Examples

Basic Usage

To create a data table, use reactable() on a data frame or matrix. The table will be sortable and paginated by default:

library(reactable)

reactable(iris)

Column definitions

Columns can be customized by providing a named list of column definitions created by colDef() to columns:

reactable(
  iris[1:5, ],
  columns = list(
    Sepal.Length = colDef(name = "Sepal Length"),
    Sepal.Width = colDef(name = "Sepal Width"),
    Species = colDef(align = "center")
  )
)

For convenience, you can also specify a default colDef() to use for all columns in defaultColDef:

reactable(
  iris[1:5, ],
  defaultColDef = colDef(
    header = function(value) gsub(".", " ", value, fixed = TRUE),
    cell = function(value) format(value, nsmall = 1),
    align = "center",
    minWidth = 70,
    headerStyle = list(background = "#f7f7f8")
  ),
  columns = list(
    Species = colDef(minWidth = 140)  # overrides the default
  ),
  bordered = TRUE,
  highlight = TRUE
)

Sorting

Tables are sortable by default. You can sort a column by clicking on its header, or sort multiple columns by holding the shift key while sorting.

Sorting toggles between ascending and descending order by default. To clear the sort, hold the shift key while sorting, and the sorting will additionally toggle between ascending, descending, and unsorted order.

Note: Ascending order means the lowest, first, or earliest values will appear first. Descending order means the largest, last, or latest values will appear first.

Default sorted columns

You can set the default sorted columns by providing a vector of column names to defaultSorted:

reactable(iris[48:52, ], defaultSorted = c("Species", "Petal.Length"))

You can also provide a named list to customize the default sort orders. Use "asc" for ascending order, or "desc" for descending order:

reactable(iris[48:52, ], defaultSorted = list(Species = "asc", Petal.Length = "desc"))

Default sort order

Columns are sorted in ascending order first by default. To change the default sort order for all columns in the table, set defaultSortOrder in reactable() to "asc" for ascending order, or "desc" for descending order.

To change the sort order of an individual column, set defaultSortOrder in its colDef() to "asc" or "desc". The default sort order of the column takes precedence over the table.

reactable(
  iris[48:52, ],
  defaultSortOrder = "desc",
  columns = list(
    Species = colDef(defaultSortOrder = "asc")
  ),
  defaultSorted = c("Species", "Petal.Length")
)

Sort missing values last

You can ignore missing values when sorting by setting sortNALast on a column:

reactable(
  data.frame(
    n = c(1, 2, 3, -Inf, Inf),
    x = c(2, 3, 1, NA, NaN),
    y = c("aa", "cc", "bb", NA, NA)
  ),
  defaultColDef = colDef(sortNALast = TRUE),
  defaultSorted = "x"
)

No sorting

You can disable sorting by setting sortable to FALSE on the table or column. When only some columns are sortable, it can help to indicate sortable columns using showSortable:

reactable(
  iris[1:5, ],
  sortable = FALSE,
  showSortable = TRUE,
  columns = list(
    Petal.Width = colDef(sortable = TRUE),
    Petal.Length = colDef(sortable = TRUE)
  )
)

Hide sort icons

You can hide sort icons by setting showSortIcon to FALSE. This is only recommended when you want to use a custom sort indicator.

reactable(iris[1:5, ], showSortIcon = FALSE)

Filtering

You can make columns filterable by setting filterable = TRUE in reactable():

data <- MASS::Cars93[1:20, c("Manufacturer", "Model", "Type", "AirBags", "Price")]

reactable(data, filterable = TRUE, minRows = 10)

To make specific columns filterable (or not), set filterable to TRUE or FALSE in colDef():

reactable(
  data,
  filterable = TRUE,
  columns = list(
    Price = colDef(filterable = FALSE)
  ),
  defaultPageSize = 5
)

Custom filtering

Column filtering can be customized using the filterMethod and filterInput arguments in colDef(). See the Custom Filtering guide for more details and examples.

This example shows basic usage of a custom filter method, changing filtering on the Manufacturer column to be case-sensitive rather than case-insensitive. (Try filtering for “bmw” and then “BMW”).

data <- MASS::Cars93[, c("Manufacturer", "Model", "Type", "Price")]

reactable(
  data,
  columns = list(
    Manufacturer = colDef(
      filterable = TRUE,
      # Filter by case-sensitive text match
      filterMethod = JS("function(rows, columnId, filterValue) {
        return rows.filter(function(row) {
          return row.values[columnId].indexOf(filterValue) !== -1
        })
      }")
    )
  ),
  defaultPageSize = 5
)

Searching

You can make the entire table searchable by setting searchable = TRUE in reactable():

data <- MASS::Cars93[1:20, c("Manufacturer", "Model", "Type", "AirBags", "Price")]

reactable(data, searchable = TRUE, minRows = 10)

Custom searching

The table search method can be customized using the searchMethod argument in reactable(). See the Custom Filtering guide for details and examples.

