default-header
Home GGIR Software

GGIR Software

I created GGIR in 2013 as a generic solution to handle and analyse the data collected with modern accelerometers. GGIR would not be what it is today without the input I have received from other researchers and developers from around the world, who have provided valuable feedback and contributions in the validation of GGIR’s functionalities.

Open Source

GGIR comes with an open source license, which is essential for reproducible and transparent research. Although open source software is free to download and use, the time investment to professionally maintain it and to support users is not for free. Voluntary community efforts are valuable. Accelting provides commercial support around the user needs that are hard to address by voluntary efforts: Dedicated support, Training, and Substantial software or algorithm enhancements.

Open Source

Process and analyse multi-day data

GGIR helps you to gain insight from multi-day accelerometer recordings. Key GGIR functionalities are:

  • Auto-calibration, an algorithm that automatically corrects for poor sensor calibration based on your own data without the need for additional experiments.
  • Non-wear detection and missing data imputation. A major concern on behavioral monitoring is participant compliance with the study protocol. GGIR allows you to detect when the participant did not wear the accelerometer.
  • Estimation of the Sleep period time window with or without sleep diary to guide the detection. An algorithm that was published in 2018 and has directly been used in high profile papers. Further, estimates are given of sleep efficiency.
  • Freedom to choose your own indicator of magnitude of acceleration, and from that derive estimates of energy expenditure.
  • 24 hours time-use indicators to facilitate research on the compositional nature of human behavior.
Process and analyse multi-day data

Insightful reports

Quantitative reports are generated on data quality and to describe patterns in movements and sleep. Complementary to these reports GGIR generates several visualisations to facilitate additional quality assurance and to aid interpretation of the quantitative reports.

Watch the video for a brief tutorial.

200+ Publications

GGIR is widely used as witnessed by an increasing number of citations from end-users (non-exhaustive list). Studies vary in size from a few dozen participants in clinical studies like the study by Bachasson and colleagues (Neurology 2017), a few thousand in epidemiological cohorts like the publication by Sabia and colleagues (Nature Scientific Reports 2017), and a hundred thousand in biobanks like UK Biobank as shown by Jones and colleagues (Nature Communications 2019).

200+ Publications

GGIR training

GGIR comes with extensive free documentation. Nonetheless, it can be valuable to have someone else guide you through all the essential of GGIR. Therefore, we are offering paid GGIR training options.

Check out GGIR training options
GGIR training

What do others think about GGIR?

"GGIR enables us to efficiently generate numerous movement behaviour outcomes from very large accelerometer datasets. It has become an essential tool for our physical activity research at the Leicester Diabetes Centre."
Alex Rowlands, PhD
Diabetes Research Centre, University of Leicester, Leicester General Hospital, Leicester, UK
"Open-science is the only way forward to allow analyses of accelerometer data. This allows harmonisation of estimators between studies to allow comparison of findings. GGIR has been developed with that purpose and its regular updates make it an cutting-edge and up-to-date software for movement and sleep research."
Séverine Sabia, PhD
Université de Paris, Inserm U1153, France; University College London, UK
"Open-source R package GGIR offers a great functionality and user flexibility by streamlining data processing pipeline and deriving meaningful analysis-ready summaries of quality and timing of sleep, volume and intensity of physical activity, and the strength of circadian rhythms."
Vadim Zipunnikov, PhD
Associate Professor at the Department of Biostatistics at Johns Hopkins Bloomberg School of Public Health

Let’s talk...

and address your research challenges together.