OTIS a01 high-level causal analysis (MatchIt + IRM-DML).

Wraps the full causal pipeline for the canonical Restrictive Confinement Detailed Dataset: 8-state alert-combo encoding -> MatchIt 1:1 NN PSM -> IRM-DML with RF nuisances -> multi-way clustered SE.

Usage

morie_otis_analyze_a01(data = NULL, out_dir = NULL)

Arguments

data

a01 data.frame (loaded from a01_restrictive_confinement_detailed_dataset.csv). Pass NULL to indicate "use registered loader" – the R port requires you supply the data because we don't ship the loader side-effect from R.

out_dir

Optional output directory.

Value

A morie_otis_analysis_result. If the morie causal helpers aren't loaded, returns a "not yet ported" stub.

Examples

# \donttest{
otis_a01 <- morie_synth_otis("a01", n = 120L, seed = 1L)
morie_otis_analyze_a01(otis_a01)
#> OTIS a01 -- high alert complexity (ac >= 2) -> regional volatility (vm count)
#> =============================================================================
#>   Source file                   a01_restrictive_confinement_detailed_dataset.csv
#>   Person-years (after MatchIt)  42
#>   ATE                           1.8394
#>   ATE 95% CI                    [1.4082, 2.2706]
#>   ATTE                          2.0133
#>   ATTE 95% CI                   [1.2871, 2.7396]
#>   Standard error type           cluster:UniqueIndividual_ID
#> 
#> MatchIt-then-IRM-DML reproduction of the published res_pool finding on the canonical Restrictive Confinement Detailed Dataset. 
# }