Interactive Pivot Tables and Charts from R

Why Another Interactive Charts Package?

Drag-and-drop plotting is useful for quick EDA: glance at how the data is shaped, spot outliers. It also works well when you need to share a playground with someone less technical or on a different stack — any browser and most modern IDEs render HTML just fine.

R already has good options here. GWalkR positions itself as a Tableau-like drag-and-drop interface. esquisse lets you build ggplots interactively. Both are great for a GUI-first workflow.

perspectiveR does something a bit different. It’s an htmlwidgets binding for FINOS Perspective, a WebAssembly data engine originally built for financial dashboards. The focus is on pivot tables with live grouping, splitting, filtering, and multiple chart types — all running client-side. It also pairs well with Shiny: data updates stream in without re-rendering the chart, which makes for a smooth experience.

# Stable release
install.packages("perspectiveR")

# Development version --- new features and bug fixes land here first
remotes::install_github("EydlinIlya/perspectiveR")

library(perspectiveR)

Just the Data

# adjust height to taste
perspective(mtcars, height = "700px")

That’s it. Take a look at the settings panel (top-right). From there you can drag columns into Group By, Split By, add filters, change sort order, switch between 11 chart types, and pick a theme. The R code sets a starting point — the viewer is yours to play with.

Combining Features

Things get more interesting when you layer group_by, split_by, filters, and aggregates:

aq <- airquality
aq$Month <- sprintf("%02d-%s", aq$Month, month.abb[aq$Month])

perspective(
  aq,
  plugin = "Heatmap",
  group_by = "Month",
  split_by = "Day",
  columns = "Temp",
  aggregates = list(Temp = "avg"),
  title = "NYC Daily Temperatures, 1973",
  theme = "Solarized Dark",
  height = "700px"
)

Months on one axis, days on the other, temperature as color intensity. It was a cold May in 1973. If you’re not American you might want Celsius — click the New Column button (the plus icon next to the column list) and paste this expression:

// Temp in Celsius
("Temp" - 32) * 5 / 9

Then swap the original Temp column for your new one in the columns panel.

The Sunburst

Hadley Wickham famously doesn’t like pie charts. But we’re not Hadley Wickham, so here’s a sunburst:

states <- as.data.frame(state.x77)
states$Region <- state.region
states$Division <- state.division

perspective(
  states,
  plugin = "Sunburst",
  group_by = c("Region", "Division"),
  columns = "Population",
  aggregates = list(Population = "sum"),
  title = "US Population by Region & Division",
  theme = "Vaporwave",
  height = "700px"
)

Inner ring is region, outer ring is division, slice size is population. Vaporwave theme because if you’re committing crimes against ggplot2 orthodoxy you might as well go all in.

Try switching to Treemap in the plugin dropdown for a rectangular version, and pick Pro Light or Pro Dark theme for a more classic look.

Shiny

The examples above work in Quarto, R Markdown, and the RStudio Viewer. But perspectiveR also has a Shiny proxy interface — you can stream rows into a viewer without re-rendering, listen for click and select events, and save or restore user configurations:

# Streaming stock market data with a rolling window
perspectiveR::run_example("shiny-basic")

# Editable CRUD table
perspectiveR::run_example("crud-table")

What’s Next