ers: Environmental Risk Score

In umich-biostatistics/ers: Environmental Risk Score

Description

Estimate environmental risk scores from a number of potentially correlated risk factors using adaptive elastic net (AENET) regression. ENET regression does variable selection and can select multiple non-zero collinear variables without overfitting. AENET is an adaptive version of elastic net (ENET) that satisfies the asymptotic normality assumption that allows us to conduct statistical inference any hypothesis testing by providing large sample standard errors and p-values.

Usage

ers(
  x,
  y,
  covar = NULL,
  control = list(lambda2.start = NULL, lambda2.adapt = NULL),
  method = "ls",
  scaled = FALSE,
  nfold = 5,
  seed = NULL,
  ...
)

Examples

set.seed(7794)
fit = ers(x = metal, y = as.numeric(Y), covar = covs,
          control = list(lambda2.start = seq(0.001, 0.5, by = 0.01),
                         lambda2.adapt = seq(0.001, 0.5, by = 0.01)))

umich-biostatistics/ers documentation built on Oct. 29, 2020, 8:15 a.m.