R: Divides various data types into *blocks* for use in parallel processing

divide {StabPerf}

R Documentation

Divides various data types into blocks for use in parallel processing

Description

Divides matrices, data.frames, lists, and vectors into lists of subsets of the orginal data.

Usage

divide(data, divisions=NULL, by.row=FALSE, by.col=FALSE)

Arguments

`data`	matrix, vector, list, or data.frame. Must be `numeric`.
`divisions`	integer. How many times to break up the `data`
`by.row`	logical. Whether matrices and data.frames should be split along rows
`by.col`	logical. Whether matrices and data.frames should be split along columns

Details

If divisions is not specified, it is based on the number of currently active nodes in the running cluster, if one is running. Divisions cannot be less than 2. If the size of the input data cannot be evenly divided by divisions the data will be distributed as evenly as possible. The order of the original data is not changed. You cannot use by.row and by.col together in the same call.

Value

list list of things of the same type as data

Examples

# Returns a list of two vectors, equally long
divide(1:10, 2)
# Returns a list of matrices, one matrix will have 2 columns, the other 3
divide(matrix(1:10,nrow=2), 5, by.col=TRUE)
# Note: this returns what it was passed, without changing anything
divide(list(1:32), 3)
# This, however, is a list of sub-lists and the sub-lists will be rearranged
# such that two lists are returned. Each of these is a list of lists of the
# original data.
divide(list(list(0:10),list(20:30), list(40:50)), 2)

[Package StabPerf version 0.5 Index]