Predict method for sglasso objects

Usage

# S3 method for class 'sglasso'
predict(
  object,
  newx = NULL,
  type = c("response", "coefficients", "vars", "groups", "link"),
  lambda,
  d,
  s = c("ALL", "opt"),
  opt_beta = NULL,
  which = 1:length(object$lambda),
  drop = TRUE,
  ...
)

Arguments

object: Fitted sglasso object.
newx: New design matrix.
type: Prediction type. For binomial objects, "response" returns probabilities of Y=1 and "link" returns log odds. For Gaussian objects both return the linear predictor. No classification threshold is fitted.
lambda: Lambda values.
d: Scaling parameter values.
s: Selection mode.
opt_beta: Optional coefficient vector.
which: Indices of lambda values.
drop: Logical; should dimensions be dropped?
...: Additional arguments.