Thank you very much! I'm using the first example.

For others: there are two simple typos in Henrik's email below:  pretty sure
y <- this$.; should be  y <- this$.y;

Thanks again...

Corrected below...

Ben

> Hello,
>
> I'd like to 'do work' on data members upon construction (i.e. without
> implementing it in a get method). Is this the best way to create data
member
> 'z' upon construction? I'm thinking if .z=paste(x,y) below gets more
complex
> I'll run into issues.
>
> setConstructorS3("MyClass", function(x=NA,y=NA,...) {
>  this <- extend(Object(), "MyClass",
>    .x=x,
>    .y=y,
>    .z=paste(x,y)
>  )

Looks good to me and is standard R procedure where you work with the
*arguments*.  You can also work on the object after it's been
instantiated, e.g.

 setConstructorS3("MyClass", function(x=NA,y=NA,...) {
  this <- extend(Object(), "MyClass",
   .x=x,
   .y=y,
   .z=NULL
  )

 # Assign z
 this$.z <- paste(this$.x, this$.y);

 this;
})

Note that if .z always a function of .x and .y it is redundant.  Then
an alternative is to create it on the fly, i.e.

setMethodS3("getZ", "MyClass", function(this, ...) {
  x <- this$.x;
 y <- this$.y;
 z <- paste(x, y);
 z;
})

You can then make it clever and cache the results so that it is only
calculated once, e.g.

setMethodS3("getZ", "MyClass", function(this, ...) {
  z <- this$.z;
 if (is.null(z)) {
   x <- this$.x;
   y <- this$.y;
   z <- paste(x, y);
   this$.z <- z;
 }
 z;
})

However, you then have to make sure to reset z (this$.z <- NULL)
whenever .x or .y is changed.

/Henrik




On Thu, Sep 22, 2011 at 10:18 AM, Henrik Bengtsson <h...@biostat.ucsf.edu>wrote:

> On Thu, Sep 22, 2011 at 9:06 AM, Ben qant <ccqu...@gmail.com> wrote:
> > Hello,
> >
> > I'd like to 'do work' on data members upon construction (i.e. without
> > implementing it in a get method). Is this the best way to create data
> member
> > 'z' upon construction? I'm thinking if .z=paste(x,y) below gets more
> complex
> > I'll run into issues.
> >
> > setConstructorS3("MyClass", function(x=NA,y=NA,...) {
> >  this <- extend(Object(), "MyClass",
> >    .x=x,
> >    .y=y,
> >    .z=paste(x,y)
> >  )
>
> Looks good to me and is standard R procedure where you work with the
> *arguments*.  You can also work on the object after it's been
> instantiated, e.g.
>
>  setConstructorS3("MyClass", function(x=NA,y=NA,...) {
>   this <- extend(Object(), "MyClass",
>    .x=x,
>    .y=y,
>    .z=NULL
>   )
>
>  # Assign z
>  this$.z <- paste(this$.x, this$.y);
>
>  this;
> })
>
> Note that if .z always a function of .x and .y it is redundant.  Then
> an alternative is to create it on the fly, i.e.
>
> setMethodS3("getZ", "MyClass", function(this, ...) {
>   x <- this$.x;
>  y <- this$.;
>  z <- paste(x, y);
>  z;
> })
>
> You can then make it clever and cache the results so that it is only
> calculated once, e.g.
>
> setMethodS3("getZ", "MyClass", function(this, ...) {
>   z <- this$.z;
>  if (is.null(z)) {
>    x <- this$.x;
>    y <- this$.;
>    z <- paste(x, y);
>    this$.z <- z;
>  }
>  z;
> })
>
> However, you then have to make sure to reset z (this$.z <- NULL)
> whenever .x or .y is changed.
>
> /Henrik
>
> >
> > })
> > setMethodS3("getX", "MyClass", function(this, ...) {
> >  this$.x;
> > })
> > setMethodS3("getY", "MyClass", function(this, ...) {
> >  this$.y;
> > })
> > setMethodS3("getZ", "MyClass", function(this, ...) {
> >  this$.z;
> > })
> >
> >> mc = MyClass('a','b')
> >> mc$x
> > [1] "a"
> >> mc$y
> > [1] "b"
> >> mc$z
> > [1] "a b"
> >
> >
> > Thanks,
> >
> > ben
> >
> >        [[alternative HTML version deleted]]
> >
> > ______________________________________________
> > 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.
> >
>

        [[alternative HTML version deleted]]

______________________________________________
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.

Reply via email to