.loss {starnet}

Loss

Description

Calculate loss from predicted and observed values

Usage

.loss(y, x, family, type.measure, foldid = NULL, grouped = TRUE)

Arguments

y

observed values: numeric vector of length n

x

predicted values: numeric vector of length n

family

character "gaussian", "binomial", "poisson", "mgaussian", or "multinomial" (to implement: "cox")

type.measure

character "deviance", "mse", "mae", "class", or "auc"

foldid

fold identifiers: integer vector of length n, or NULL

grouped

logical (for "cox" only)

Examples

NA