Message-ID: <4993201E.8000509@idi.ntnu.no>
Date: 2009-02-11T18:59:42Z
From: Wacek Kusnierczyk
Subject: Variables captured in closures get copied?
In-Reply-To: <49930870.3010409@biostat.ku.dk>
Peter Dalgaard wrote:
> Titus von der Malsburg wrote:
>> Hi list! I have a data frame called fix and a list of index vectors
>> called rois:
>>
>> > head(rois, 3)
>> [[1]]
>> [1] 2 1
>>
>> [[2]]
>> [1] 3
>>
>> [[3]]
>> [1] 6 7 28 26 27 24 25
>>
>> The part that's causing the issue is the following line:
>>
>> lapply(rois, function(roi) fix$x[roi] <- 100)
>>
>> So for every index vector I'd like to set the respective entries in the
>> data frame (fix) to 100.
>>
>> I expected the data frame would be changed after lapply but instead it
>> remains unchanged. I understand that when I pass an argument into a
>> function it gets passed as a value and not as a reference. But here fix
>> is not an argument but captured in the closure. Do my questions are:
>> What's going on here and what is the idiomatic way of achieving my goal?
>>
>
> It's a local variable in the function. Not in principle different from
>
> function(roi) { fix <- fix ; ... }
>
> You could use superassignment (<<-), but a simpler idiom is
>
> for (roi in rois) fix$x[roi] <- 100
>
interesting. i'm not sure if this is something one should consider
obvious, though it does make sense.
actually, it seems that r gets caught by surprise on such sort of semantics:
d = data,frame(x=1:10^8)
(function() d$x[1] = 0)()
prints 'Error: cannot allocate vector of size 762.9 Mb' and stops
responding with 100% cpu (sometimes even 101, as reported by top, hehe)
occupied by r.
platform i686-pc-linux-gnu
arch i686
os linux-gnu
system i686, linux-gnu
status
major 2
minor 8.0
year 2008
month 10
day 20
svn rev 46754
language R
version.string R version 2.8.0 (2008-10-20)
vQ