SparseArray-abind: Combine multidimensional SparseArray objects

In Bioconductor/SparseArray: High-performance sparse data representation and manipulation in R

Combine multidimensional SparseArray objects

Description

Like ordinary matrices and arrays in base R, SparseMatrix derivatives can be combined by rows or columns, with rbind() or cbind(), and multidimensional SparseArray derivatives can be bound along any dimension with abind().

Note that arbind() can also be used to combine the objects along their first dimension, and acbind() can be used to combine them along their second dimension.

See Also

• cbind in base R.

• abind in the S4Arrays package.

• SparseArray objects.

• Ordinary array objects in base R.

Examples

## ---------------------------------------------------------------------
## COMBINING SparseMatrix OBJECTS
## ---------------------------------------------------------------------

m1a <- matrix(1:15, nrow=3, ncol=5,
              dimnames=list(NULL, paste0("M1y", 1:5)))
m1b <- matrix(101:135, nrow=7, ncol=5,
              dimnames=list(paste0("M2x", 1:7), paste0("M2y", 1:5)))
sm1a <- SparseArray(m1a)
sm1b <- SparseArray(m1b)

rbind(sm1a, sm1b)

## ---------------------------------------------------------------------
## COMBINING SparseArray OBJECTS WITH 3 DIMENSIONS
## ---------------------------------------------------------------------

a2a <- array(1:105, dim=c(5, 7, 3),
             dimnames=list(NULL, paste0("A1y", 1:7), NULL))
a2b <- array(1001:1105, dim=c(5, 7, 3),
             dimnames=list(paste0("A2x", 1:5), paste0("A2y", 1:7), NULL))
sa2a <- SparseArray(a2a)
sa2b <- SparseArray(a2b)

abind(sa2a, sa2b)               # same as 'abind(sa2a, sa2b, along=3)'
abind(sa2a, sa2b, rev.along=0)  # same as 'abind(sa2a, sa2b, along=4)'

a3a <- array(1:60, dim=c(3, 5, 4),
             dimnames=list(NULL, paste0("A1y", 1:5), NULL))
a3b <- array(101:240, dim=c(7, 5, 4),
             dimnames=list(paste0("A2x", 1:7), paste0("A2y", 1:5), NULL))
sa3a <- SparseArray(a3a)
sa3b <- SparseArray(a3b)

arbind(sa3a, sa3b)  # same as 'abind(sa3a, sa3b, along=1)'

## ---------------------------------------------------------------------
## Sanity checks
## ---------------------------------------------------------------------

sm1 <- rbind(sm1a, sm1b)
m1 <- rbind(m1a, m1b)
stopifnot(identical(as.array(sm1), m1), identical(sm1, SparseArray(m1)))

sa2 <- abind(sa2a, sa2b)
stopifnot(identical(sa2, abind(sa2a, sa2b, along=3)))
a2 <- abind(a2a, a2b, along=3)
stopifnot(identical(as.array(sa2), a2), identical(sa2, SparseArray(a2)))

sa2 <- abind(sa2a, sa2b, rev.along=0)
stopifnot(identical(sa2, abind(sa2a, sa2b, along=4)))
a2 <- abind(a2a, a2b, along=4)
stopifnot(identical(as.array(sa2), a2), identical(sa2, SparseArray(a2)))

sa3 <- arbind(sa3a, sa3b)
a3 <- arbind(a3a, a3b)
stopifnot(identical(as.array(sa3), a3), identical(sa3, SparseArray(a3)))

Bioconductor/SparseArray documentation built on Nov. 14, 2024, 6:22 p.m.