Bump version to 3.7.0#661
Conversation
Release the work merged to main since v3.6.2 (41 commits): the ContinuousDiD capability push (covariate support, discrete-treatment saturated regression, lowest-dose-as-control), the systematic Callaway/staggered standard-error parity sweep vs canonical R backed by the se-audit test locks, EfficientDiD and Callaway performance work (incl. Rust batched-Cholesky and HC2 kernels), imputation leave-one-out variance, and the completed removal of the deprecated HAD survey=/weights= kwargs across all 8 surfaces. Converts the curated [Unreleased] CHANGELOG section to [3.7.0] - 2026-07-08 (leading with the Removed breaking change) and syncs the version string across __init__.py, pyproject.toml, rust/Cargo.toml, llms-full.txt, and CITATION.cff. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Overall assessment✅ Looks good — no unmitigated P0/P1 findings. Executive summary
MethodologyNo findings. Impact: This PR does not change estimator implementations or methodology behavior. The only methodology-adjacent changelog text concerns previously shipped HAD survey API consolidation, which is documented in the methodology registry. Concrete fix: None. Code QualityNo findings. Impact: The changed files are straightforward metadata/string updates. Concrete fix: None. PerformanceNo findings. Impact: No runtime code paths are changed. Concrete fix: None. MaintainabilityNo findings. Impact: Release metadata is synchronized across Python, Rust, citation, changelog, and bundled guide surfaces. No Concrete fix: None. Tech DebtNo findings. Impact: The PR does not add deferred work or new limitations. Concrete fix: None. SecurityNo findings. Impact: No secrets, credentials, network logic, or executable workflow changes are introduced. Concrete fix: None. Documentation/TestsNo blocking findings. Severity: P3-informational |
Summary
3.6.2->3.7.0across all surfaces (diff_diff/__init__.py,pyproject.toml,rust/Cargo.toml,diff_diff/guides/llms-full.txt,CITATION.cff).## [Unreleased]CHANGELOG section to## [3.7.0] - 2026-07-08and add thev3.6.2...v3.7.0comparison link. No auto-generated content - the hand-written entries (leading with### Removed) are preserved verbatim.This is a minor release cutting the 41 commits merged to
mainsince v3.6.2. Highlights:Removed (breaking, deprecation-cycle complete):
survey=/weights=kwargs removed from all 8 surfaces -HeterogeneousAdoptionDiD.fit(refactor(had): remove deprecated survey=/weights= kwargs from HeterogeneousAdoptionDiD.fit (3.7.0) #633) and the 7 pretest helpers (refactor(had): remove deprecated survey=/weights= kwargs from the 7 pretest helpers (3.7.x) #659). Onlysurvey_design=remains. Deprecation warnings shipped since v3.3.1.Added:
allow_unbalanced_panelRC-on-panel parity with Rdid(feat(callaway): allow_unbalanced_panel — RC-on-panel parity with R did #631); imputation leave-one-out conservative variance (BJS 2024 App. A.9) (feat(imputation): leave-one-out conservative variance (BJS 2024 App. A.9) #630); Rust one-way HC2 robust vcov kernel (feat(rust): one-way HC2 robust vcov kernel #651).Changed / performance:
Fixed / SE parity (behavior changes on reported standard errors):
vcov_typecontract under replicate designs (fix(survey): unify explicit-vcov_type contract under replicate designs (warn + hc1 remap) #654) - backed by the se-audit machine-precision test locks (test(se-audit): CI-lock flagship + previously-unasserted SE parity (G2 + C1-C3/C5) #624/test(se-audit): fiddly-bits SE parity locks (C2 DOF, C3 df, C4 intercept SE, C5 Yatchew p/sigma2, G2 cluster band) #628/test(se-audit): G2 hetero + cluster machine-precision fixest locks on unbalanced heteroskedastic DGP #644/test(callaway): machine-precision golden tolerances — SE-audit C6 closure #658).Methodology references (required if estimator / math changes)
docs/methodology/REGISTRY.mdupdates.Validation
Release note
Fixed/Changedsections for specifics.Security / privacy
Generated with Claude Code