ActivityIndex is an R package which provides functions to read and process raw accelerometry data.
ActivityIndex provides R functions to read raw accelerometry data collected by accelerometers. Specifically, it can directly handle .csv data files generated by accelerometer model GT3X+ and software ActiLife by ActiGraph. The package also provides functions to calculate summarizing metrics such as Activity Index (AI), using the raw data.
The AI is a way of summarizing densely sampled accelerometry data into given epochs (such as every 1 second or every 15 seconds, etc). Essentially, AI describes the variability of the raw acceleration signals, after normalizing it using systematic noise of the signal. AI is an evolvement of the original metric, Activity Intensity, proposed in the paper Normalization and extraction of interpretable metrics from raw accelerometry data by J. Bai et al (2014). The AI addresses some limitation of the orignal Activity Intensity, and has favorable properties. More details on these properties and a direct comparison of AI versus other metrics such as Activity Count could be found in the paper An Activity Index for Raw Accelerometry Data and Its Comparison with Other Activity Metrics.
ActivityIndex software can be installed via GitHub. Users should have
R installed on their computer before installing ActivityIndex.
Install devtools package
To install ActivityIndex package via GitHub, the user must have installed devtools, which could be completed by using the following R command
Install ActivityIndex package
The following R command can be used to install the latest version of ActivityIndex via GitHub:
ActivityIndex User Manual
The ActivityIndex package includes a vignette to demonstrate a typical work flow of computing AI. The vignette can either be accessed by R command
and interactively browsing or going to http://javybai.github.io/ActivityIndex/articles/ActivityIndexIntro.html.
Please contact the author and maintainer Jiawei Bai (jbai [at] jhsph [dot] edu) or open an issue at GitHub.