Specify additional options for the augmented LARS and the DAL algorithm to solve the EN problem.
en_options_aug_lars(use_gram = c("auto", "yes", "no"), eps = 1e-12)en_options_dal(
maxit = 100,
eps = 1e-08,
eta_mult = 2,
eta_start_numerator = 0.01,
eta_start,
preconditioner = c("approx", "none", "diagonal"),
verbosity = 0
)
should the Gram matrix be pre-computed.
numeric tolerance for convergence.
maximum number of iterations allowed.
multiplier to increase eta at each iteration.
if eta_start
is missing, it is defined
by eta_start = eta_start_numerator / lambda
.
the start value for eta.
preconditioner for the numerical solver. If none, a standard solver will be used, otherwise the faster preconditioned conjugate gradient is used.
verbosity of the algorithm.
a checked options list.
Other specifying additional options:
initest_options()
,
mstep_options()
,
pense_options()