mirtCluster {mirt} | R Documentation |
Define a parallel cluster object to be used in internal functions
Description
This function defines a object that is placed in a relevant internal environment defined in mirt.
Internal functions such as calcLogLik
, fscores
, etc, will utilize this object
automatically to capitalize on parallel
processing architecture. The object defined is a call from parallel::makeCluster()
.
Note that if you are defining other parallel objects (for simulation designs, for example)
it is not recommended to define a mirtCluster.
Usage
mirtCluster(spec, omp_threads, remove = FALSE, ...)
Arguments
spec |
input that is passed to |
omp_threads |
number of OpenMP threads to use (currently applies to E-step computations only). Not used when argument input is missing |
remove |
logical; remove previously defined |
... |
additional arguments to pass to |
Author(s)
Phil Chalmers rphilip.chalmers@gmail.com
References
Chalmers, R., P. (2012). mirt: A Multidimensional Item Response Theory Package for the R Environment. Journal of Statistical Software, 48(6), 1-29. doi:10.18637/jss.v048.i06
Examples
## Not run:
if(interactive()){
# use all available cores
mirtCluster()
mirtCluster(remove = TRUE)
# make 4 cores available for parallel computing
mirtCluster(4)
mirtCluster(remove = TRUE)
# create 3 core architecture in R, and 4 thread architecture with OpenMP
mirtCluster(spec = 3, omp_threads = 4)
# leave previous multicore objects, but change omp_threads
mirtCluster(spec = NULL, omp_threads = 2)
}
## End(Not run)