[R] handling big data set in R

[ONKELINX, Thierry]

Dear Shu,

Why not store your dataset in a database? Then you can start each loop
by reading the submatrix you need for the analysis. This will require
much less memory. loops from the apply-family with work better than the
for loop.

HTH,

Thierry

------------------------------------------------------------------------
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek / Research Institute for Nature
and Forest
Cel biometrie, methodologie en kwaliteitszorg / Section biometrics,
methodology and quality assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium 
tel. + 32 54/436 185
Thierry.Onkelinx op inbo.be 
www.inbo.be 

Do not put your faith in what statistics say until you have carefully
considered what they do not say.  ~William W. Watt
A statistical analysis, properly conducted, is a delicate dissection of
uncertainties, a surgery of suppositions. ~M.J.Moroney

-----Oorspronkelijk bericht-----
Van: r-help-bounces op r-project.org [mailto:r-help-bounces op r-project.org]
Namens shu zhang
Verzonden: maandag 3 maart 2008 6:35
Aan: r-help op r-project.org
Onderwerp: [R] handling big data set in R

Hello R users,

I'm wondering whether it is possible to manage big data set in R? I
have a data set with  3 million rows and 3 columns (X,Y,Z), where X is
the group id. For each X, I need to run 2 regression on the submatrix.
I used the function "split":

datamatrix<-read.csv("datas.csv", header=F, sep=",")
dim(datamatrix)
# [1] 2980523  3
names(datamatrix)<-c("X","Y","Z")

attach(datamatrix)

subX<-split(X, X)
subY<-split(Y,X)
subZ<-split(Z,X)
n<-length(subdata)  ### number of groups
s1<-s2<-rep(NA, n)  ### vector to store the regression slope

for (i in 1:n){
  a<-table(Y[[i]])
  table.x<-as.numeric(names(a))
  table.y<-as.numeric(a)
  fit1<-lm(table.y~table.x)    ##### find the slope of the histogram of
y
  s1[i]<-fit$coefficients[2]

  fit2<-lm(subY[[i]]~subZ[[i]])  ####### regress y on z
  s2[i]<-fit$coefficients[2]
}


But my R died before completing the loop... (I've thought about doing
it in SAS, but I don't know how to write a loop combined with a PROC
REG...)

One thing that might be helpful is that my data set has already been
sorted based on X. I don't know whether this can be any helpful for
managing the dataset.

Any suggestion would be appreciated!


Thanks!
-Shu

______________________________________________
R-help op 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.