README.md

In muschellij2/weartime: Wear Time Estimation from Accelerometer Data

weartime

The goal of weartime is to implement the wear time algorithm from https://github.com/shaheen-syed/CNN-Non-Wear-Time-Algorithm/. This algorithm is used to determine the wear time of a device, such as an accelerometer, that is worn by a person. The algorithm is based on a convolutional neural network (CNN) that is trained to determine if a window of data is wear time or non-wear time. The algorithm is trained on the ActiGraph GT3X+ accelerometer data, but can be used on other accelerometer data as well.

Installation

You can install the released version of weartime from CRAN with:

install.packages("weartime")

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("muschellij2/weartime")

Example

This is a basic example which shows you how to solve a common problem:

library(weartime)
## basic example code

muschellij2/weartime documentation built on April 5, 2025, 4:01 a.m.