Computing mean/median CI ratio and significant Discoveries

Perfroms full model boostrap, stepwise AIC, stepwise BIc, lasso and MCP and gives average CI for variables adjusting for post-selectino inference using hybrid, bootstrap and selective inference

Usage

ciratio(
  formula,
  data,
  family = c("gaussian", "binomial", "poisson"),
  nonselection = c("ignored", "confident_nulls", "uncertain_nulls"),
  conf.level = 0.95,
  B = 250,
  direction = "forward",
  select_factors_together = F,
  debias = F,
  inference_target = "selections",
  ...
)

Arguments

formula: a formula
data: data set
family: outcome distributional family
nonselection: A character string specifying how to handle variables not selected by model selection procedure. One of "ignored", "confident_nulls" or "uncertain_nulls" supported
conf.level: .95 by default
B: Number of bootstraps
direction: the mode of step wise search, can be one of "both", "backward", or "forward", with a default of "forward"
select_factors_together: should categorical variables be jointly selected?
debias: should estimates be debiased in bootstrap
inference_target: is inference requested on all or selected only
...: additional arguments

Value

A list with following data frames

avg_ci_ln: Mean of CI length across all variables in model excluding intercept
med_ci_ln: Median of CI length across all variables in model excluding intercept
no_sign_disc: # of significant discoveries in model excluding intercept