Simulated Accelerometer Data for Clustering
收藏DataCite Commons2025-07-07 更新2025-04-16 收录
下载链接：
https://purl.stanford.edu/cg185zq8485
下载链接
链接失效反馈官方服务：
资源简介：
The data provided correspond to an extensive simulation study to evaluate methods for applying prediction strength when the interest lies in identifying clusters based on accelerometer time series. The study contains simulated accelerometer data generated under 28 different scenarios according to procedures outlined in Yu et. al. 2023.
This repository contains 28 zip files. Each zip file corresponds to a different set of data generation parameters (scenarios) and contains 100 simulated data sets per scenario. Each data set is in RDS format.
Additional materials include an Excel spreadsheet (data info.xlsx) containing relevant data generation parameters for each scenario and a data dictionary for the simulated data sets. Included also is a sample R script (exampleCode.R) which can be used to load a data set, visualize its contents, and apply the prediction.strength and pam algorithms.
We are providing the data used in Yu et al along with code for full transparency and to facilitate others who wish to use the methods described in the paper.
Questions may be directed to Kristopher Kapphahn (kikapp@stanford.edu).
提供机构：
Stanford Digital Repository
创建时间：
2023-02-09
