README.md

In oslerinhealth-releases/adept: Adaptive Empirical Pattern Transformation

adept

adept package implements ADaptive Empirical Pattern Transformation (ADEPT) method for pattern segmentation from a time-series. ADEPT was designed for optimal use in performing fast, accurate walking strides segmentation from high-density data collected from a wearable accelerometer worn during continuous walking activity.

Installation

# install.packages("devtools")
devtools::install_github("martakarass/adept")

Vignettes

Vignettes are available to better explain package methods functionality.

Vignette 1. Introduction to adept package

Vignette Introduction to adept package intends to introduce a reader to the ADEPT method and demonstrate the usage of the segmentPattern {adept} function which implements ADEPT method. Here, we focus on illustrating segmentPattern {adept} functionality with a comprehensive set of simulated data examples.

Specifically, we show how to use adept segment pattern occurrences from a (noisy) signal in the presence of:

• different shapes of pattern in a time-series,
• various lengths of pattern occurrences in a time-series.

Vignette 2. Walking strides segmentation with adept

Vignette Walking strides segmentation with adept provides an example of walking stride segmentation from subsecond accelerometry data with adept package. We demonstrate that ADEPT method can be used to perform automatic and precise walking stride segmentation from data collected during a combination of running, walking and resting exercise. We demonstrate how to segment stride pattern:

1. with the use of stride templates that were pre-computed based on data from an external study,
2. by deriving new stride templates in a semi-manual manner.

Accelerometry data collection

Accelerometry data visualization

Segmentation results

oslerinhealth-releases/adept documentation built on Nov. 4, 2019, 11:11 p.m.