v0.1.0

0.1.0

First release.

• Re-export the deterministic reserving primitives from actuarialpy.reserving
  under the reservingmodels namespace (aliases, not copies): ChainLadder,
  completion_factors, apply_completion, develop_ultimate (chain-ladder,
  Bornhuetter-Ferguson, Benktander, Cape Cod), make_completion_triangle,
  Mack standard errors, and the supporting helpers.
• Add BootstrapODP: the semiparametric over-dispersed-Poisson bootstrap
  (England & Verrall) of the predictive reserve distribution -- estimation
  error by Pearson-residual resampling, process error by a gamma draw.
  Validated against Mack on the Taylor-Ashe triangle (same reserve; dispersion
  phi = 52,601).
• Add ReserveDistribution with predictive percentiles, VaR, TVaR (ecosystem
  empirical estimators), and a per-origin exhibit.
• BootstrapODP.sample(size, rng) implements the risksim sampling seam, so a
  reserve model is a portfolio component with no dependency between packages.
• Add residual diagnostics: pearson_residuals, residual_summary,
  calendar_year_effects.
• Add integrations.actuarialpy to build a triangle (and a fitted bootstrap)
  from a claims-listing actuarialpy.Experience.
• Add reference triangles datasets.taylor_ashe() and datasets.raa().