// dgp_irm_heterogeneous.mbt
//
// Pure-MoonBit port of upstream
// `doubleml.irm.datasets.dgp_heterogeneous_data.make_heterogeneous_data`.
//
// IRM DGP with heterogeneous treatment effects (the parameter
// theta_i varies across observations by a function of X):
//
// p_score_i = 0.5 * X_{i,1} + 0.5 * X_{i,2}
// d_i ~ Bernoulli(sigmoid(p_score_i))
// theta_i = 0.5 + 0.25 * X_{i,1} // heterogeneous causal effect
// y_i = theta_i * d_i + 0.5 * sigmoid(X_{i,1}) + 0.5 * sigmoid(X_{i,2})
// + v_i, v ~ N(0, 1)
// X ~ N(0, Sigma), Sigma_{kj} = 0.7^|j-k|
//
// The ATE (average of theta_i over the population) is still 0.5
// when X_{i,1} has mean 0 (which it does under N(0, Sigma)), but
// the conditional effect varies. Tests check that the recovered
// estimator theta_hat lands near the population ATE.
///|
struct IrmHeterogeneousData {
/// Average causal parameter (ATE of theta_i over the sample).
theta : Double
x : Matrix
y : Array[Double]
d : Array[Double]
}
///|
pub fn make_irm_heterogeneous_data(
n_obs : Int,
dim_x : Int,
seed : Int,
) -> IrmHeterogeneousData {
let rng = chacha8_rng(seed)
// 1. Pre-draw X-draw normals.
let n_normals_x = n_obs * dim_x
let z_flat : Array[Double] = Array::make(n_normals_x, 0.0)
let half_z = (n_normals_x + 1) / 2
for i = 0; i < half_z; i = i + 1 {
let (z1, z2) = box_muller_pair(rng)
let idx_a = 2 * i
let idx_b = 2 * i + 1
if idx_a < n_normals_x {
z_flat[idx_a] = z1
}
if idx_b < n_normals_x {
z_flat[idx_b] = z2
}
}
// 2. Build X.
let x_flat : Array[Double] = Array::make(n_normals_x, 0.0)
for i = 0; i < n_obs; i = i + 1 {
for k = 0; k < dim_x; k = k + 1 {
let mut s = 0.0
let mut acc = 1.0
let mut j = k
while j >= 0 {
s = s + acc * z_flat[i * dim_x + j]
acc = acc * 0.7
if j == 0 {
break
}
j = j - 1
}
x_flat[i * dim_x + k] = s
}
}
// 3. Pre-draw v ~ N(0, 1).
let v_flat : Array[Double] = Array::make(n_obs, 0.0)
let half_n = (n_obs + 1) / 2
for i = 0; i < half_n; i = i + 1 {
let (z1, z2) = box_muller_pair(rng)
let idx_a = 2 * i
let idx_b = 2 * i + 1
if idx_a < n_obs {
v_flat[idx_a] = z1
}
if idx_b < n_obs {
v_flat[idx_b] = z2
}
}
// 4. Build d and y; track per-unit theta_i for the population ATE.
let d : Array[Double] = Array::make(n_obs, 0.0)
let y : Array[Double] = Array::make(n_obs, 0.0)
let mut theta_sum = 0.0
for i = 0; i < n_obs; i = i + 1 {
let x1 = x_flat[i * dim_x + 1]
let x2 = x_flat[i * dim_x + 2]
let p_score = 0.5 * x1 + 0.5 * x2
let p = 1.0 / (1.0 + @math.exp(-p_score))
d[i] = if rng.double() < p { 1.0 } else { 0.0 }
let theta_i = 0.5 + 0.25 * x1
theta_sum = theta_sum + theta_i
let sig1 = 1.0 / (1.0 + @math.exp(-x1))
let sig2 = 1.0 / (1.0 + @math.exp(-x2))
y[i] = theta_i * d[i] + 0.5 * sig1 + 0.5 * sig2 + v_flat[i]
}
let theta = theta_sum / n_obs.to_double()
let x_mat = Matrix::from_array(x_flat, n_obs, dim_x)
{ theta, x: x_mat, y, d }
}
///|
pub fn IrmHeterogeneousData::theta_get(self : IrmHeterogeneousData) -> Double {
self.theta
}
///|
pub fn IrmHeterogeneousData::x_get(self : IrmHeterogeneousData) -> Matrix {
self.x
}
///|
pub fn IrmHeterogeneousData::y_get(
self : IrmHeterogeneousData,
) -> Array[Double] {
self.y
}
///|
pub fn IrmHeterogeneousData::d_get(
self : IrmHeterogeneousData,
) -> Array[Double] {
self.d
}