Set a matrix indexed by a pair of axes.

Usage

set_matrix(
  daf,
  rows_axis,
  columns_axis,
  name,
  mat,
  overwrite = FALSE,
  relayout = TRUE
)

Arguments

daf: A DafReader.
rows_axis: Row-axis name.
columns_axis: Column-axis name.
name: Matrix name.
mat: Dense matrix, or sparse dgCMatrix / lgCMatrix, of shape axis_length(rows_axis) x axis_length(columns_axis).
overwrite: See set_scalar.
relayout: If TRUE (the default, matching Julia set_matrix!(...; relayout = true)), also store the flipped layout so get_matrix(columns_axis, rows_axis, name) is directly available and skips the transpose-on-the-fly path. Set FALSE to store only the given orientation (saving storage at the cost of on-the-fly transposes).

Value

Invisibly the input daf.

Examples

# Mirrors writers.jl jldoctest at line 686.
m <- example_metacells_daf()
has_matrix(m, "gene", "metacell", "confidence", relayout = FALSE)    # FALSE
#> [1] FALSE
set_matrix(m, "metacell", "gene", "confidence",
           matrix(stats::runif(7L * 683L), 7L, 683L), relayout = FALSE)
has_matrix(m, "gene", "metacell", "confidence", relayout = FALSE)    # FALSE
#> [1] FALSE
has_matrix(m, "metacell", "gene", "confidence", relayout = FALSE)    # TRUE
#> [1] TRUE
set_matrix(m, "metacell", "gene", "confidence",
           matrix(stats::runif(7L * 683L), 7L, 683L),
           overwrite = TRUE, relayout = TRUE)
has_matrix(m, "gene", "metacell", "confidence", relayout = FALSE)    # TRUE
#> [1] TRUE
has_matrix(m, "metacell", "gene", "confidence", relayout = FALSE)    # TRUE
#> [1] TRUE