Struct ErrorChainComplex 

pub struct ErrorChainComplex { /* private fields */ }

Error chain complex for a Clifford circuit with noise.

Represents the chain complex C₂ →∂₂→ C₁ →∂₁→ C₀ where:

• C₀ = F₂^m (measurement/detector space)
• C₁ = F₂^p (error location space)
• C₂ = F₂^s (stabilizer space)
• ∂₁ = E (the m×p error propagation matrix)

Implementations

impl ErrorChainComplex

pub fn build(circuit: &Circuit, noise: &NoiseModel, _seed: u64) -> Result<Self>

Build the error chain complex from a Clifford circuit and noise model.

Uses backward Pauli propagation (same as the compiled noisy sampler) to determine which measurements are sensitive to each error location.

pub fn boundary_dim(&self) -> usize

dim(im(∂₂) ∩ ker(∂₁)): stabilizer generators undetectable by the measurements.

pub fn homology_dim(&self) -> usize

dim(H₁): independent logical error classes.

pub fn noisy_marginals(&self, noiseless_marginals: &[f64]) -> Vec<f64>

Compute exact noisy marginals analytically. No sampling, no rank limit.

For each measurement j, the noisy probability is: p_j^noisy = p_j + (1 - 2·p_j) · (1 - f_j) / 2 where f_j = Π_{e: E(j,e)=1} (1 - 2·p_e) is the flip attenuation factor and p_j is the noiseless marginal (0, 0.5, or 1).

Cost: O(nnz(E)). Works for any qubit count.