[Rd] LazyData: no / yes

Simon Urbanek simon.urbanek at r-project.org
Tue Nov 6 23:24:55 CET 2012


On Nov 6, 2012, at 5:13 PM, Christophe Genolini wrote:

> I get it from windows tack manager (under Window 7). I guess it is in K something.
> My point was not about "how big is my dataset" (anyway, it is a fake dataset, so it can be as big as I want) but more about "where on hell are lost the 52 760 - 39 668 K ?"
> :-)

Ask your system ;) - you're comparing wrong things: a) you didn't run garbage collection so there will be temporary objects around and b) see FAQ 7.42 why what you're looking at has no real meaning.

Cheers,
Simon



> Christophe
> 
>>  On Nov 6, 2012, at 1:03 PM, Christophe Genolini wrote:
>> 
>>> Hi the list
>>> 
>>> I have package foo0 with a big dataset 'myData'.
>>> In DESCRIPTION, if I use 'LazyData: no', then I get:
>>> 
>>> - when I open a R session : memory used=20 908
>>> - when I attach 'library(foo0)' : memory used=24364
>>> - then I load the set 'data(myData)' : memory used=39 668
>>> 
>>> If I use LazyData: yes', then I get
>>> - when I open a R session : memory used=20 908
>>> - when I attach 'library(foo0)' : memory used=52 760.
>>> 
>>> In this second example, after 'library(foo0)', I was expecting the memory to rize up to 39 668, not to 52 760... Where does the difference come from?
>>> 
>> What do you mean by "memory used" - i.e. where do you get that from? After GC?
>> This certainly doesn't look like a "big dataset" by the numbers - I would classify that as tiny :)
>> 
>> Cheers,
>> Simon
>> 
>> 
>>> Thanks
>>> Christophe
>>> 
>>> 
>>> -- 
>>> Christophe Genolini
>>> Maître de conférences en bio-statistique
>>> Vice président Communication interne et animation du campus
>>> Université Paris Ouest Nanterre La Défense
>>> 
>>> ______________________________________________
>>> R-devel at r-project.org mailing list
>>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>> 
>>> 
>> 
> 
> 
> -- 
> Christophe Genolini
> Maître de conférences en bio-statistique
> Vice président Communication interne et animation du campus
> Université Paris Ouest Nanterre La Défense
> 
> 



More information about the R-devel mailing list