No history yet

Introduction to Teal

From Static Reports to Interactive Apps

Imagine you're analyzing data from a clinical trial. You have patient demographics, lab results, and treatment outcomes. Your team asks for a report on the safety profile of a new drug. You create it. Then they ask, "What if we only look at patients over 50?" You create another report. Then, "How about just females over 50 in the high-dose group?" You're now stuck in a cycle of re-running scripts and generating endless static reports.

This is a common bottleneck in data analysis. The R package teal is designed to solve this problem. It's a framework for building interactive web applications for data exploration, built on top of R's popular shiny package. While it was created with clinical trial data in mind, its tools are flexible enough for many types of analysis.

teal turns your data analysis from a one-way street of static reports into a two-way conversation with your data.

Core Functionalities

teal's power comes from a set of core features that work together seamlessly. At its heart, it provides a structured environment where different analytical components, or modules, can coexist.

Key features include:

  • Filtering: A built-in, powerful filtering panel allows users to slice and dice the data in real-time without writing any code. This is the feature that breaks the static report cycle.

  • Reproducibility: Every filter, selection, and analysis step a user takes in a teal app is captured. The app can generate the R code needed to reproduce the exact state of the analysis, which is crucial for validation and documentation.

  • Modularity: teal apps are built from modules. You might have one module for demographic tables, another for a specific type of plot, and a third for safety analysis. This modular design makes the code organized and reusable.

Lesson image

How Teal Works with Shiny

If you've used shiny, you know it provides the foundation for web apps in R. So, where does teal fit in? teal is not a replacement for shiny; it's a high-level framework that handles much of the boilerplate code you would otherwise write yourself.

Think of shiny as a box of tools: UI elements, server logic, and reactivity. You can build anything with it, but you have to assemble everything from scratch. teal, on the other hand, is like a pre-assembled toolkit specifically for data exploration. It provides the application layout, the filtering logic, and the connections between modules, letting you focus on the analysis itself.

This pre-built structure is what makes development with teal so rapid. Let's see what a minimal teal app looks like. We'll use the classic iris and mtcars datasets.

library(teal)

# The datasets we want to explore
data <- teal_data(
  iris = iris,
  mtcars = mtcars
)

# Define the UI and Server for the app
app <- init(
  data = data,
  modules = modules(
    # A basic module to show the data
    tm_data_table(
      label = "View Data"
    )
  )
)

# Run the app
shinyApp(app$ui, app$server)

When you run this code, you get a fully functional web app. It has a sidebar with filtering options for both datasets and a main panel showing the data table. teal handled the layout, the filter logic, and the connection between them automatically. You only had to provide the data and specify which module to use.

Quiz Questions 1/5

What is the primary problem the teal package is designed to solve in data analysis?

Quiz Questions 2/5

How does teal relate to the shiny package?

By providing a standardized framework, teal helps teams build consistent, maintainable, and highly effective tools for data exploration.