bs_var_tf: Estimate variance of a trendfiltering fit by bootstrapping

In jean997/jadeTF:

Description

Estimate variance of a trendfiltering fit by bootstrapping

Usage

bs_var_tf(Y, lambda, sds = NULL, READS = NULL, sd.type = c("Equal",
  "Binomial", "Poisson"), calc.sds = TRUE, pos = NULL, ord = 2,
  n.rep = 100)

Arguments

Y	A matrix of complete (not average) data for one group.
sds	A matrix of estimated standard deviations of Y. If missing observations are assumed to have equal variance.
READS	A matrix of reads corresponding to Y (Optional). If READS is provided, the elements in Y are assumed to be binomial observations and sds will be calculated unless calc.sds=FALSE
pos	Vector of positions corresponding to the rows of Y
ord	Order of trendfiltering. Can be 0, 1 or 2.
n.rep	Number of bootstrap samples to take.

Value

A list with three elements

all.fits

All of the fits from bootstrapped samples. p by n.rep matrx

avg.fit

Equivalent to rowMeans(all.fits)

var

Estimated variance.

jean997/jadeTF documentation built on May 18, 2019, 11:44 p.m.