Skip to contents

🧊 sugaR

Continuous glucose monitoring analysis for chronobiology research.

License: MIT R

⚠️ Early development. sugaR is a fresh scaffold — the pipeline below describes the intended functionality, but no functions are implemented yet. See NEWS.md for current status.

📖 What is sugaR?

sugaR processes continuous glucose monitoring (CGM) data for circadian and chronobiology research. It smooths raw interstitial glucose traces, computes an adaptive pre-wake baseline, detects the first post-wake glucose peak, and derives cycle-level regularity metrics — most notably fGPTstd, the day-to-day standard deviation of first post-wake glucose peak timing.

sugaR is a peer package to zeitR (actigraphy) and slumbR (sleep diaries) within the Circadia Lab ecosystem, intended to feed into syncR::sync() alongside those packages once published.

✨ Features

  • 📉 Savitzky-Golay smoothing of CGM traces
  • ⚖️ Adaptive baseline computation with diabetic/non-diabetic tolerances
  • 📈 First post-wake glucose peak detection with dynamic prominence/distance
  • 🔀 Hierarchical onset/offset determination
  • 🌙 Sleep-wake-anchored glucose cycle construction and QC
  • 📊 First post-wake glucose peak timing variability (fGPTstd)
  • 🖼️ Actogram-style plotting for visual QC of automated peak detection

🗂️ Project Structure 

sugaR/
├── R/
│   ├── smoothing.R        # Savitzky-Golay smoothing
│   ├── baseline.R         # Pre-wake baseline computation & correction
│   ├── peak-detection.R   # Candidate peak detection & grouping
│   ├── onset-offset.R     # Peak onset/offset rules
│   ├── cycle.R            # Sleep-anchored cycle construction & QC
│   ├── fgptstd.R          # fGPTstd metric
│   ├── plot-actogram.R    # Visual QC plotting
│   ├── constructors.R     # sugaR_trace / sugaR_cycle S3 objects
│   └── sugaR-package.R    # Package-level documentation
├── tests/testthat/        # Unit tests (currently stubs)
├── dev/                   # Real-data smoke tests (never committed with data)
└── man/                   # Generated documentation

🚀 Getting Started

Prerequisites

  • R >= 4.1.0
  • signal package (Savitzky-Golay filtering)

Installation 

# Not yet published to r-universe. Once functional:
remotes::install_github("circadia-bio/sugaR")

Basic usage (planned) 

library(sugaR)

trace <- glucose_trace(time = cgm$time, glucose = cgm$glucose, diabetic = FALSE)
cycle <- build_glucose_cycle(cgm = cgm, sleep_episode = sleep_episode)
cycle <- validate_cycle(cycle)

fgptstd <- compute_fgptstd(cycles = participant_cycles)

📦 Dependencies

Package Purpose
signal Savitzky-Golay filtering
stats Base statistical functions
ggplot2 (Suggests) Actogram plotting
testthat (Suggests) Unit testing
  • 🌙 zeitR — wrist actigraphy analysis
  • 📔 slumbR — sleep diary processing
  • 🔄 syncR — ecosystem integrator, pulling sugaR/zeitR/slumbR/tallieR into a unified participant database
  • 🔬 circadia-bio — the Circadia Lab GitHub organisation

📄 Licence

Released under the MIT License.

Copyright © sugaR authors, 2026