reductions: Arithmetic Reductions: Sums, Means, and Prods

In pbdDMAT: 'pbdR' Distributed Matrix Methods

Description

Arithmetic reductions for distributed matrices.

Usage

rowMin(x, ...)

rowMax(x, ...)

colMin(x, ...)

colMax(x, ...)

## S4 method for signature 'ddmatrix'
rowSums(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
colSums(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
rowMeans(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
colMeans(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
rowMin(x, na.rm = FALSE)

## S4 method for signature 'matrix'
rowMin(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
colMin(x, na.rm = FALSE)

## S4 method for signature 'matrix'
colMin(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
rowMax(x, na.rm = FALSE)

## S4 method for signature 'matrix'
rowMax(x, na.rm = FALSE)

## S4 method for signature 'ddmatrix'
colMax(x, na.rm = FALSE)

## S4 method for signature 'matrix'
colMin(x, na.rm = FALSE)

Arguments

x	numeric distributed matrix
...	additional arguments
na.rm	logical. Should missing (including NaN) be removed?

Details

Performs the reduction operation on a distributed matrix.

There are several legitimately new operations, including rowMin(), rowMax(), colMin(), and colMax(). These implementations are not really necessary in R because one can easily (and reasonably efficiently) do something like

apply(X=x, MARGIN=1L, FUN=min, na.rm=TRUE)

But apply() on a ddmatrix is very costly, and should be used sparingly.

sd() will compute the standard deviations of the columns, equivalent to calling apply(x, MARGIN=2, FUN=sd) (which will work for distributed matrices, by the way). However, this should be much faster and use less memory than apply(). If reduce=FALSE then the return is a distributed matrix consisting of one (global) row; otherwise, an R vector is returned, with ownership of this vector determined by proc.dest.

Value

Returns a global numeric vector.

pbdDMAT documentation built on May 1, 2019, 6:34 p.m.