OmicSelector_propensity_score_matching — OmicSelector_propensity_score

Lightweight propensity score matching (nearest neighbor) using base R.

Usage

OmicSelector_propensity_score_matching(
  dataset,
  match_by = c("age_at_diagnosis", "gender.x"),
  class_col = "Class",
  positive = "Case",
  method = "nearest",
  distance = "logit",
  ratio = 1,
  replace = FALSE,
  caliper = NULL,
  na_action = c("omit", "median")
)

Arguments

dataset: Original dataset.
match_by: Character vector of covariate names to match on.
class_col: Outcome column name (default: "Class").
positive: Label for the treated/positive class (default: "Case").
method: Matching method (currently only "nearest").
distance: Propensity model type (currently only "logit").
ratio: Number of controls per treated (default: 1).
replace: Logical; allow matching controls with replacement.
caliper: Optional maximum absolute propensity score distance.
na_action: How to handle missing values in match_by: "omit" or "median".

Value

Matched dataset (treated + matched controls).