Uploaded March 2025 | Updated September 2026, 2 weeks ago
#ShinyConf #ShinyConf2024
Abstract: Your shiny app is slow. What should you do? ….
Google Lighthouse is a tool that assesses the time taken to start any web app (among other things). It provides a report that can be analysed in R. Here, Lighthouse is used to assess a range of different Shiny apps to see if there are any common issues that slow down these apps.
The start-up performance for over one hundred applications from the Shiny Contest 2021 were analysed using Lighthouse. For those apps that were still working, Lighthouse report-generation was automated using R.
Unsurprisingly, simpler apps tended to load quicker than more complicated apps, although several complex apps were found with very good Lighthouse scores. Consequently, I sought to quantify the impact on start-up times of making a shiny app more complex. This was done by creating a simple app and iteratively adding more features (using {plotly} and {DataTable}). Lighthouse was used to reveal the effects of making an app incrementally more complex, with the additional widgets increasing start-up times.
This will demonstrate how to use Google Lighthouse on your own Shiny apps. I will cover the analysis performed on the data obtained from the Lighthouse reports. Finally, I will cover the results from my experiments on the impact of different features on Shiny apps and discuss the areas where Lighthouse may not give a complete picture of the start-up issues in your application.|
____________________________
Join us this year! Sign up for ShinyConf 2025 here: go.appsilon.com/register-shinyconf2025
#ShinyConf #ShinyConf2024
Abstract: Your shiny app is slow. What should you do? ….
Google Lighthouse is a tool that assesses the time taken to start any web app (among other things). It provides a report that can be analysed in R. Here, Lighthouse is used to assess a range of different Shiny apps to see if there are any common issues that slow down these apps.
The start-up performance for over one hundred applications from the Shiny Contest 2021 were analysed using Lighthouse. For those apps that were still working, Lighthouse report-generation was automated using R.
Unsurprisingly, simpler apps tended to load quicker than more complicated apps, although several complex apps were found with very good Lighthouse scores. Consequently, I sought to quantify the impact on start-up times of making a shiny app more complex. This was done by creating a simple app and iteratively adding more features (using {plotly} and {DataTable}). Lighthouse was used to reveal the effects of making an app incrementally more complex, with the additional widgets increasing start-up times.
This will demonstrate how to use Google Lighthouse on your own Shiny apps. I will cover the analysis performed on the data obtained from the Lighthouse reports. Finally, I will cover the results from my experiments on the impact of different features on Shiny apps and discuss the areas where Lighthouse may not give a complete picture of the start-up issues in your application.|
____________________________
Join us this year! Sign up for ShinyConf 2025 here: go.appsilon.com/register-shinyconf2025
