clerkR — Publication-ready tables for R • clerkR

A clerk keeps tabs — clerkR keeps yours publication-ready.

⚠️ clerkR is in early development and has not been formally tested. The API may change without notice, estimation results have not yet been validated against a reference implementation, and the package has not undergone peer review. Use with caution and verify outputs independently before using in any research context.

📖 What is clerkR?

clerkR transforms standard R data frames into publication-ready tables for biomedical and neuroscience manuscripts. Rather than wrestling with formatting each time, clerkR reduces the most common table types to a handful of opinionated constructor functions that share consistent theming, domain grouping, footnote handling, and a unified rendering pipeline for Word/PDF (gt), interactive HTML (reactable), and LaTeX output.

✨ Features

📋 Five table archetypes covering ~90% of what appears in a biomed paper
🎨 clerkR theme — light teal headers, navy text, clean borders, consistent throughout
🗂️ Domain/section grouping — organise rows under labelled section headers, with optional nested sub-sections (e.g. repeated timepoints within a domain)
📝 Footnotes — automatic notes for log-transformed variables and FDR correction, plus your own blanket or row/column-targeted footnotes
🖨️ Three render targets — gt for Word/PDF, reactable for HTML, LaTeX for manuscripts
🔢 Output baked in at construction — set output = "gt" once, then just |> clerk_render()
🔗 R-itable compatible — herit_batch() output pipes straight into tbl_heritability()
🧩 Composable — all constructors return a clerk_tbl S3 object

📋 Table archetypes

Function	Use case	Example
`tbl_descriptive()`	Sample characteristics by group, mean ± SD, t/χ²	Table 1
`tbl_simple()`	Descriptive summary, no inferential test	Supplementary table
`tbl_correlation()`	Partial correlations, r, p, p†	Correlation results
`tbl_regression()`	β, SE, 95% CI, p, FDR — accepts `broom::tidy()`	Linear/logistic models
`tbl_heritability()`	h², 95% CI, LRT p, σ²a/σ²e — accepts `herit_batch()`	Heritability results

🚀 Getting Started

Installation

# clerkR, from r-universe
install.packages("clerkR", repos = "https://circadia-bio.r-universe.dev")

# For the heritability workflow, also install R-itable (same repo)
install.packages("Ritable", repos = "https://circadia-bio.r-universe.dev")

Or from GitHub directly:

remotes::install_github("circadia-bio/clerkR")
remotes::install_github("circadia-bio/R-itable")

The one-two pattern

library(clerkR)

tbl_descriptive(
  clerk_example,
  group   = sex,
  domains = list(
    "Metabolic"      = c("hdl", "glucose", "bmi"),
    "Anthropometric" = c("waist", "systolic_bp"),
    "Cognitive"      = c("tmt_time", "verbal_fluency"),
    "Mental health"  = c("bdi", "panas_neg", "life_satisfaction")
  ),
  log_vars = "tmt_time",
  fdr      = TRUE,
  output   = "gt"
) |>
  clerk_render(title = "Table 1. Sample characteristics by sex")

Heritability from R-itable

herit_batch(traits, grm = A, data = dat, covs_list = covs_list) |>
  tbl_heritability(
    model    = "covariates",
    sigma2_a = "sigma2_a",
    sigma2_e = "sigma2_e",
    fdr      = TRUE,
    output   = "gt"
  ) |>
  clerk_render(title = "Heritability estimates")

Nested domains & custom footnotes

A domain can itself hold sub-sections — handy for repeated timepoints — and footnotes aren’t limited to one blanket note per table:

tbl_descriptive(
  longitudinal_example,
  group   = sex,
  domains = list(
    "Mental health" = list(
      "Baseline"    = c("bdi_bl",  "panas_neg_bl"),
      "Follow-up 1" = c("bdi_fu1", "panas_neg_fu1")
    )
  ),
  output = "gt"
) |>
  clerk_render(
    title     = "Mental health by timepoint",
    footnote  = c("Data collected 2024-2025.", "Listwise exclusion applied."),
    footnotes = list(list(text = "Self-report questionnaire.", rows = "bdi_bl"))
  )

See vignette("formatting-options") for the full write-up, including how nested domains render (compound row-group label for gt/latex, a true expandable tree for output = "html").

🎨 Colour palette

clerk_palette()     # full named palette
clerk_diverging()   # terracotta → off-white → navy (9 steps)
clerk_sequential()  # near-white → navy (7 steps)

📦 Dependencies

Package	Version	Purpose
`dplyr`	≥ 1.1.0	Data manipulation
`tidyr`	any	Reshaping
`rlang`	any	Tidy evaluation
`gt`	≥ 0.10.0	Word/PDF table rendering
`reactable`	≥ 0.4.0	Interactive HTML rendering
`htmltools`	any	Reactable title/footnote chrome
`knitr`	any	LaTeX output
`grDevices`	any	Colour ramps
`stats`	any	t-test, ANOVA, chi-squared, FDR correction
`utils`	any	Internal helpers

👥 Authors

Role	Name
Author, maintainer	Lucas França
Author	Mario Leocadio-Miguel

⌚️ zeitR — actigraphy analysis and circadian metrics
🛌 slumbR — sleep diary processing
🧮 tallieR — questionnaire and sociodemographic data
🔄 syncR — integrates zeitR, slumbR, and tallieR
🧬 R-itable — pedigree-based heritability estimation
🔬 circadia-bio — the Circadia Lab GitHub organisation

📄 Licence

Released under the MIT License.

📋 clerkR