Shanshan Chen

Statistical Signal Processing, Wearable Sensing & Digital Health

Research · Software · Publications · Teaching · Notes

CircaCP

Sleep and circadian metrics estimation from actigraphy data

CircaCP provides an unsupervised approach to detecting sleep–wake cycles in unlabeled actigraphy data. It combines a robust cosinor model with change point detection to estimate sleep and wake onset times.

Open Shiny app · R package on CRAN · R source code · MATLAB toolbox

What it supports

Estimation of sleep and wake onset times from actigraphy sequences.
Estimation of circadian rhythm metrics, including mesor, phasor, interdaily stability, and intradaily variability.
Actigraphy quality screening, cosinor modeling, and parametric change point detection.
Segmentation of long activity sequences for physical activity research.

Activity counts, estimated sleep–wake cycles, and matched sleep and wake event markers across seven work schedules — Example CircaCP sleep–wake estimates across different work schedules. This research figure illustrates the method; it is not a screenshot of the Shiny app.

Getting started

For interactive exploration, launch CircaCP Shiny. For an analysis in R, install the package from CRAN:

install.packages("CircaCP")
library(CircaCP)
help(package = "CircaCP")

The CRAN package page provides the reference manual and vignettes. The actigraphy research page describes the methodological background and validation.

Reference

Chen, S. and Sun, X. (2024). Validating CircaCP: a generic sleep–wake cycle detection algorithm for unlabelled actigraphy data. Royal Society Open Science, 11, 231468. doi:10.1098/rsos.231468.

The algorithm was validated against event markers in the MESA Sleep study. For package-specific citation information, use citation("CircaCP") in R.

This site is open source. Improve this page.