Microlensing modeling rung: PARALLAX with escalation guidance, evidence-gate checkpoints, and default priors.
name roman-modeling-microlensing-parallax description Microlensing modeling rung: PARALLAX with escalation guidance, evidence-gate checkpoints, and default priors. Roman Modeling Microlensing: PARALLAX Purpose Escalate when long-timescale asymmetry suggests annual parallax. Use This Skill When Current lower-complexity rung leaves structured residual anomalies. Known degeneracies indicate this rung should be tested. Escalation Gate Run this rung only after the previous rung fails anomaly or degeneracy checks. Keep added complexity only if evidence improves enough (use skills/modeling/model-selection/evidence-bic). Default Priors Inherit FSPL/PSPL priors piE_N, piE_E: zero-centered broad normal priors Weak regularization against unphysically large parallax amplitudes Output Contract Return: status: ok | warning | error summary: fit/evidence verdict artifacts: fit products and diagnostics validation: convergence and residual checks provenance: data paths, seed/config, timestamp