Suggest an ordered analysis plan based on a dataset profile.

Uses the inferred measurement levels, binary indicators, and detected treatment/outcome/weight columns to recommend epidemiological analyses (descriptive profile, propensity scores, IPW-ATE, AIPW, ATT/ATC, double-ML, GATE, survey-weighted estimates).

Usage

morie_dataset_suggest_plan(profile)

Arguments

profile

A morie_dataset_profile.

Value

A list of suggestion lists; each has analysis, rationale, and required_vars.

Examples

df <- data.frame(
  treat = c(0, 1, 0, 1, 0, 1),
  outcome_y = c(1.1, 2.2, 1.4, 3.0, 0.9, 2.8),
  age = c(20, 25, 30, 35, 40, 45)
)
p <- morie_dataset_profile(df, hint_treatment = "treat", hint_outcome = "outcome_y")
plan <- morie_dataset_suggest_plan(p)
vapply(plan, function(s) s$analysis, character(1))
#> [1] "descriptive_profile" "propensity_scores"   "ipw_ate"            
#> [4] "aipw"                "att_atc"             "double_ml_plr"