loom

Local Shiny app for building submission-ready clinical Tables, Figures & Listings (TFLs) from ADaM data. Point-and-click interface that generates reproducible R scripts using arframe as the rendering engine.

No server infrastructure needed — runs entirely on your laptop.

Quick Start

git clone git@github.com:vthanik/loom.git
cd loom
Rscript -e "shiny::runApp('.')"

Prerequisites

• R >= 4.1.0
• arframe — rendering engine (install from GitHub)

Install Dependencies

# Install arframe first
remotes::install_github("vthanik/arframe")

# Install CRAN dependencies
install.packages(c(
  "shiny", "bslib", "htmltools", "dplyr", "tidyr",
  "tibble", "readr", "glue", "reactable", "rlang"
))

Or Install as R Package

remotes::install_github("vthanik/loom")
loom::launch()

How It Works

1. Load Data — select ADaM datasets (CDISC pilot data bundled) or upload your own (.rds, .csv)
2. Select Template — choose Demographics (more templates coming)
3. Configure Analysis — treatment variable, group variable, per-variable statistics
4. Format Output — titles, footnotes, columns, headers, page layout, rules
5. Generate Preview — Ctrl+Enter renders the table instantly
6. Export — download RTF/PDF/HTML + standalone R script

Bundled Data

10 CDISC pilot datasets from pharmaverseadam:

Dataset	Rows	Description
ADSL	306	Subject-level demographics
ADAE	1,191	Adverse events
ADCM	7,510	Concomitant medications
ADEX	6,315	Exposure
ADLB	83,652	Laboratory results
ADMH	1,818	Medical history
ADRS	3,694	Tumor response
ADTR	181	Tumor results
ADTTE	512	Time-to-event
ADVS	65,032	Vital signs

Keyboard Shortcuts

Shortcut	Action
Ctrl+1-5	Switch tabs (Data/Template/Analysis/Format/Output)
Ctrl+Enter	Generate preview
Ctrl+S	Export RTF
Ctrl+Shift+S	Download R script
Ctrl+B	Toggle sidebar

Architecture

ADaM Data → ARD (wide) → arframe spec → RTF/PDF/HTML
                ↓
         R script (reproducible)

• arframe = rendering engine (pure R, no UI)
• loom = this app (Shiny UI that drives arframe)
• Generated R scripts are standalone — run anywhere with dplyr + arframe

Project Structure

R/
  app_ui.R / app_server.R    # Shiny app shell
  mod_*.R                     # UI modules (data, template, analysis, format, output)
  fct_ard_demog.R             # Demographics ARD builder
  fct_render.R                # arframe rendering (consumes IR)
  fct_codegen.R               # R script generator (consumes IR)
  fct_spec_ir.R               # Shared intermediate representation
  spec_demog.R                # Demographics template defaults

inst/
  app/www/                    # CSS, JS, highlight.js
  data/                       # Bundled CDISC pilot datasets (.rds)

tests/
  testthat/                   # 57+ regression tests

Corporate Proxy

If behind a firewall for first-time package installs:

Sys.setenv(http_proxy = "http://proxy:port", https_proxy = "http://proxy:port")

License

MIT