steven -- > > This works for me > > setClass("A", representation=representation(df="data.frame")) > > setMethod("as.data.frame", "A", > function(x, row.names=NULL, optional=FALSE, ...) > { > ## implementation, e.g., > callGeneric(x@df, row.names=row.names, optional=optional, ...) > }) > > > this makes no sense to me. > > Looking at this in the manual: > "A call to callGeneric can only appear inside a method definition. It > then results in a call to the current generic function. The value of > that call is the value of callGeneric. While it can be called from any > method, it is useful and typically used in methods for group generic > functions."
callGeneric is not necessary, and I should just have said as.data.frame(x@df, row.names=row.names, optional=optional, ...) > I'm further confused. what is the "current" generic function? I think of a generic 'foo' as a function, and inside that function are functions foo,A-method, foo,B-method etc for classes A, B, .... When in one of these methods, foo,A-method, the current generic is the function 'foo'. For as.data.frame, the generic is unambiguous ('as.data.frame' !) and the interesting case are the group generics (?Logic, for instance), where a single method setMethod("Logic", function(e1, e2) callGeneric(<YOUR CODE HERE>)) and you'll have in effect written 'methods' for all the operators defined in the 'Logic' group. Cool. > > > as.data.frame(new("A")) > Object of class "data.frame" > data frame with 0 columns and 0 rows > > as.data.frame(matrix(0, 3, 5)) > V1 V2 V3 V4 V5 > 1 0 0 0 0 0 > 2 0 0 0 0 0 > 3 0 0 0 0 0 > > > Maybe you call setGeneric (no need to, setMethod will promote > as.data.frame automatically) in a way that does not specify the default > (arg useAsDefault) correctly? > > > I think that may have been the mistake.. what do you mean by > no need to call setGeneric? The only code I had was what was written above -- setClass and setMethod. I could also have setGeneric("as.data.frame") and would have been ok -- the default behavior of setGeneric in this case is to make a generic function as.data.frame, AND a method as.data.frame,ANY-method. The as.data.frame,ANY-method is implemented as base::as.data.frame, and is where objects not handled by methods I implement might end up being dispatched to. In a new R session, try setGeneric("as.data.frame") showMethods(as.data.frame) selectMethod(as.data.frame, "ANY") If I had done > setGeneric("as.data.frame", function(x) standardGeneric("as.data.frame")) Creating a generic for 'as.data.frame' in package '.GlobalEnv' (the supplied definition differs from and overrides the implicit generic in package 'base': Formal arguments differ: (x), (x, row.names, optional, ...)) [1] "as.data.frame" then I'm in trouble -- I've created a generic 'as.data.frame', but since the signature of my generic differs from the signature of base::as.data.frame, the default behavior does NOT create a as.data.frame,ANY method. Not sure if this helps or not... Martin > > I really like chambers book, but there are certain parts where the lack > of simple examples > really makes it difficult to follow. > > > Martin > > > > > > > > So, I guess my question is what do I do, write an s3 method for > > as.data.frame that takes a "df" object as a paramter? > > The book wasn't exactly clear ( or I'm not that bright), or is > there a way > > to make the S4 method I wrote "as.data.frame" > > call the S3 method if needed? > > > > [[alternative HTML version deleted]] > > > > ______________________________________________ > > R-help@r-project.org <mailto:R-help@r-project.org> mailing list > > 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. > > > -- > Computational Biology > Fred Hutchinson Cancer Research Center > 1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109 > > Location: M1-B861 > Telephone: 206 667-2793 > > -- Computational Biology Fred Hutchinson Cancer Research Center 1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109 Location: M1-B861 Telephone: 206 667-2793 ______________________________________________ R-help@r-project.org mailing list 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.