sim.probe.specific: Simulate a Probe-Specific Variable

In Sage-Bionetworks/snm: Supervised Normalization of Microarrays

Description

This function is designed to simulate a probe-specific variable. It accepts one variable at a time. The variable need be either a factor or numeric vectors, any character vector will be modeled as a factor.

Usage

sim.probe.specific(data = NULL, variable = NULL, rows = NULL, sample.from = NULL)

Arguments

data	The data matrix of simulated data the user wishes to adjust.
variable	The variable the user wishes to introduce.
rows	Either a number between 0 and 1 or a vector of row IDs. A number between 0 and 1 specifies the proportion of data to be influenced by the variable. A vector specifies the rows to be influenced.
sample.from	A list specifying a function that coefficients can be sampled from (e.g. rnorm, rchisq, etc.) and the relevant parameters (e.g. mean, sd, df, etc). The first item in the list should be named func and the second named params.

Value

Matrix of dimension equal to that of data containing the probe-specific effects.

Author(s)

Brig Mecham <[e-mail]>

Examples

data <- matrix(rchisq(100,1,2),nr=1000,nc=12)
bio.var <- data.frame(groups=rep(c("A","B"),each=6))
group.effect <- sim.probe.specific(data, bio.var$groups, 0.3, list(func=rnorm,params=c(mean=1,sd=0.3)))

Sage-Bionetworks/snm documentation built on May 9, 2019, 12:14 p.m.