R help - Aug 2011 - Ignoring loadNamespace errors when loading a file

On a Unix machine I ran caret::rfe using the multicore package, and I 
saved the resulting object using save(lm2, file = "lm2.RData").  
[Reproducible example below.]

When I try to load("lm2.RData") on my Windows laptop, I get

Error in loadNamespace(name) : there is no package called 'multicore'

I completely understand the error and I would like to ignore it and 
still load the saved object data.

I imagine that I can make myself an empty multicore package for Windows 
and load the data file successfully, but is there another way?

(I am not going to use any multicore functionality; I just want to 
inspect some of the data stored in the object and the reference in 
question is just an unfortunately stored link from the original call.)

(I did search for this question in the archives, but could only find it 
discussed in connection with starting R with a .RData file where the 
consensus seems to be to start R in vanilla mode and install the missing 
package.  This situation is different and installing a Unix-only package 
on Windows is obviously a non-starter, except as I proposed above.)

Obligatory reproducible example: On the Unix machine do

library("multicore")
a <- list(data = 1:10, fun = mclapply)
save(a, file = "a.RData")

and then try to load the "a.RData" file on Windows.  The question is
if
I can recover the data (1:10) on that platform.

Allan
>  sessionInfo()
R version 2.13.1 (2011-07-08)
Platform: x86_64-pc-mingw32/x64 (64-bit)

locale:
[1] LC_COLLATE=English_United Kingdom.1252
[2] LC_CTYPE=English_United Kingdom.1252
[3] LC_MONETARY=English_United Kingdom.1252
[4] LC_NUMERIC=C
[5] LC_TIME=English_United Kingdom.1252

attached base packages:
[1] compiler  stats     graphics  grDevices utils     datasets  methods
[8] base

other attached packages:
[1] caret_4.98      cluster_1.14.0  reshape_0.8.4   plyr_1.6
[5] lattice_0.19-31 boot_1.3-2      ctv_0.7-3

loaded via a namespace (and not attached):
[1] grid_2.13.1  tools_2.13.1

On 22.08.2011 08:52, Allan Engelhardt wrote:
> On a Unix machine I ran caret::rfe using the multicore package, and I
> saved the resulting object using save(lm2, file = "lm2.RData").
> [Reproducible example below.]
>
> When I try to load("lm2.RData") on my Windows laptop, I get
>
> Error in loadNamespace(name) : there is no package called
'multicore'
>
> I completely understand the error and I would like to ignore it and
> still load the saved object data.
>
> I imagine that I can make myself an empty multicore package for Windows
> and load the data file successfully, but is there another way?
>
> (I am not going to use any multicore functionality; I just want to
> inspect some of the data stored in the object and the reference in
> question is just an unfortunately stored link from the original call.)
>
> (I did search for this question in the archives, but could only find it
> discussed in connection with starting R with a .RData file where the
> consensus seems to be to start R in vanilla mode and install the missing
> package. This situation is different and installing a Unix-only package
> on Windows is obviously a non-starter, except as I proposed above.)
>
> Obligatory reproducible example: On the Unix machine do
>
> library("multicore")
> a <- list(data = 1:10, fun = mclapply)
> save(a, file = "a.RData")
>
> and then try to load the "a.RData" file on Windows. The question
is if I
> can recover the data (1:10) on that platform.
Short answer: No.

The element "fun" within "a" relies on the Namespace being
present. And
then you want additionally just part of an object. If this is really 
relevant, you have to digg into ./src/main/saveload.R and serialize.R 
and try to write your own interface to extract parts of objects. So this 
is probably not worth the effort.

Uwe Ligges


> Allan
>
>> sessionInfo()
> R version 2.13.1 (2011-07-08)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United Kingdom.1252
> [2] LC_CTYPE=English_United Kingdom.1252
> [3] LC_MONETARY=English_United Kingdom.1252
> [4] LC_NUMERIC=C
> [5] LC_TIME=English_United Kingdom.1252
>
> attached base packages:
> [1] compiler stats graphics grDevices utils datasets methods
> [8] base
>
> other attached packages:
> [1] caret_4.98 cluster_1.14.0 reshape_0.8.4 plyr_1.6
> [5] lattice_0.19-31 boot_1.3-2 ctv_0.7-3
>
> loaded via a namespace (and not attached):
> [1] grid_2.13.1 tools_2.13.1
>
> ______________________________________________
> 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.

