← What's New

Tier 0.1 — Full PLDS identifiability (canonical gauge)

2026-05-28 · nstat.extras.em.dynamax_bridge · PR #113

The problem

The PP_EM (fit_point_process_em) and mPPCO_EM (fit_hybrid_em) trainers fit a Poisson linear dynamical system, which is gauge-free up to the full GL(d) group: the reparameterization (A, C, x) → (TAT⁻¹, CT⁻¹, Tx) leaves the observable log-rate Cx — and hence the likelihood — invariant. Only the d scale degrees of freedom were pinned, so A/C drifted across random seeds.

A previous attempt to canonicalize every iteration destabilized the EM (it reshapes the optimization landscape each step and fights the Newton trust-region) — producing NaN and across-seed |ΔC| ≈ 460 on identical data.

The fix

Result

QuantityBeforeAfter
Across-seed |ΔC| (PP_EM)~460 (+ NaN)~0.75
Across-seed |ΔC| (hybrid)large~0.15
Returned CᵀCarbitrarydiag(S²), machine-precision
Hybrid Gaussian R~0.09~0.09 (unchanged, true 0.09)

What remains is local-optima multiplicity (genuinely different likelihoods across seeds), not gauge freedom — addressed by multi-restart selection (Tier 0.3).

Files changed