feat: add likelihood-function, estimated-CDF, and potential/observed-outcome macros

[d-morrison]

Clears the open-issue backlog (issue #56 intentionally left out of scope). Each issue is a separate commit; every new macro has a matching interpretations.tsv entry, so the macros-table.qmd build still passes.

#64 — likelihood / log-likelihood function forms

Macro	Expands to	Meaning
\Likf{x}	\Lik(x)	likelihood function, ℒ(·)
\loglikf{x}	\llik(x)	log-likelihood function, ℓ(·) (alias of \llikf)
\loglf{x}	\llik(x)	log-likelihood function (alias of \llikf)
\likf{x}	\llik(x)	log-likelihood function (alias of \llikf)

Completes the function-form family for the existing \Lik / \llik symbols. The lowercase-\lik-means-log-likelihood convention is pre-existing (\lik is already an alias of \llik); the TSV entries flag each alias accordingly.

#69 — estimated CDF function macros

Macro	Expands to	Meaning
\cdff{x}	\cdf(x)	CDF applied to argument, F(·)
\hcdf	F̂	estimated CDF symbol
\hcdff{x}	\hcdf(x)	estimated CDF applied to argument, F̂(·)

Mirrors the existing \Q/\Qf/\hQ/\hQf and \surv/\survf/\hsurv/\hSurvf distribution-function families; \hcdf uses the same \mathop{\hat{…}}\nolimits pattern as \hsurv.

#70 — potential-outcome vs observed-outcome notation

Following the convention in the issue — subscripts for potential (intervened) exposures, parentheses for observed exposures:

Macro	Expands to	Meaning
\potout{a}	Y_a	potential outcome under exposure a
\potprob{a}	π_a	potential outcome probability (e.g. Pr(Y_a = 1) for binary Y)
\potmean{a}	μ_a	potential outcome mean (e.g. E[Y_a])
\obsprob{a}	π(a)	observed outcome probability
\obsmean{a}	μ(a)	observed outcome mean

These reuse the existing \prob (π) and \mean (μ) symbols, so they are notation variants of existing estimands rather than a new estimand group, and need no separate hat-wrapped estimators.

Verification

• Replicated the macros-table.qmd build check: every new macro has an interpretations.tsv entry (no missing/orphan entries; no name collisions).
• Loaded the full macros.qmd into a lualatex preamble and rendered every new macro to PDF — all produce the intended symbols (ℒ(θ), ℓ(θ), F(x), F̂, F̂(x), Y_a, π_a, μ_a, π(a), μ(a)).
• Re-rendered demo-shortcode.qmd to PDF to confirm the include-shortcode path still loads macros.qmd cleanly.

Full HTML render via CI (Quarto Preview), since DT/knitr aren't installable against the container's bleeding-edge R.

Closes #64
Closes #69
Closes #70

---

Generated by Claude Code

[d-morrison]

Opened and actively iterating this PR to clear the #64/#69/#70 macro backlog (#56 intentionally left out of scope). Driving it to clean via ARDI — paws off until I'm done.

---

Generated by Claude Code

[Copilot]

Pull request overview

Adds several new LaTeX macros to macros.qmd to complete common notation “families” (function-form likelihood/log-likelihood, estimated CDF, and potential/observed outcome notation), and updates interpretations.tsv so the macros-table.qmd render check continues to pass.

Changes:

• Added \Likf plus log-likelihood function aliases (\loglikf, \loglf, \likf) to mirror existing \Lik / \llik symbols.
• Added estimated CDF macros (\cdff, \hcdf, \hcdff) following the existing hat-wrapped distribution-function pattern.
• Added potential vs observed outcome macros (\potout, \potprob, \potmean, \obsprob, \obsmean) consistent with the subscript-vs-parentheses convention described in the linked issue.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
macros.qmd	Introduces the new likelihood/CDF/potential-outcome macros and aliases.
interpretations.tsv	Adds interpretation rows for each new macro to keep the macros-table merge/check passing.

[d-morrison]

Added short aliases for the potential/observed macros (per request):

Shorthand	Canonical	Renders
\py	\potout	Y_a
\ppi	\potprob	π_a
\pmu	\potmean	μ_a
\opi	\obsprob	π(a)
\omu	\obsmean	μ(a)

p = potential, o = observed, plus the symbol's letter-name. The potential-mean shorthand is \pmu (symbol-based) because \pmean is already the population mean. Each alias has a matching interpretations.tsv entry; verified the interpretation-completeness check and a lualatex render of all five.

---

Generated by Claude Code

[d-morrison]

Fully clean and ready for your merge:

• ✅ build-deploy (Quarto render — exercises all four macros.qmd consumption paths) — success
• ✅ Copilot review — reviewed both files, no comments
• ✅ @claude review — full pass on the head commit, no findings

No open review threads. Merges Closes #64, #69, #70; #56 intentionally left out of scope. Leaving the merge to you.

---

Generated by Claude Code

[github-actions]

PR Preview Action v1.8.1
Preview removed because the pull request was closed.