Skip to content

v3.3.0

Choose a tag to compare

@igerber igerber released this 25 Apr 11:48
· 170 commits to main since this release
9c908f1

What's Changed

  • Precompute stratum-PSU scaffolding in aggregate_survey (~19x BRFSS speedup) by @igerber in #338
  • Close BR/DR gap #4: canonical-dataset regression guards + wording fixes by @igerber in #341
  • Phase 1c: Bias-corrected local-linear CI (CCT 2014) by @igerber in #340
  • Delete compute_synthetic_weights shim; inline Frank-Wolfe in rank_control_units by @igerber in #344
  • Remove dead Rust compute_synthetic_weights (follow-up to PR #344) by @igerber in #345
  • Unify Rust TROP inner solver to SVD (close finding #23 grid-search divergence) by @igerber in #348
  • Phase 2a: HeterogeneousAdoptionDiD class (single-period, 3 design paths) by @igerber in #346
  • Close BR/DR gap #6: target-parameter block in schemas by @igerber in #347
  • Fix SyntheticDiD bootstrap p-value dispatch and SE formula by @igerber in #349
  • HAD Phase 2b: multi-period event-study extension (Appendix B.2) by @igerber in #350
  • HAD Phase 3: pre-test diagnostics (qug_test, stute_test, yatchew_hr_test, composite workflow) by @igerber in #352
  • Add SyntheticDiD variance_method='bootstrap_refit' and coverage MC study by @igerber in #351
  • Fix TROP bootstrap SE backend divergence under fixed seed by @igerber in #354
  • HAD Phase 3 follow-up: joint Stute pretest + event-study workflow by @igerber in #353
  • dCDH: add by_path per-path event-study disaggregation by @igerber in #357
  • Restore SDID survey-bootstrap via weighted Frank-Wolfe + Rao-Wu composition by @igerber in #355
  • Fix TROP local-method backend parity: drop Rust weight normalization + Python cache-fallthrough by @igerber in #358
  • HAD Phase 4.5: survey support on continuous-dose paths by @igerber in #359
  • dCDH by_path R-parity fixtures + TestDCDHDynRParityByPath by @igerber in #360
  • Fix CHANGELOG: rename survey-bootstrap PR placeholder #352 to #355 by @igerber in #361
  • Add profile_panel() + autonomous LLM guide for agent-facing workflows by @igerber in #356
  • HAD Phase 4.5 B: weighted mass-point 2SLS + event-study survey composition + sup-t bootstrap by @igerber in #363
  • dCDH by_path + n_bootstrap support (library-consistent percentile CI) by @igerber in #364
  • Restore SDID survey support for placebo and jackknife variance methods by @igerber in #365
  • HAD Phase 4.5 C0: QUG-under-survey decision gate by @igerber in #367
  • Release 3.3.0: HAD estimator, profile_panel, dCDH by_path, SDID survey complete by @igerber in #368

Full Changelog: v3.2.0...v3.3.0