Hello,
I have 21 covariates that I would like to screen using GAM in xpose. I
tried using the extra file in xpose. I couldn't figure out how to
communicate with xpose and tell it to use the additional variables in
the extra file as covariates during GAM. Could someone comment on how to
use the extra file in xpose?
As a work around to the above problem; I tried to hard code step.gam
with 21 explanatory variables in R. I ran into a problem again with the
"Error: cannot allocate vector of size 437.9 Mb". This problem comes up
if the list of explanatory variables is longer than 17. I have to
comment out 4 of the variables and can't test more than 17 variables at
a time (see below). I am wondering if there is a work around this
problem or should I run the steps of a step.gam manually in R?
Please help,
Mahesh
gam.object <- gam(YVAR~1, data=gam.data)
step.object <-step.gam(gam.object, scope=list( "FAC1"=~1+FAC1,
"FAC2"=~1+FAC2, "FAC3"=~1+FAC3, "NUM1"=~1+NUM1+ns(NUM1,df=2),
"NUM2"=~1+NUM2+ns(NUM2,df=2), "NUM3"=~1+NUM3+ns(NUM3,df=2),
"NUM4"=~1+NUM4+ns(NUM4,df=2), "NUM5"=~1+NUM5+ns(NUM5,df=2),
"NUM6"=~1+NUM6+ns(NUM6,df=2), "NUM7"=~1+NUM7+ns(NUM7,df=2),
"NUM8"=~1+NUM8+ns(NUM8,df=2), "NUM9"=~1+NUM9+ns(NUM9,df=2),
"NUM10"=~1+NUM10+ns(NUM10,df=2), "NUM11"=~1+NUM11+ns(NUM11,df=2),
"NUM12"=~1+NUM12+ns(NUM12,df=2),
"NUM13"=~1+NUM13+ns(NUM13,df=2),"NUM14"=~1+NUM14+ns(NUM14,df=2)
#"NUM15"=~1+NUM15+ns(NUM15,df=2),
#"NUM16"=~1+NUM16+ns(NUM16,df=2),
#"NUM17"=~1+NUM17+ns(NUM17,df=2),
#"NUM18"=~1+NUM18+ns(NUM18,df=2)
))
Difficulty with gam and step.gam
Dear Mahesh,
The problem ("Error: cannot allocate vector of size
437.9 Mb") you are facing could be related to memory of R.
Though, I never used GAM function in R. I had same type of problem when I was
trying
to read 2GB size simulated dataset. May
be following solution might work.
You can increase the memory of R by
1. Right-click on your R program icon (the icon on your
desktop or in your programs directory). Select ``Properties'', and then select
the ``Shortcut'' tab. Look for the ``Target'' field and after the closing
quotes around the location of the R executible, add
--max-mem-size=500M
as shown in the figure below..
If you get
the error that R cannot allocate a vector of length x, close out of R and add
the following line to the ``Target'' field:
--max-vsize=500M
or as appropriate.
You can
always check to see how much memory R has available by typing at the R prompt
> round(memory.limit()/2^20, 2)
which gives
you the amount of available memory in MB.
2. If you have lower RAM capacity then you may
have to install additional RAM and then reinstall R in order to take advantage
of the additional capacity to the maximum amount of physical RAM you have
installed.
Best Regards,
Venkatesh Pilla Reddy
PhD Fellow,
Dept. of Pharmacokinetics, Toxicology and Targeting,
University of Groningen,
Antonius Deusinglaan 1,
9713 AV Groningen, The Netherlands
Office Tel: +31 50 363 7627
Mobile Phone: +31 61 988 3918
Room: (3211) 473
Email: [email protected]
--- On Tue, 24/8/10, Samtani, Mahesh [PRDUS] <[email protected]> wrote:
Quoted reply history
From: Samtani, Mahesh [PRDUS] <[email protected]>
Subject: [NMusers] Difficulty with gam and step.gam
To: [email protected]
Date: Tuesday, 24 August, 2010, 7:52 PM
Hello,
I have 21 covariates that I would like to screen using
GAM in xpose. I tried using the extra file in xpose. I couldn't figure out how
to communicate with xpose and tell it to use the additional variables in the
extra file as covariates during GAM. Could someone comment on how to use the
extra file in xpose?
As a work around to the above problem; I tried to hard
code step.gam with 21 explanatory variables in R. I ran into a problem again
with the "Error: cannot allocate vector of size 437.9 Mb". This problem
comes up if the list of explanatory variables is longer than 17. I have to
comment out 4 of the variables and can't test more than 17 variables at a time
(see below). I am wondering if there is a work around this problem or should I
run the steps of a step.gam manually in R?
Please help,
Mahesh
gam.object <- gam(YVAR~1, data=gam.data)
step.object <-step.gam(gam.object, scope=list(
"FAC1"=~1+FAC1, "FAC2"=~1+FAC2, "FAC3"=~1+FAC3,
"NUM1"=~1+NUM1+ns(NUM1,df=2), "NUM2"=~1+NUM2+ns(NUM2,df=2),
"NUM3"=~1+NUM3+ns(NUM3,df=2), "NUM4"=~1+NUM4+ns(NUM4,df=2),
"NUM5"=~1+NUM5+ns(NUM5,df=2), "NUM6"=~1+NUM6+ns(NUM6,df=2),
"NUM7"=~1+NUM7+ns(NUM7,df=2), "NUM8"=~1+NUM8+ns(NUM8,df=2),
"NUM9"=~1+NUM9+ns(NUM9,df=2),
"NUM10"=~1+NUM10+ns(NUM10,df=2), "NUM11"=~1+NUM11+ns(NUM11,df=2),
"NUM12"=~1+NUM12+ns(NUM12,df=2),
"NUM13"=~1+NUM13+ns(NUM13,df=2),"NUM14"=~1+NUM14+ns(NUM14,df=2)
#"NUM15"=~1+NUM15+ns(NUM15,df=2),
#"NUM16"=~1+NUM16+ns(NUM16,df=2),
#"NUM17"=~1+NUM17+ns(NUM17,df=2),
#"NUM18"=~1+NUM18+ns(NUM18,df=2)
))