biom.options {BioMark}

Set or return options for stability-based biomarker selection

Description

A function to set options for stability-based biomarker selection in the BioMark package, or to return the current options.

Usage

biom.options(..., reset = FALSE)

Arguments

...

a single list, a single character vector, or any number of named arguments (name = value).

reset

logical: if TRUE all options are set to their factory defaults.

Details

If called with no arguments, or with an empty list as the single argument, biom.options returns the current options.

If called with a character vector as the single argument, a list with the arguments named in the vector are returned.

If called with a non-empty list as the single arguments, the list elements should be named, and are treated as named arguments to the function.

Otherwise, biom.options should be called with one or more named arguments name = value. For each argument, the option named name will be given the value value.

The options are saved in an envirtonment variable .biom.Options, and remain in effect until the end of the session. If the environment is saved upon exit, they will be remembered in the next session.

The recognised options are:

max.seg

Maximal number of jackknife iterations. Default: 100.

oob.size, oob.fraction

Size of the out-of-bag fraction, either given as an absolute number (oob.size) or as a fraction. Default is to leave out ten percent. If oob.size is given explicitly, it takes precedence over oob.fraction. Default: oob.fraction = .1.

variable.fraction

Use 1 to always include all variables - use a smaller fraction to have a different random subset of all variables in each iteration (stability-based identification). Default: .7.

ntop

The number of "top" coefficients taken into account in stability-based biomarker identification. If a variable appears consistently among the ntop biggest coefficients, it is said to be stable. If ntop is a number between 0 and 1, it is taken to indicate the fraction of variables to be included in the model. Default: 10.

min.present

The minimal fraction of times a variable should be in the top list to be considered as a potential biomarker (stability-based identification). Setting this argument to 0 will lead to a list containing all coefficients that were present in the top list at least once - a value of 1 only returns those variables that are selected in every iteration. Default: .1.

nset

The number of permutations to establish null distributions for PCR, PLS and VIP statistics in the Higher-Criticism approach. Default: 10,000.

fmethods

All biomarker selection methods available within BioMark. Currently equal to c("studentt", "shrinkt", "pcr", "pls", "vip", "lasso".

univ.methods

The names of the univariate biomarker selection methods currently known to BioMark. Currently equal to c("studentt", "shrinkt")

HCalpha

The default of the alpha parameter in the HC method. Value: 0.1.

lasso

a list of arguments passed to the underlying glmnet function, such as family, nlambda, alpha, lambda, or lambda.min.ratio. For binary classification, the "binomial" family is the default, but the most similar setting compared to the other methods in the package is family = "gaussian". For choices other than the default, a warning is printed to the screen.

Value

A list with the (possibly changed) options. If any named argument (or list element) was provided, the list is returned invisibly.

Side Effects

If any named argument (or list element) was provided, biom.options updates the elements of the option list .biom.Options$options.

Note

This function is based on the pls.options function in package pls.

Author(s)

Ron Wehrens

See Also

glmnet

Examples

## Return current options:
biom.options()
biom.options("max.seg")

## Set options:
biom.options(max.seg = 100, oob.fraction = .2)
biom.options(lasso = list(alpha = .75, nlambda = 50))
biom.options()
## the next line removes some options - for these, glmnet defaults will be used
biom.options(lasso = list(alpha = .9, family = "binomial"))

## Restore factory settings:
biom.options(reset = TRUE)