[R] [R-1.4.0] minimum spanning tree of large ontology

[Luca.Toldo@merck.de]

Dear all,
I have an ontology which I would like to load in R to then compute the
minimum distance between elements.
Since this ontology is not only DAG, the problem is non trivial and
therefore decided to use R with the powerful e1071
library (allShortestPaths)  to do that.  That method requires to provide a
distance matrix between the
various objects.  My ontology contains 32768 objects and therefore I have
to load a matrix of 32768x32768.

Values in the matrix are set to 1 for those pair of terms that have a link
(semantic or hyper/hypo-nimic).

I am trying to reproduce at least part of the infrastructure described in
the following work
|--------------------------------------------------------------------------------|
|   1. Sugato Basu, Raymond J. Mooney, Krupakar V. Pasupuleti and Joydeep Ghosh  |
|      Evaluating the Novelty of Text-Mined Rules using Lexical Knowledge        |
|      Submitted to the Seventh ACM SIGKDD International Conference on Knowledge |
|      Discovery and Data Mining (KDD-2001), 2001                                |
|--------------------------------------------------------------------------------|


Unfortunately, when I try to do that I get into the reported problem
(cannot allocate vector of length 1073741824).
I am operating on an Irix Origin 2000 with 24 processors and 16 GB of
shared memory. My process has no limits
set to the amount of memory it can use.

I have noticed in the mailing as well as in the FAQ that since version >1.2
the memory management has been changed
and I got the information that somebody already successfully managed in
loading a 2GB dataset in previous versions.

I would be delighted to learn your opinion and instructions on how to
proceed.

Regards and compliments

Dr. Luca Toldo
Bioinformatician at Merck KGaA, Darmstadt

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._