Skip to content

docs: Tobit half-life now runs end-to-end through pk.nca() - #4

Open
billdenney wants to merge 1 commit into
mainfrom
claude/tobit-pipeline-lloq
Open

docs: Tobit half-life now runs end-to-end through pk.nca()#4
billdenney wants to merge 1 commit into
mainfrom
claude/tobit-pipeline-lloq

Conversation

@billdenney

Copy link
Copy Markdown
Member

Summary

Upstream humanpred/pknca issue 553
(closed 2026-06-11) wired the LLOQ from PKNCAconc() through to the Tobit
half-life method, so hl_method = "tobit" now works through the full
pk.nca() pipeline and the book's direct-call workaround note is obsolete.

Verified empirically against PKNCA 0.12.1.9000 (GitHub eff79252, packaged
2026-07-24, R 4.6.1):

  • PKNCAconc(..., lloq = 0.1) + PKNCAdata(options = list(hl_method = "tobit"))
    • pk.nca() runs cleanly and returns λz = 0.2659 / half-life = 2.607 h —
      identical to the direct pk.calc.half.life(..., hl_method = "tobit") call
      and distinct from the log-linear control (2.393 h), with the two BLQ points
      reported in lambda.z.n.points_blq.
  • lloq works as a numeric scalar and as a column name.
  • Omitting lloq still errors ("lloq must be provided when hl_method is
    'tobit'"), so the LLOQ on the concentration object is required.

Changes

  • user/halflife.qmd — replaced the "tracked for a future PKNCA update"
    workaround note with a runnable end-to-end pipeline example in a new
    section, "Tobit through the pk.nca() pipeline (≥ 0.12.2)".
  • CONTRIBUTING.md — the Tobit known-gotcha now documents the remaining
    pitfall: supply the LLOQ via PKNCAconc(lloq = ).

quarto render passes with no warnings or errors; the rendered chapter shows
the pipeline and direct-call results matching.

🤖 Generated with Claude Code

Upstream humanpred/pknca issue 553 (closed 2026-06-11) added an lloq
argument to PKNCAconc() (numeric scalar or column name) and wired it
through to pk.calc.half.life(), so hl_method = "tobit" set via
PKNCAdata(options = ...) no longer needs the direct pk.calc.half.life()
workaround. Verified against the installed PKNCA 0.12.1.9000 (GitHub
eff79252, packaged 2026-07-24): the pipeline result matches the direct
Tobit call exactly and errors only when lloq is absent from PKNCAconc.

- user/halflife.qmd: replace the pipeline-workaround note with a
  runnable end-to-end example, tagged (>= 0.12.2)
- CONTRIBUTING.md: update the Tobit known-gotcha to the remaining
  pitfall (lloq must be set on PKNCAconc)

Co-Authored-By: Claude Fable 5 <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant