perspectiveR

R htmlwidgets binding for the FINOS Perspective library -- a high-performance WebAssembly-powered data visualization engine with interactive pivot tables and multiple chart types.

Installation

# Install from CRAN
install.packages("perspectiveR")

# Or install the development version from GitHub
# remotes::install_github("EydlinIlya/perspectiveR")
install.packages("perspectiveR", repos = c('https://eydlinilya.r-universe.dev', 'https://cloud.r-project.org'))

Quick Start

library(perspectiveR)

# Interactive data grid with full self-service UI
perspective(mtcars)

# Bar chart grouped by cylinder count
perspective(mtcars, group_by = "cyl", plugin = "Y Bar")

# Filtered scatter plot
perspective(iris,
  columns = c("Sepal.Length", "Sepal.Width", "Species"),
  filter = list(c("Species", "==", "setosa")),
  plugin = "Y Scatter"
)

Features

• Multiple visualization types: Datagrid, bar, line, area, scatter, heatmap, treemap, sunburst, and more
• Self-service interactive UI: Drag-and-drop columns, switch chart types, add filters/sorts/pivots, create computed expressions
• High performance: WebAssembly-powered compute engine runs entirely in the browser
• Shiny integration: Output/render bindings plus proxy interface for streaming data updates, indexed/keyed tables, rolling-window tables (limit), data export, table metadata queries, and state save/restore
• Filter operator: Combine multiple filters with filter_op = "and" or filter_op = "or"
• Arrow IPC support: Optional arrow package integration for efficient serialization of large datasets
• Works everywhere: RStudio Viewer, R Markdown, Quarto, and Shiny

Examples

Pivot tables with aggregation

Summarize data by grouping rows and splitting columns, with custom aggregates:

# Average horsepower by cylinders and transmission type
perspective(mtcars,
  columns    = c("hp", "mpg", "wt"),
  group_by   = "cyl",
  split_by   = "am",
  aggregates = list(hp = "avg", mpg = "avg", wt = "sum"),
  plugin     = "Datagrid"
)

Sorting

Sort by one or more columns, including absolute-value sorts for financial data:

# Sort airquality by Ozone descending, then Temp ascending
perspective(airquality,
  sort   = list(c("Ozone", "desc"), c("Temp", "asc")),
  filter = list(c("Ozone", "is not null"))
)

Multiple filters with filter_op

Combine filters with "and" (default) or "or":

# Cars that are either very efficient OR very powerful
perspective(mtcars,
  columns   = c("mpg", "hp", "cyl", "wt"),
  filter    = list(c("mpg", ">", 30), c("hp", ">", 200)),
  filter_op = "or",
  sort      = list(c("mpg", "desc"))
)

Computed expressions

Create derived columns using Perspective's expression language:

# Compute power-to-weight ratio and bin it
perspective(mtcars,
  expressions = c(
    '"hp_per_ton" = "hp" / ("wt" * 1000)',
    '"efficiency" = if ("mpg" > 25) { "high" } else { "low" }'
  ),
  columns  = c("hp", "wt", "mpg", "hp_per_ton", "efficiency"),
  group_by = "efficiency",
  plugin   = "Y Bar"
)

3D rendering with expressions

Render a 3D torus entirely in Perspective's expression engine — only the parametric angles come from R, while all geometry and shading math runs in WebAssembly (a hat-tip to the classic donut.c):

# 3D donut — geometry computed in Perspective's expression engine
torus <- expand.grid(
  u = seq(0, 2 * pi, length.out = 80),
  v = seq(0, 2 * pi, length.out = 80)
)

perspective(torus,
  expressions = c(
    '"x" = (2 + cos("v")) * cos("u")',
    '"y" = (2 + cos("v")) * sin("u")',
    '"shade" = cos("v") * sin("u") + sin("v")'
  ),
  columns  = c("x", "y", "shade"),
  plugin   = "XY Scatter",
  theme    = "Pro Dark",
  settings = FALSE,
  title    = "Donut"
)

Add streaming rotation in Shiny — see run_example("spinning-donut").

