varImp: Compute variable importance of various survival models object

In CYGUBICKO/glmnetpostsurv: Fit Survival Models with 'glmnet'

Compute variable importance of various survival models object

Description

Compute variable importance of various survival models object

Usage

varImp(
  object,
  type = c("param", "variable"),
  scale = TRUE,
  newdata,
  nrep = 20,
  ...
)

Arguments

object	fitted glmnetsurv, pcoxtime, coxph, etc, object.
type	if type = "param" absolute value of estimated coefficients are used. If type = "variable" variable level importance is computed using permutation.
scale	if TRUE the scores are divided by the absolute sum of the coefficients.
newdata	optional data frame containing the variables appearing on the right hand side of glmnetsurv formula. Required if type = "variable"
nrep	number of replicates for permulations. Only if type = "variable".
...	not implemented.

Details

Absolute value of the coefficients (parameters) corresponding the tuned model are used type = param. Otherwise, variable level importance is computed using permutation. See permuteImp.

See Also

plotImp

Examples

data(veteran, package="survival")
# glmnet
gfit1 <- glmnetsurv(Surv(time, status) ~ factor(trt) + karno + diagtime + age + prior
	, data = veteran
	, lambda = 0.02
	, alpha = 0.8
)
imp1 <- varImp(gfit1, type = "param")
print(imp1)
imp2 <- varImp(gfit1, type = "variable", newdata = veteran)
print(imp2)

CYGUBICKO/glmnetpostsurv documentation built on Sept. 1, 2022, 7:26 p.m.