similar to: Rmpi task-pull

I am running R 2.7.0 on a Suse 9.1 linux cluster with a job scheduler dispatching jobs and openmpi-1.0.1. I have tried running one of the examples at http://ace.acadiau.ca/math/ACMMaC/Rmpi/examples.html in Rmpi and they seem to be working, except mpi.close.Rslaves() hangs. The slaves are closed, but the master doesn't finish its script. Below is the example script and the call to R. The job is

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I've written a function that uses Rmpi to perform a calculation in parallel. It works fine, but I'm trying to improve efficiency in terms of memory usage and the amount of data being passed back and forth between mater and slaves. Calculations are performed on a symmetrical matrix in order to zero-out some of the cells. In the parallel

Hi everybody! I have the next code which makes a reduction of the *a *variable in two slaves, using the Rmpi package. library(Rmpi) mpi.spawn.Rslaves(nslaves=2) reduc<-function(){ a<-mpi.comm.rank()+2 mpi.reduce(a,type=2, op="prod") return(paste("a=",a)) } mpi.bcast.Robj2slave(reduc) mpi.remote.exec(reduc()) cat("Product: ")

Hi everybody! I have the next code which makes a reduction of the *a *variable in two slaves, using the Rmpi package. library(Rmpi) mpi.spawn.Rslaves(nslaves=2) reduc<-function(){ a<-mpi.comm.rank()+2 mpi.reduce(a,type=2, op="prod") return(paste("a=",a)) } mpi.bcast.Robj2slave(reduc) mpi.remote.exec(reduc()) cat("Product: ")

I'm invoking R from withing a shell script like this R --no-save --no-restore --gui=none > `hostname` 2>&1 <<BYE # various commands here BYE I would like to regain control from the invoking terminal at some point. I tried source(stdin()) but got a syntax error, presumably stdin is the little shell here snippet (the part between <<BYE and BYE). Is there some way to

Dear all, I have used the Rmpi package many times before however this time I've installed it as I always do with openMPI tar.gz file direct from the website. I'm installing on my ubuntu 8.10. Linux 2.6.27-11-generic #1 SMP Thu Jan 29 19:28:32 UTC 2009 x86_64 GNU/Linux All i get is: > library(Rmpi) Segmentation fault :~$ Which dumps me back into the shell, and doens't give me much

R-users, My question is related to earlier posts about benefits of quadcore over dualcore computers; I am trying to setup a cluster of windows xp computers so that eventually I could make use of 10-20 cpu:s, but for learning how to do this, I am playing around with two laptops. I thought that the package snow would come handy in this situation, but to use snow, I would probably need to install

Hello, I try to write parts of my code in C to accelerate the for-loops. But basic operations I want to do in R (e.g. start cluster). My R code looks something like this: library(Rmpi) mpi.spawn.Rslaves() mpi.remote.exec(....) dyn.load("test.so") erg <- .Call("test", ....) .... mpi.close.Rslaves() mpi.quit() And my C function looks something like this: #include

Hi I am trying to utilize my dual core processor (and later a High-performance clusters (HPC) ) by using the Rmpi, snow, snowfall, ... packages, but I am struggling at the beginning, i.e. to initialise the "cluster" on my dual core computer. Whenever I try to initialize it (via sfInit(parallel=TRUE, cpus=2) or mpi.spawn.Rslaves(nslaves=2) ), I get an error message: >

Hi all, I'm handling massive data.frames and matrices in R (30000 x 400). In the 1st column, say, I have 0s and 1s indicating rows that matter; other columns have probability values. One simple task I would like to do would be to get the column mean values for signaled rows (the ones with 1) As a very fresh "programmer" I have build a simple function in R which should not be very

Dear R list, In the course of learning to work with Rmpi, we are confused about a few points. The following simple program is based on some examples we retrieved from the web. Each slave is writing the same output line multiple times (a multiple equal to the number of slaves). In other words, the write statements are being executed a number of times equal to the number of slaves. I am

Hello, the package snow is not working on a windows cluster with MPICH2 and Rmpi. There is an error in makeCluster: launch failed: CreateProcess(/usr/bin/env "RPROG="C:\Programme\R\R-2.5.1\bin\R" "OUT=/dev/null" "R_LIBS=" C:/Programme/R/R-2.5.1/library/snow/RMPInode.sh) on 'cl1' failed, error 3 - Das System kann den angegbenen Pfad nicht finden. I

Hi, I am using a cluster with LAM 7.1.3/MPI 2 and R 2.6.0. Rmpi version 0.5-5 is working very well. Now I have tested "Rmpi 0.5-6". During spawning the Rslaves I get an error: MPI_Error_string: error spawning process > sessionInfo() R version 2.6.0 (2007-10-03) x86_64-unknown-linux-gnu locale:

I'm aiming to compare the workings of Rmpi and multicore on a duel processor quad core machine with 64 bit R-2.11.1 Kubuntu 10.4. It's impossible for me to get a small reproducable code segment to show what I mean, but if I show what works for mclapply, I'd hope it's possible to be shown what would be the equivalent way with mpi.apply. The function lr.gbm has variables trees,

I have just installed Rmpi on a Suse 9.1 linux cluster with openmpi-1.0.1. I am trying the example included below from the tutorial website. However, I keep getting the following error: > # Load the R MPI package if it is not already loaded. > if (!is.loaded("mpi_initialize")) { + library("Rmpi") + } > > # Spawn as many slaves as possible >

Dear R People: I'm having some trouble with mpiexec and Rmpi. I would like to be able to pass in the number of "children" via the mpiexec command (from the command line). this is in SUSE10.1, with R-2.7.1 Here are my files: cat eb.R library(Rmpi) mpi.remote.exec(paste("i am",mpi.comm.rank(),"of",mpi.comm.size())) mpi.quit() hodgesse at

Hello: Anybody knows how to run Rmpi through PBS (Portable Batch System) on a cluster computer. I'm using a supercomputer which require to submit jobs to PBS queue for dispatching. I tried use mpirun in my PBS script. But all my Rslaves are spawned to the same node. This is not desired. Any suggestions are welcome! Thanks in advance. ======================================== Shengqiao Li

Hi all, I'm having difficulties passing data from the master to the children. I haven't seen any complete tutorial, all the tutorial broadcast data where I would like to send parts of my big tablechip list. Everytime I run this, the last line always jams. I even tried to reduce the number of rows in data to 10 and still doesn't work. I even tried to do a remote.exec(ls()) and

Hi all, now I'm testing R in a "virtual cluster", made it with VirtualBox. This one has 3 nodes, running CentOS 5 and OpenMPI 1.2.8, and the principal node (called "server") exports the /home to other nodes. I have installed R and OpenMPI in /home, in fact, it seems work OK. Editing the openmpi-default-hostfile and run "mpirun -np 3 hostname" I can see the

I'm just curious if anyone else has had problems with this configuration. I added the CRAN repository to apt and installed 2.5.0 with apt-get. I then did an install.packages("Rmpi") on cluster nodes. Rmpi loads and lamhosts() shows the nodes, but mpi.spawn.Rslaves() fails (something to do with temp files?). Rmpi works fine with the Edgy-native version of R (2.3.x) and installing