Overlap-stressed synthetic DGP for CausalStress (v1.3.0) — dgp_synth_overlap_stressed

Implements the synth_overlap_stressed design from the DGP registry. Shares covariates, baseline outcome, treatment effect, and Gaussian noise with synth_baseline, but uses a sharper propensity to stress overlap: p = plogis(3 * X1 + 3 * X2).

Usage

dgp_synth_overlap_stressed_v130(
  n,
  seed = NULL,
  include_truth = TRUE,
  oracle_only = FALSE
)

Arguments

n: Integer, number of observations.
seed: Optional seed for reproducibility (passed to cs_set_rng()).
include_truth: Logical; if TRUE, include oracle truth tables where supported.
oracle_only: Logical; if TRUE, return only columns needed for oracle truth generation where supported.

Value

A list with:

df: tibble with columns y, w, y0, y1, p, structural_te, X1, X2, X3, X4, X5
true_att: numeric scalar (ATT on treated units)
true_qst: tibble with columns tau, value
meta: list with fields dgp_id, type, structural_te

Details

Goal: Stress-test weighting stability under severe overlap violations.

Structure: Same outcomes/noise as synth_baseline.

Propensity: \(p(X) = \mathrm{plogis}(3.0 X_1 + 3.0 X_2)\), intentionally pushing probability mass toward \(0\) or \(1\).