howbig: howbig

Description

Determines the memory usage for a dense, in-core, numeric matrix of specified rows/columns.

Usage

howbig(nrow, ncol, representation = "dense", unit = .UNIT, unit.prefix = .PREFIX, unit.names = .NAMES, ..., sparsity = 0.05, type = "double", intsize = 4)
howbig.par(nrow, ncol, cores = 1, par = "balanced", unit = .UNIT, unit.prefix = .PREFIX, unit.names = .NAMES, ..., type = "double", intsize = 4, ICTXT = 0, bldim = c(4, 4))

Arguments

nrow,ncol

Number of (global) rows/columns of the matrix.

representation

The kind of storage the object would be in, i.e. "dense" or "sparse".

unit

string; the unit of storage, such as "MiB" or "MB", depending on prefix. Case is ignored.

unit.prefix

string; the unit prefix, namely IEC or SI. Case is ignored.

unit.names

string; control for whether the unit names should be printed out or their abbreviation should be used. Options are "long" and "short", respectively. Case is ignored.

...

Additional arguments.

sparsity

The proportion of sparsity of the matrix if representation="sparse"

type

"double" or "int"; the storage type of the data matrix. If you don't know the type, it is probably stored as a double, so the default value will suffice.

intsize

The size (in bytes) of an integer. Default is 4, but this is platform dependent.

cores

The number of cores/processors

par

Type of data distribution. Choices are "dmat" or "balanced". The former is for pbdDMAT objects, the latter is the simple, locally load-balanced block partitioning.

ICTXT

BLACS context number; only used with howbig.par() with par="dmat".

bldim

Blocking factor for block-cyclically decomposed (dmat) data.

Value

howbig() returns a memuse class object.howbig.par() returns a list of 2 elements, each of class memuse. One is the total memory usage, the other is the local memory usage.

Details

These functions provide the memory usage of an unallocated, dense, in-core, numeric matrix. As the names suggest, howbig() simply returns the size (as a memuse object), while howbig.par() is the parallel, distributed analogue. The latter returns the memory usage of a distributed, object

Examples

Run this code

## Not run: 
# # size of a 1000x1000 matrix
# howbig(1000, 1000)
# ## End(Not run)