R help - May 2012 - Repost: Expressions returned by GlobalEnv functions and package functions

Hello,
(Sorry for the repost, i am resending in plain text)

I have a function 'ewrap' (see below for definition).
It takes 3 expressions and returns another expression e.g.

map <- ewrap({
    len <- length(r$addon)
    rhcollect(len,1)
  })

becomes:

expression({
    NULL
    result <- mapply(function(.index, k, r) {
        {
            len <- length(r$addon)
            rhcollect(len, 1)
        }
    }, 1:length(map.values), map.keys, map.values)
    NULL
})
attr(,"class")
[1] "expression" "rhmr-map"

ewrap is defined in the GlobalEnv. In my package (Rhipe), a function
rhwrap has the exact same definition

rhwap <-  function(co1=NULL,before=NULL,after=NULL){
  co <- substitute(co1); before=substitute(before)
  j <- as.expression(bquote({
    .(BE)
    result <- mapply(function(.index,k,r){
      .(CO)
    },1:length(map.values),map.keys,map.values)
    .(AF)
  },list(CO=co,BE=before,AF=after)))
  class(j) <- c(class(j),"rhmr-map")
  j
}

but the following two are different,

map <- ewrap({
    len <- length(r$addon)
    rhcollect(len,1)
  })

and

map2 <- rhwrap({
    len <- length(r$addon)
    rhcollect(len,1)
  })

(because serialize(map,NULL) != serialize(map2,NULL))

I guess this is because both functions(ewrap and rhwrap) return an
environment in which they are defined and in the case
of rhwrap this is the Rhipe package namespace/environment (i'm not
sure what jargon i should use here).

So my questions are:

1. how do i inspect the extra information that rhwrap is adding to its
return value
2. How do i remove this, so that it behaves like ewrap


Thanks in advance
Saptarshi

Hello Saptarshi,

  If two functions behave differently even though their definitions
are identical, then it is because they refer to variables that were in
scope when the functions were defined (i.e. lexical scoping). To
prevent such differences, either define both functions in the same
scope or have them read only arguments and local variables (i.e. pass
all needed input variables as arguments).

     Take care
     Oliver

On Thu, May 31, 2012 at 5:25 PM, Saptarshi Guha
<saptarshi.guha at gmail.com> wrote:
> Hello,
> (Sorry for the repost, i am resending in plain text)
>
> I have a function 'ewrap' (see below for definition).
> It takes 3 expressions and returns another expression e.g.
>
> map <- ewrap({
> ? ?len <- length(r$addon)
> ? ?rhcollect(len,1)
> ?})
>
> becomes:
>
> expression({
> ? ?NULL
> ? ?result <- mapply(function(.index, k, r) {
> ? ? ? ?{
> ? ? ? ? ? ?len <- length(r$addon)
> ? ? ? ? ? ?rhcollect(len, 1)
> ? ? ? ?}
> ? ?}, 1:length(map.values), map.keys, map.values)
> ? ?NULL
> })
> attr(,"class")
> [1] "expression" "rhmr-map"
>
> ewrap is defined in the GlobalEnv. In my package (Rhipe), a function
> rhwrap has the exact same definition
>
> rhwap <- ?function(co1=NULL,before=NULL,after=NULL){
> ?co <- substitute(co1); before=substitute(before)
> ?j <- as.expression(bquote({
> ? ?.(BE)
> ? ?result <- mapply(function(.index,k,r){
> ? ? ?.(CO)
> ? ?},1:length(map.values),map.keys,map.values)
> ? ?.(AF)
> ?},list(CO=co,BE=before,AF=after)))
> ?class(j) <- c(class(j),"rhmr-map")
> ?j
> }
>
> but the following two are different,
>
> map <- ewrap({
> ? ?len <- length(r$addon)
> ? ?rhcollect(len,1)
> ?})
>
> and
>
> map2 <- rhwrap({
> ? ?len <- length(r$addon)
> ? ?rhcollect(len,1)
> ?})
>
> (because serialize(map,NULL) != serialize(map2,NULL))
>
> I guess this is because both functions(ewrap and rhwrap) return an
> environment in which they are defined and in the case
> of rhwrap this is the Rhipe package namespace/environment (i'm not
> sure what jargon i should use here).
>
> So my questions are:
>
> 1. how do i inspect the extra information that rhwrap is adding to its
> return value
> 2. How do i remove this, so that it behaves like ewrap
>
>
> Thanks in advance
> Saptarshi
>
> ______________________________________________
> R-help at 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.


-- 
Oliver Ruebenacker
Bioinformatics Consultant (http://www.knowomics.com/wiki/Oliver_Ruebenacker)
Knowomics, The Bioinformatics Network (http://www.knowomics.com)
SBPAX: Turning Bio Knowledge into Math Models (http://www.sbpax.org)