Skip to contents

Introduction to dataviewR

Source: vignettes/Introduction-to-dataviewR.Rmd
Introduction-to-dataviewR.Rmd

Introduction

dataviewR provides a clean and intuitive Shiny interface for interactively exploring R data frames and tibbles. Instead of using the default View() from the utils package or writing temporary filter code, you can open a dataset visually, browse its structure, filter rows, inspect values, and export results — all within seconds. Unlike View(), dataviewR does not dynamically reflect when the dataset is altered in RStudio after being loaded.

Why dataviewR?

Whether you work in clinical research, data analysis, or academia, a strong understanding of datasets is important. dataviewR helps you:

  • Inspect datasets interactively within seconds
  • View two or more datasets side-by-side
  • Review data quality and structure before analysis

While functions like View() offer a quick static snapshot of a dataset, they are limited when working with large datasets, inspecting variable values, or applying interactive filters.

dataviewR offers a flexible, dynamic interface that enables efficient and intuitive data exploration, whether within the viewer pane or through a standalone R Shiny application.

dataviewR does not modify any values/attributes of the dataset(s) passed

This guide walks you through how to launch, navigate, and work with datasets in dataviewR.

For feature-specific articles and examples, visit the Articles section of the documentation.

Launching dataviewR

There are three ways to start the application, depending on your workflow:

1️⃣ Launch with a dataset

When you pass a dataset(s), dataviewR opens directly with that dataset(s) loaded.

For better user experience, it is recommended to open dataviewR in new window from Viewer pane.

To open the dataset in a shiny application, user is requested to disable the background option using background = FALSE argument but it keeps the console busy.

library(dataviewR)

dataviewer(iris) # Opens in RStudio Viewer pane or default web browser

dataviewer(iris, background = FALSE) # Opens the Shiny application and run in foreground to enable the "Import Dataset" tab alongside 'iris'

You’ll see the interactive table immediately, along with variable info and filtering options.

dataviewR viewing the iris dataset

2️⃣ Launch with multiple datasets (side-by-side viewing)

You can now pass two or more datasets to view them within the same session.

dataviewer(iris, mtcars) # Opens multiple datasets in separate tabs

This is useful for:

  • Comparing raw vs derived datasets

  • Inspecting multiple SDTM/ADaM datasets

  • Reviewing train vs test splits

  • Comparing outputs across pipelines

dataviewR comparing iris and mtcars

Each dataset gets its own tab, allowing you to switch seamlessly.

Please check the Exploring Multiple Datasets article for more examples.

3️⃣ Launch without data (use the Import Panel)

Calling dataviewer() with no arguments opens the Import Panel in shiny application by default, letting you load the datasets either from the global environment or from the loaded packages.

Once imported, the viewer opens the selected dataset(s) in separate tab(s).

dataviewR import panel

What Happens Inside the App?

When a dataset is loaded, you can:

✔ Apply flexible row filters (with dplyr style expressions)
✔ Select or deselect variables and sort columns
✔ See distinct values for character/factor columns (including NA) in the quick filter box.
✔ View metadata such as variable classes and attributes
✔ Export filtered data
✔ Copy auto-generated dplyr code for reproducibility

This helps integrate interactive exploration into a scripted workflow.

Typical Workflow Example 

library(dataviewR)

# Explore data interactively
dataviewer(iris)

Hover to watch how dataviewR helps you explore a dataset and generate reproducible code. 

# After exploring, use the exported code:
library(dplyr)

iris |>
  filter(Species == "setosa" & Petal.Width > 0.2) |>
  select(Petal.Length, Petal.Width, Species)

The combination of interactive exploration + reproducible code makes dataviewR a strong addition to both analysis and QC workflows.

Next Steps

To dive deeper into individual features, see:

All available under the Articles section of the package documentation.

Feedback

If you have suggestions or run into issues, please visit the GitHub repository: https://github.com/madhankumarnagaraji/dataviewR

We welcome ideas, feedback, and contributions.

🏁 That’s It!

You’re ready to start using dataviewR for fast, intuitive, and reproducible data exploration.