[Rd] exporting s3 and s4 methods

Thomas Lumley tlumley at u.washington.edu
Wed Mar 18 17:42:31 CET 2009


I'm still having problems getting a package to define both S3 and S4 methods for the same new generic, on a current r-devel (version 48144).

Symptoms
> example(bigglm)

bigglm> data(trees)

bigglm> ff<-log(Volume)~log(Girth)+log(Height)

bigglm> a <- bigglm(ff,data=trees, chunksize=10, sandwich=TRUE)
Error in UseMethod("bigglm", data) : no applicable method for "bigglm"

> showMethods("bigglm")
Function: bigglm (package biglm)
formula="ANY", data="ANY"
formula="ANY", data="DBIConnection"
formula="formula", data="data.frame"
     (inherited from: formula="ANY", data="ANY")

> bigglm          
standardGeneric for "bigglm" defined from package "biglm"

function (formula, data, family = gaussian(), ...) 
standardGeneric("bigglm")
<environment: 0x8554240>
Methods may be defined for arguments: formula, data
Use  showMethods("bigglm")  for currently available ones.
> getMethod("bigglm",c("ANY","ANY"))
Method Definition (Class "derivedDefaultMethod"):

function (formula, data, family = gaussian(), ...) 
UseMethod("bigglm", data)
<environment: namespace:biglm>

Signatures:
         formula data 
target  "ANY"   "ANY"
defined "ANY"   "ANY"

>  methods("bigglm")
[1] bigglm.data.frame* bigglm.function*   bigglm.RODBC*

    Non-visible functions are asterisked
Warning messages:
1: In findGeneric(generic.function, parent.frame()) :
   'bigglm' is a formal generic function; S3 methods will not likely be found
2: In methods("bigglm") : function 'bigglm' appears not to be generic

In the NAMESPACE file I have
import(stats)
useDynLib(biglm)
importClassesFrom(DBI)
exportMethods(bigglm)
export(biglm)
export(bigglm)
S3method(bigglm,data.frame)
S3method(bigglm,"function")
S3method(bigglm, RODBC)

and in the code
bigglm<-function(formula, data, family=gaussian(),...)
     UseMethod("bigglm", data)
setGeneric("bigglm", signature=c("formula","data"))


bigglm.data.frame<-function(formula, data, ..., chunksize=5000){ <snip>

setMethod("bigglm",
            c("ANY","DBIConnection"),
            function(formula, data, family = gaussian(),
                                    tablename, ..., chunksize=5000){
              terms<-terms(formula)
              modelvars<-all.vars(formula)
      <snip>


Any suggestions?

     -thomas

Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle



More information about the R-devel mailing list