On 08/21/2011 11:52 PM, Allan Engelhardt wrote:
> On a Unix machine I ran caret::rfe using the multicore package, and I
> saved the resulting object using save(lm2, file = "lm2.RData").
> [Reproducible example below.]
>
> When I try to load("lm2.RData") on my Windows laptop, I get
>
> Error in loadNamespace(name) : there is no package called
'multicore'
>
> I completely understand the error and I would like to ignore it and
> still load the saved object data.
>
> I imagine that I can make myself an empty multicore package for Windows
> and load the data file successfully, but is there another way?
>
> (I am not going to use any multicore functionality; I just want to
> inspect some of the data stored in the object and the reference in
> question is just an unfortunately stored link from the original call.)
>
> (I did search for this question in the archives, but could only find it
> discussed in connection with starting R with a .RData file where the
> consensus seems to be to start R in vanilla mode and install the missing
> package. This situation is different and installing a Unix-only package
> on Windows is obviously a non-starter, except as I proposed above.)
>
> Obligatory reproducible example: On the Unix machine do
>
> library("multicore")
> a <- list(data = 1:10, fun = mclapply)
> save(a, file = "a.RData")
>
> and then try to load the "a.RData" file on Windows. The question
is if I
> can recover the data (1:10) on that platform.
Is this a more realistic reproducible example (from ?rfe, modified to 
use computeFunction=mclapply)?

   data(BloodBrain)

   x <- scale(bbbDescr[,-nearZeroVar(bbbDescr)])
   x <- x[, -findCorrelation(cor(x), .8)]
   x <- as.data.frame(x)

   set.seed(1)
   lmProfile <- rfe(x, logBBB,
                    sizes = c(2:25, 30, 35, 40, 45, 50, 55, 60, 65),
                    rfeControl = rfeControl(functions = lmFuncs,
                      number = 5,
                      computeFunction=mclapply))

Maybe provide a computeFunction that only indirectly references mclapply

   computeFunction=function(...) {
       if (require(multicore)) mclapply(...)
       else lapply(...)
   }

or editing the result object to remove references to mclapply

   lmProfile$control$computeFunction <- NULL

Martin

>
> Allan
>
>> sessionInfo()
> R version 2.13.1 (2011-07-08)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United Kingdom.1252
> [2] LC_CTYPE=English_United Kingdom.1252
> [3] LC_MONETARY=English_United Kingdom.1252
> [4] LC_NUMERIC=C
> [5] LC_TIME=English_United Kingdom.1252
>
> attached base packages:
> [1] compiler stats graphics grDevices utils datasets methods
> [8] base
>
> other attached packages:
> [1] caret_4.98 cluster_1.14.0 reshape_0.8.4 plyr_1.6
> [5] lattice_0.19-31 boot_1.3-2 ctv_0.7-3
>
> loaded via a namespace (and not attached):
> [1] grid_2.13.1 tools_2.13.1
>
> ______________________________________________
> 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.

-- 
Computational Biology
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109

Location: M1-B861
Telephone: 206 667-2793

On 22/08/11 12:26, Martin Morgan wrote:
> On 08/21/2011 11:52 PM, Allan Engelhardt wrote:
>> [...]
>> Obligatory reproducible example: On the Unix machine do
>>
>> library("multicore")
>> a <- list(data = 1:10, fun = mclapply)
>> save(a, file = "a.RData")
>>
>> and then try to load the "a.RData" file on Windows. The
question is if I
>> can recover the data (1:10) on that platform.
>
> Is this a more realistic reproducible example (from ?rfe, modified to 
> use computeFunction=mclapply)?
>
>   data(BloodBrain)
>
>   x <- scale(bbbDescr[,-nearZeroVar(bbbDescr)])
>   x <- x[, -findCorrelation(cor(x), .8)]
>   x <- as.data.frame(x)
>
>   set.seed(1)
>   lmProfile <- rfe(x, logBBB,
>                    sizes = c(2:25, 30, 35, 40, 45, 50, 55, 60, 65),
>                    rfeControl = rfeControl(functions = lmFuncs,
>                      number = 5,
>                      computeFunction=mclapply))
>
> Maybe provide a computeFunction that only indirectly references mclapply
>
>   computeFunction=function(...) {
>       if (require(multicore)) mclapply(...)
>       else lapply(...)
>   }
>
> [...]
Yes, that workaround works for my usage.  Thanks!

Allan