Project Goal
This project builds a transparent, reproducible inference pipeline that confronts three dark-energy models — ΛCDM, CPL and an oscillating equation of state (Ξosc) — with Pantheon+SH0ES supernovae and DESI DR2 BAO distances. The question it asks is deliberately modest: can current distance data accommodate a small oscillatory component in w, and how does that compare with standard parameterisations once model complexity is penalised?
The project is released as v0.1.0 and is exploratory. It does not claim new physics: the aim is to document the pipeline, its constraints and its limits. See Known Limitations.
What the project delivers
| Deliverable | Where |
|---|---|
| Three models with explicit equations and priors | Three Model Families · Prior Specification |
| Combined supernova + BAO likelihood with full covariances | Inference Pipeline |
| MCMC chains, convergence diagnostics and AIC/BIC comparison | Convergence · Model Comparison |
| Constraints, corner plots and the oscillation-amplitude diagnostic | Results |
| A pipeline you can rerun, from a smoke test to the final run | Reproducibility |
Where the project stands
The four figures below summarise the final Protocol 1b run (Pantheon+SH0ES + DESI DR2 BAO). Their reading, and what they do not show, is on Key Findings.