Document estimator aliases for user discoverability

[igerber]

Summary

• Add alias reference table to README (13 aliases mapped to full names and methods)
• Show # or: DiD comment on Quick Start import in README
• Upgrade quickstart.rst alias tip from Python comment to RST .. tip:: admonition
• Add (alias: ...) labels to all 5 estimator headings in API docs
• Fix wrong alias names in CHANGELOG (DID→DiD, MPID→EventStudy, CSAN→CS, SDID→SDiD)

Methodology references (required if estimator / math changes)

• N/A — no methodology changes

Validation

• Tests added/updated: No test changes (documentation only)
• All 13 aliases verified importable: python -c "from diff_diff import DiD, TWFE, EventStudy, SDiD, CS, CDiD, SA, BJS, Gardner, DDD, Stacked, Bacon, EDiD"
• Alias names cross-checked against diff_diff/__init__.py lines 167-180

Security / privacy

• Confirm no secrets/PII in this PR: Yes

---

Generated with Claude Code

[igerber]

/ai-review

[github-actions]

🔁 AI review rerun (requested by @igerber)

Head SHA: 69574a644df8609fb9d1caf67979449b7b7dcb82

---

Overall Assessment

✅ Looks good

Executive Summary

• This is a docs-only PR. I found no changes to estimator logic, identification assumptions, weighting, variance/SE computation, or defaults, so there is no methodology regression against the registry.
• The alias names documented in the README and changelog are consistent with the public exports in diff_diff/init.py#L168, diff_diff/init.py#L180, and diff_diff/init.py#L244.
• Severity: P3. The new quickstart tip points readers to api/estimators for the “full list,” but that page’s alias example omits EDiD, which is a public alias. This is already tracked as EfficientDiD API-doc debt in TODO.md#L49.
• Severity: P3. The new README section is titled “Estimator Aliases,” but it includes Bacon, which is a decomposition/diagnostic rather than an estimator.
• I could not rerun the PR-body import smoke test in this environment because numpy is not installed; I verified alias existence from source exports instead.

Methodology

• No findings. The diff only changes documentation/changelog text, and the documented aliases match the public API exports in diff_diff/init.py#L168, diff_diff/init.py#L180, and diff_diff/init.py#L244.

Code Quality

• No findings.

Performance

• No findings.

Maintainability

• No findings.

Tech Debt

• Severity: P3. Impact: docs/quickstart.rst#L29 says the API reference has the “full list,” but docs/api/estimators.rst#L24 omits EDiD even though it is exported at diff_diff/init.py#L180 and diff_diff/init.py#L244. This aligns with the existing tracked item in TODO.md#L49, so it is informational, not blocking. Concrete fix: add EDiD to the alias list on docs/api/estimators.rst, or change the quickstart wording from “full list” to “common aliases” until the EfficientDiD API docs land.

Security

• No findings. Docs-only changes and no obvious secret exposure.

Documentation/Tests

• Severity: P3. Impact: README.md#L87 introduces an “Estimator Aliases” table, but README.md#L104 includes Bacon, which maps to BaconDecomposition. The methodology registry classifies Bacon under Diagnostics & Sensitivity at docs/methodology/REGISTRY.md#L22, and the class itself is documented as a TWFE decomposition at diff_diff/bacon.py#L288. Concrete fix: rename the section to something like “Public Aliases” or split Bacon into a separate diagnostic-alias note.
• No testing finding from the diff. I could not rerun the alias-import smoke test locally because the review environment is missing numpy.