big_cprodVec: Cross-product with a vector

In bigstatsr: Statistical Tools for Filebacked Big Matrices

Cross-product with a vector

Description

Cross-product between a Filebacked Big Matrix and a vector.

Usage

big_cprodVec(
  X,
  y.row,
  ind.row = rows_along(X),
  ind.col = cols_along(X),
  center = NULL,
  scale = NULL,
  ncores = 1
)

Arguments

X	An object of class FBM.
y.row	A vector of same size as ind.row.
ind.row	An optional vector of the row indices that are used. If not specified, all rows are used. Don't use negative indices.
ind.col	An optional vector of the column indices that are used. If not specified, all columns are used. Don't use negative indices.
center	Vector of same length of ind.col to subtract from columns of X.
scale	Vector of same length of ind.col to divide from columns of X.
ncores	Number of cores used. Default doesn't use parallelism. You may use nb_cores.

Value

X^T \cdot y.

Examples

X <- big_attachExtdata()
n <- nrow(X)
m <- ncol(X)
y <- rnorm(n)

test <- big_cprodVec(X, y)             # vector
true <- crossprod(X[], y)  # one-column matrix
all.equal(test, as.numeric(true))

# subsetting
ind.row <- sample(n, n/2)
ind.col <- sample(m, m/2)

tryCatch(test2 <- big_cprodVec(X, y, ind.row, ind.col),
         error = function(e) print(e))
# returns an error. You need to use the subset of y:
test2 <- big_cprodVec(X, y[ind.row], ind.row, ind.col)
true2 <- crossprod(X[ind.row, ind.col], y[ind.row])
all.equal(test2, as.numeric(true2))

bigstatsr documentation built on Sept. 11, 2024, 7:08 p.m.