Arrays with arbitrary offsets

Description

The traditional R array has extents which are indexed with integers that start at 1. This is generalized to arbitrary offsets, where extent i is indexed with integers that start at offset[i], which must be no less than zero to accomodate the R convention of dropping components with negative indices. In order to use negative offsets, the flag drop.negative can be set FALSE.

Usage

Oarray(data=NA, dim=length(data), dimnames=NULL, offset=rep(1, length(dim)),
  drop.negative=TRUE)
as.Oarray(x, offset=rep(1, length(dim)), drop.negative=TRUE)
## S3 method for class 'Oarray'
as.array(x, ...)
## S3 method for class 'Oarray'
print(x, ...)

Arguments

Details

Typical uses are

  x[i, j]
  x[i, j] <- someValues

where x is an object of class ‘Oarray’ and i, j are indices specifying which elements to extract or replace.

Indexing may be via a logical matrix, which indicates which elements to select.

Indexing may be via a single numeric matrix with the one column for each dimension: the offset is sweep()-ed out. See Extract.Rd for details.

The use of drop.negative = FALSE will only work in [.Oarray where it is provided as the final argument inside the square brackets.

Value

Typically an array, either with or without the ‘Oarray’ class attribute. Extracting from an ‘Oarray’ object unclasses the result which is then a simple array, but assigning into an ‘Oarray’ object leaves the result as an ‘Oarray’ object.

The print method provides more informative extent labelling in the case where dimnames are not provided.

Side effects

The function base::as.array is redefined as generic, to provide an as.array.Oarray method.

Author(s)

Jonathan Rougier, j.c.rougier@bristol.ac.uk

See Also

array,Extract

Examples

fred <- Oarray(1:24, 2:4, list(c("sad", "happy"), NULL, NULL),
  offset=rep(7, 3))

tmp <- as.array(fred)
fred1 <- as.Oarray(tmp, offset=rep(7, 3))
stopifnot(identical(fred, fred1))

print.default(fred)  # print method provides numbers for
fred                 # non-named extents

# examples of extraction

fred[] # unclasses fred
fred["sad", 7, -9]
fred["sad", 7, -9, drop=FALSE]
fred[-8, , 7:8]

i <- 8:9; fred[, , i+1]
how.I.feel <- "happy"; fred[how.I.feel, , -(7:8)]

# examples of assignment

fred["sad", 7, -9] <- NA
fred[, , i] <- 100
fred[how.I.feel, , -(7:8)] <- Inf

# now use negative offsets and suppress usual behaviour

fred <- Oarray(24:1, 2:4, offset=c(-1, -2, 7), drop.negative=FALSE)
fred[] <- 1:24
fred[-(1:0), , 7:8]
fred[-(1:0), , 7:8] <- 100
dimnames(fred) <- list(c("sad", "happy"), NULL, NULL)
fred["sad", -2, 10] <- NA


# array and logical indexing

a <- Oarray(0, dim=rep(2,4),offset=rep(0,4))
a[diag(4)] <- 1

a[a == 0] <- NA

Internal functions for package Oarray

Description

Internal functions for package Oarray

Usage

.handleTheOffset(mc, dim, offset, dn)

Arguments

Details

Function .handleTheOffset is the workhorse function that is called by the subsetting functions [.Oarray and [<-.Oarray.

Value

Function .handleTheOffset returns an object with indices in the original values (ie with offset one) to be evaluated.

as.array.default is simply as.array from the base.

Author(s)

Jonathan Rougier, j.c.rougier@bristol.ac.uk

Examples

#### nothing to run here