Chart types

Perspective supports a wide range of chart types:

# Line chart — time series
perspective(EuStockMarkets,
  plugin = "Y Line",
  title  = "European Stock Indices"
)

# Heatmap — correlation-style view
perspective(mtcars,
  group_by = "cyl",
  split_by = "gear",
  columns  = c("mpg", "hp", "wt"),
  plugin   = "Heatmap"
)

# Sunburst — hierarchical breakdown
perspective(mtcars,
  group_by = c("cyl", "gear", "am"),
  columns  = "mpg",
  plugin   = "Sunburst"
)

Themes

Nine built-in themes for light and dark environments:

# Dark mode with Monokai theme
perspective(iris,
  plugin = "Y Scatter",
  theme  = "Monokai",
  title  = "Iris Measurements"
)

# Available themes: "Pro Light" (default), "Pro Dark", "Monokai",
# "Solarized Light", "Solarized Dark", "Vaporwave", "Dracula",
# "Gruvbox", "Gruvbox Dark"

Editable datagrid

Enable inline editing for manual data entry or corrections:

perspective(data.frame(
    Task   = c("Review PR", "Write tests", "Deploy"),
    Status = c("Done", "In Progress", "Pending"),
    Owner  = c("Alice", "Bob", "Carol")
  ),
  editable = TRUE,
  title    = "Task Tracker"
)

Shiny Demos

Three interactive demos are bundled with the package:

library(perspectiveR)
run_example()             # list all available demos
run_example("shiny-basic")
run_example("crud-table")
run_example("spinning-donut")

• shiny-basic — Streaming stock market line chart (DAX, SMI, CAC, FTSE 1991-1998) with a 100-row sliding window, Arrow IPC toggle, computed columns, and named state save/restore.
• crud-table — Editable indexed CRUD table with click events, add/update/delete rows by key, downloadable CSV/JSON export, and an update activity log.
• spinning-donut — Animated 3D torus with all geometry computed in Perspective's expression engine and rotation driven by streaming data updates.

Shiny Usage

library(shiny)
library(perspectiveR)

ui <- fluidPage(
  selectInput("dataset", "Dataset:",
    choices = c("mtcars", "iris", "airquality")
  ),
  perspectiveOutput("viewer", height = "600px")
)

server <- function(input, output, session) {
  output$viewer <- renderPerspective({
    data <- switch(input$dataset,
      "mtcars" = mtcars,
      "iris" = iris,
      "airquality" = airquality
    )
    perspective(data)
  })
}

shinyApp(ui, server)

Proxy Functions

• psp_update(proxy, data) — append new rows (upserts when table has an index)
• psp_replace(proxy, data) — replace all data
• psp_clear(proxy) — clear all rows
• psp_restore(proxy, config) — apply a saved config
• psp_reset(proxy) — reset viewer to defaults
• psp_remove(proxy, keys) — remove rows by primary key (indexed tables)
• psp_export(proxy, format) — export data (JSON, CSV, columns, or Arrow); supports windowed export with start_row/end_row/start_col/end_col
• psp_save(proxy) — retrieve current viewer state
• psp_on_update(proxy, enable) — subscribe to data change events
• psp_schema(proxy) — get table schema (column names and types)
• psp_size(proxy) — get table row count
• psp_columns(proxy) — get table column names
• psp_validate_expressions(proxy, expressions) — validate expression strings

Related Projects

• Perspective — the upstream WebAssembly data visualization engine this package wraps
• GWalkR — interactive data exploration widget powered by Graphic Walker
• esquisse — drag-and-drop ggplot2 builder as a Shiny module

Building the JS Bundle

The pre-built JS bundle is included. To rebuild from source:

cd tools
npm install
npm run build
npm run copy-themes

License

Apache License 2.0

'Perspective' is a project of the OpenJS Foundation. perspectiveR is an independent community package and is not affiliated with or endorsed by the OpenJS Foundation.