我试图将下面的bigrf示例与doMPI后端并行化。
# Libraries
library(doMPI)
library(bigrf)
# Load data
data(Cars93, package="MASS")
y <- Cars93$Type
x <- Cars93
# Parallelization
cl <- startMPIcluster(count=2, verbose=TRUE)
exportDoMPI(cl, c("big.matrix"))
registerDoMPI(cl)
# Select variables with which to train model.
vars <- c(4:22)
# Run model, grow 50 trees on the first 60 examples.
forest1 <- bigrfc(x[1:60, ], y[1:60], ntree=50L, varselect=vars)
# Stop the Cluster
closeCluster(cl)
然而，看起来有些线程看不到＆＃34; modaC＆＃34;，bigrf包的C函数，给出以下错误，
task 2 failed - ""modaC" not available for .Call() for package "bigrf""
回溯
7 stop(simpleError(msg, call = expr))
6 e$fun(obj, substitute(ex), parent.frame(), e$data)
5 foreach(treenum = (forest@ntrees + 1):(forest@ntrees + ntrees),
.combine = combine.treeresults, .init = forest, .inorder = FALSE,
.verbose = FALSE) %dopar% {
if (trace >= 1L) ... at grow.R#161
4 .local(forest, ...)
3 grow(forest, x, ntrees, printerrfreq = printerrfreq, printclserr = printclserr,
trace = trace) at grow.R#1
2 grow(forest, x, ntrees, printerrfreq = printerrfreq, printclserr = printclserr,
trace = trace) at bigrfc.R#257
1 bigrfc(x[1:60, ], y[1:60], ntree = 50L, varselect = vars)
和会话信息
R version 3.0.2 (2013-09-25)
Platform: x86_64-pc-linux-gnu (64-bit)
locale:
[1] LC_CTYPE=pt_BR.UTF-8 LC_NUMERIC=C LC_TIME=pt_BR.UTF-8
[4] LC_COLLATE=pt_BR.UTF-8 LC_MONETARY=pt_BR.UTF-8 LC_MESSAGES=pt_BR.UTF-8
[7] LC_PAPER=pt_BR.UTF-8 LC_NAME=C LC_ADDRESS=C
[10] LC_TELEPHONE=C LC_MEASUREMENT=pt_BR.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] bigrf_0.1-11 bigmemory_4.4.6 BH_1.55.0-3 bigmemory.sri_0.1.3
[5] doMPI_0.2.1 Rmpi_0.6-5 iterators_1.0.7 foreach_1.4.2
loaded via a namespace (and not attached):
[1] codetools_0.2-8 compiler_3.0.2 parallel_3.0.2 tools_3.0.2
我已经知道doMPI只传递明确请求的变量，例如big.matrix，但我不知道如何obtain和pass modaC。