glmBootstrapStability: Analyze stability of top n coefficients

In nbenn/singleCellFeatures: Analyze single cell features stored in an openBIS instance

Description

Resamples the data n.rep times, fits a glm modle using the function glm.fun and records the top n.hit coefficient names. Returns a table with frequencies of features that were among the top n.hit coefficients.

Usage

glmBootstrapStability(well.a, well.b, n.rep = 100, n.hit = 20,
  frac.sample = 0.7, seed = 7, glm.fun = "glmnet", norm.feat = "all",
  norm.method = "zScore", norm.sep = FALSE, select.inf = "all",
  n.cores = getNumCores(), ...)

Arguments

well.a	A WellData object for glm analysis.
well.b	A second WellData object for glm analysis.
n.rep	The number of bootstrap repeats.
n.hit	Record this many top coefficients.
frac.sample	Size of the subsample as fraction of all available data.
seed	Seed for reproducible sampling.
glm.fun	The glm function to use.
norm.feat	Which features to normalize. Either a regular expression or "all".
norm.method	How to normalize (see function normalizeMoltenData for details).
norm.sep	Separate data into wells form normalization?
select.inf	Run stability analysis only on "infected", "uninfected" or on "all"?
n.cores	The number of cores to use for bootstrapping.
...	Further arguments to be passed to the glm function. For example using glmnet, the alpha value.

Value

A vector of frequencies for features that were among the top n.hit coefficients.

Examples

wells <- findWells(plates=c("J110-2C", "J104-2D", "J107-2L"),
                   well.names=c("H2", "H6"))
data  <- unlist(getSingleCellData(wells), recursive=FALSE)
a <- cleanData(data[["J110-2C.H2"]], "lower")
b <- cleanData(data[["J110-2C.H6"]], "lower")
test <- glmBootstrapStability(a, b, n.rep=20)

nbenn/singleCellFeatures documentation built on May 23, 2019, 12:24 p.m.