RxLocalParallel: Generate Local Parallel Compute Context

Artigo
07/15/2019

Description

Creates a local compute context object that uses the doParallel back-end for HPC computations performed using rxExec. This compute context can be used only to distribute computations via the rxExec function; it is ignored by Revolution HPA functions. This is the main generator for S4 class RxLocalParallel.

Usage

  RxLocalParallel( object, dataPath = NULL, outDataPath = NULL  )

Arguments

`object`

a compute context object. If object has slots for dataPath and/or outDataPath, they will be copied to the equivalent slots for the new RxLocalParallel object. Explicit specifications of the dataPath and/or outDataPath arguments will override this.

`dataPath`

NULL or character vector defining the search path(s) for the input data source(s). If not NULL, it overrides any specification for dataPath in rxOptions

`outDataPath`

NULL or character vector defining the search path(s) for new output data file(s). If not NULL, this overrides any specification for dataPath in rxOptions

Details

A job is associated with the compute context in effect at the time the job was submitted. If the compute context subsequently changes, the compute context of the job is not affected.

Note that dataPath and outDataPath are only utiltized by data sources used in RevoScaleR analyses. They do not alter the working directory for other R functions that read from or write to files.

Value

object of class RxLocalParallel.

Author(s)

Microsoft Corporation Microsoft Technical Support

Examples


 ## Not run:


# Create a RxLocalParallel object to use with rxExec  
parallelContext <- RxLocalParallel()

# Tell RevoScaleR to use parallelContext compute context
rxSetComputeContext( parallelContext )

# Subsequent calls to rxExec will use the doParallel package 
# behind the scenes.

## End(Not run)

Partilhar via