Shiny application for Systematic measurement changes: A simulation study

This repository contains the source code for the Shiny application accompanying the paper
“Assessing systematic measurement changes in single-wave study data: A simulation study”. The simulation study underlying this app, including code to reproduce all results, is available in a separate GitHub repository.

🚀 Try the app

The interactive Shiny app is available online at: https://sys-measurement-changes.bips.eu/.

The app allows users to explore results from a range of simulation settings across different sample sizes using LOESS-based visualizations and boxplots. All resulting figures can be downloaded directly from the application, and brief usage instructions are provided within the app interface.

📖 About the app

This application enables interactively visualization of results from a large simulation study investigating systematic changes in measurements. The study compares seven statistical methods for quantifying systematic changes in measurements across different settings:

• Autoregressive integrated moving average (ARIMA)
• Fused lasso signal approximator (FLSA)
• Generalized additive model (GAM)
• Locally weighted scatterplot smoothing (LOWESS)
• Moving average (MA)
• Pruned exact linear time (PELT)
• Piecewise regression (PR)

Users can explore how these methods behave under different data distributions, signal-to-noise ratios, and systematic change patterns.

📁 Repository structure

• data/: Contains the results of the simulation study.
• renv/: Contains project-specific environment configuration files.
• output_shiny_application.pdf: All figures from the Shiny application. This corresponds to Supplementary data 3 of the paper.
• renv.lock: Package versions required to run the Shiny application locally (R 4.4.3).
• renv.lock.server-r4.1: Package versions used for hosted server deployment (R 4.1.2).
• server.R: Shiny server logic.
• ui.R: Shiny user interface definition.

📚 R environment (for local development)

This project uses the renv R package to manage package versions for reproducibility. Developers who wish to run the app locally can restore the exact package versions recorded in renv.lock using:

install.packages("renv")
renv::restore()

Note: This step is not required to use the hosted version of the app.