rgcca_bootstrap: Bootstrap confidence intervals and p-values

In Tenenhaus/RGCCA: Regularized (or Sparse) Generalized Canonical Correlation Analysis (R/SGCCA) for multi-block data analysis

Bootstrap confidence intervals and p-values

Description

Bootstrap confidence intervals and p-values for evaluating the significance/ stability of the block-weight vectors produce by S/RGCCA.

Usage

rgcca_bootstrap(
  rgcca_res,
  n_boot = 100,
  n_cores = 1,
  balanced = TRUE,
  keep_all_variables = FALSE,
  verbose = TRUE
)

Arguments

rgcca_res	A fitted RGCCA object (see rgcca)
n_boot	Number of bootstrap samples. Default: 100.
n_cores	Number of cores for parallelization.
balanced	A boolean indicating if a balanced bootstrap procedure is performed or not (default is TRUE).
keep_all_variables	A boolean indicating if all variables have to be kept even when some of them have null variance for at least one bootstrap sample (default is FALSE).
verbose	Logical value indicating if the progress of the bootstrap procedure is reported.

Value

A list containing two objects: 'bootstrap' and 'rgcca'. 'bootstrap' is a list containing for each block, a matrix with the variables of the block in row and the block weight vector calculated across bootstrap sample in column. 'rgcca' is the fitted rgcca object obtained from the original data. (see rgcca)

See Also

plot.bootstrap, print.bootstrap

Examples

# Bootstrap confidence intervals and p-values for RGCCA
data("Russett")
blocks <- list(
  agriculture = Russett[, seq(3)],
  industry = Russett[, 4:5],
  politic = Russett[, 6:11]
)

fit.rgcca <- rgcca(blocks, ncomp = c(2, 1, 2))
boot.out <- rgcca_bootstrap(fit.rgcca, n_boot = 20, n_cores = 2)

plot(boot.out, type = "weight", block = 3, comp = 1)

print(boot.out, comp = 2, block = 1)

fit.rgcca <- rgcca(blocks, method = "mcoa")
boot.out <- rgcca_bootstrap(fit.rgcca, n_boot = 50, n_cores = 2)

plot(boot.out, type = "weight", block = 1)
## Not run: 
# Stability of the selected variables for SGCCA
# Download the dataset's package at http://biodev.cea.fr/sgcca/ and install
# it from the package archive file.
# You can do it with the following R commands:
if (!("gliomaData" %in% rownames(installed.packages()))) {
  destfile <- tempfile()
  download.file("http://biodev.cea.fr/sgcca/gliomaData_0.4.tar.gz", destfile)
  install.packages(destfile, repos = NULL, type = "source")
}

data("ge_cgh_locIGR", package = "gliomaData")
A <- ge_cgh_locIGR$multiblocks
A[[3]] <- A[[3]][, -3]
Loc <- factor(ge_cgh_locIGR$y)
levels(Loc) <- colnames(ge_cgh_locIGR$multiblocks$y)
C <- matrix(c(0, 0, 1, 0, 0, 1, 1, 1, 0), 3, 3)

# rgcca algorithm using the dual formulation for X1 and X2
# and the dual formulation for X3

fit.rgcca <- rgcca(A,
  connection = C, tau = c(1, 1, 0),
  ncomp = c(2, 2, 1), scheme = "factorial",
  verbose = TRUE
)
boot.out <- rgcca_bootstrap(fit.rgcca, n_boot = 50, n_cores = 2)
plot(boot.out, block = 1, type = "weight", ncomp = 1, n_marks = 30)
plot(boot.out, block = 1, type = "weight", ncomp = 2, n_marks = 30)

# stability analysis prior bootstrap for sgcca

## End(Not run)

Tenenhaus/RGCCA documentation built on March 16, 2023, 2:04 p.m.