[R] Parallel computations using snow: how to combine boot objects?

BEER Michael michael.beer at unifr.ch
Fri Jan 21 10:35:00 CET 2005


Hello,

I'm trying to do some bootstrapping in a parallel environment (Linux
cluster) in order to estimate confidence intervals for a certain
parameter. Following the example in the documentation of the "snow"
package (http://www.stat.uiowa.edu/~luke/R/cluster/cluster.html), I
launch my computations by something like

> cl.nuke.boot <-
+             clusterCall(cl,boot,nuke.data, nuke.fun, R=500, m=1,
+                         fit.pred=new.fit, x.pred=new.data)

which gives me a list of n boot objects (where n is the number of nodes
in my cluster). So far, so good.

However, if I now want to go further, I need to combine all these boot
objects to a single one which I can pass to boot.ci for example. Is
there a recommended way to do this?

Thanks,
Michael




More information about the R-help mailing list