R devel - May 2011 - Reference Classes copy(shallow=FALSE) unexpected behavior.

Dear all,

I''ve just discovered the ''Reference Classes''.
In a previous attempt ---a year ago--- to re-implement in a Object Oriented
fashion the AMORE package using S4 classes I strongly felt the need of such
capability. It''s great to have the Reference Classes now available.
Along
with the discovery of the Rcpp package, this new programming paradigm has
boosted my interest in rewriting that package.

Nevertheless, I have found a surprising behavior in the $copy(shallow=FALSE)
method. Let''s have a look at the results which I believe are
self-explanatories. The ".Data" field is a list which contains objects
from
the "Con" class  ---connections for what is worth---.
> p$copy(shallow=FALSE) -> k
> p$.Data
[[1]]
An object of class "Con"
<environment: 0x15a370c4>

[[2]]
An object of class "Con"
<environment: 0x15b6ed94>

[[3]]
An object of class "Con"
<environment: 0x15b4ba08>
> k$.Data
[[1]]
An object of class "Con"
<environment: 0x15a370c4>

[[2]]
An object of class "Con"
<environment: 0x15b6ed94>

[[3]]
An object of class "Con"
<environment: 0x15b4ba08>

As you can see the elements of both lists share the memory addresses while I
was expecting the shallow=FALSE to perform a deep copy and create copies of
those objects as well.

As I''m new with reference classes this causes me a lot of confussion.
Is
this a bug, is this a feature?

Any insights about the matter are welcome

Thanks in advance

-M

-- 


 Manuel Castejón Limas

 University of León,

 Spain

 

 manuel.castejon@unileon.es





	[[alternative HTML version deleted]]

2011/5/10 Manuel Castej?n Limas <manuel.castejon at
unileon.es>:
> Dear all,
>
> I've just discovered the 'Reference Classes'.
> In a previous attempt ---a year ago--- to re-implement in a Object Oriented
> fashion the AMORE package using S4 classes I strongly felt the need of such
> capability. It's great to have the Reference Classes now available.
Along
> with the discovery of the Rcpp package, this new programming paradigm has
> boosted my interest in rewriting that package.
>
> Nevertheless, I have found a surprising behavior in the
$copy(shallow=FALSE)
> method. Let's have a look at the results which I believe are
> self-explanatories. The ".Data" field is a list which contains
objects from
> the "Con" class ?---connections for what is worth---.
What sort of class is the Con class? S4 or reference?

Hadley


-- 
Assistant Professor / Dobelman Family Junior Chair
Department of Statistics / Rice University
http://had.co.nz/