Pagination

You can change the default page size by configuring defaultPageSize:

reactable(iris[1:6, ], defaultPageSize = 4)

You can also set the minimum rows per page using minRows. This may be useful when rows don’t completely fill the page, or if the table has filtering:

reactable(iris[1:6, ], defaultPageSize = 4, minRows = 4, searchable = TRUE)

Page size options

You can show a dropdown of page sizes for users to choose from using showPageSizeOptions. The page size options can be customized through pageSizeOptions:

reactable(
  iris[1:12, ],
  showPageSizeOptions = TRUE,
  pageSizeOptions = c(4, 8, 12),
  defaultPageSize = 4
)

Alternative pagination types

You can use an alternative pagination type by setting paginationType to:

• "jump" to show a page jump
• "simple" to show previous/next buttons only

Page jump

reactable(iris[1:50, ], paginationType = "jump", defaultPageSize = 4)

Simple

reactable(iris[1:50, ], paginationType = "simple", defaultPageSize = 4)

Hide page info

You can hide page info by setting showPageInfo to FALSE:

reactable(iris[1:12, ], showPageInfo = FALSE, defaultPageSize = 4)

reactable(iris[1:12, ], showPageInfo = FALSE, showPageSizeOptions = TRUE, defaultPageSize = 4)

Always show pagination

By default, pagination is hidden if the table only has one page. To keep the pagination shown, set showPagination to TRUE. This is especially useful if you want to keep the page info showing the number of rows in the table.

reactable(iris[1:5, ], showPagination = TRUE)

No pagination

Tables are paginated by default, but you can disable pagination by setting pagination to FALSE:

reactable(iris[1:20, ], pagination = FALSE, highlight = TRUE, height = 250)

Tip: Disabling pagination is not recommended for large tables with many interactive elements (such as links, expand buttons, or selection checkboxes), as that can make it difficult for keyboard users to navigate the page.

Grouping and Aggregation

You can group rows in a table by specifying one or more columns in groupBy:

data <- MASS::Cars93[10:22, c("Manufacturer", "Model", "Type", "Price", "MPG.city")]

reactable(data, groupBy = "Manufacturer")

When rows are grouped, you can aggregate data in a column using an aggregate function:

data <- MASS::Cars93[14:38, c("Type", "Price", "MPG.city", "DriveTrain", "Man.trans.avail")]

reactable(
  data,
  groupBy = "Type",
  columns = list(
    Price = colDef(aggregate = "max"),
    MPG.city = colDef(aggregate = "mean", format = colFormat(digits = 1)),
    DriveTrain = colDef(aggregate = "unique"),
    Man.trans.avail = colDef(aggregate = "frequency")
  )
)

You can use one of the built-in aggregate functions:

colDef(aggregate = "sum")        # Sum of numbers
colDef(aggregate = "mean")       # Mean of numbers
colDef(aggregate = "max")        # Maximum of numbers
colDef(aggregate = "min")        # Minimum of numbers
colDef(aggregate = "median")     # Median of numbers
colDef(aggregate = "count")      # Count of values
colDef(aggregate = "unique")     # Comma-separated list of unique values
colDef(aggregate = "frequency")  # Comma-separated counts of unique values

Or a custom aggregate function in JavaScript:

colDef(
  aggregate = JS("
    function(values, rows) {
      // input:
      //  - values: an array of all values in the group
      //  - rows: an array of row data values for all rows in the group (optional)
      //
      // output:
      //  - an aggregated value, e.g. a comma-separated list
      return values.join(', ')
    }
  ")
)

Multiple groups

data <- data.frame(
  State = state.name,
  Region = state.region,
  Division = state.division,
  Area = state.area
)

reactable(
  data,
  groupBy = c("Region", "Division"),
  columns = list(
    Division = colDef(aggregate = "unique"),
    Area = colDef(aggregate = "sum", format = colFormat(separators = TRUE))
  ),
  bordered = TRUE
)

Custom aggregate function

Custom aggregate functions are useful when none of the built-in aggregate functions apply, or when you want to aggregate values from multiple columns. For example, when calculating aggregate averages or percentages.

Within a custom aggregate function, you can access the values in the column using the values argument, and the values in other columns using the rows argument:

columns = list(
  Price = colDef(
    aggregate = JS("function(values, rows) {
      values
      // [46.8, 27.6, 57]

      rows
      // [
      //   { "Model": "Dynasty", "Manufacturer": "Dodge", "Price": 46.8, "Units": 2 },
      //   { "Model": "Colt", "Manufacturer": "Dodge", "Price": 27.6, "Units": 5 },
      //   { "Model": "Caravan", "Manufacturer": "Dodge", "Price": 57, "Units": 5 }
      // ]
    }")
  )
)

Here’s an example that calculates an aggregate average price by dividing the the sum of two columns, Price and Units:

library(dplyr)

set.seed(10)

data <- sample_n(MASS::Cars93[23:40, ], 30, replace = TRUE) %>%
  mutate(Price = Price * 3, Units = sample(1:5, 30, replace = TRUE)) %>%
  mutate(Avg.Price = Price / Units) %>%
  select(Model, Manufacturer, Price, Units, Avg.Price)

reactable(
  data,
  groupBy = "Manufacturer",
  columns = list(
    Price = colDef(aggregate = "sum", format = colFormat(currency = "USD")),
    Units = colDef(aggregate = "sum"),
    Avg.Price = colDef(
      # Calculate the aggregate Avg.Price as `sum(Price) / sum(Units)`
      aggregate = JS("function(values, rows) {
        let totalPrice = 0
        let totalUnits = 0
        rows.forEach(function(row) {
          totalPrice += row['Price']
          totalUnits += row['Units']
        })
        return totalPrice / totalUnits
      }"),
      format = colFormat(currency = "USD")
    )
  )
)

Include sub rows in pagination

By default, sub rows are excluded from pagination and always shown on the same page when expanded. To include sub rows in pagination, you can set paginateSubRows to TRUE. This is recommended for grouped tables with a large number of rows where expanded rows may not all fit on one page.

data <- MASS::Cars93[, c("Manufacturer", "Model", "Type", "Price", "MPG.city")]

reactable(data, groupBy = "Type", paginateSubRows = TRUE)

Column Formatting

You can format data in a column by providing colFormat() options to the format argument in colDef().

The formatters for numbers, dates, times, and currencies are locale-sensitive and automatically adapt to language preferences of the user’s browser. This means, for example, that users will see dates formatted in their own timezone or numbers formatted in their own locale.

To use a specific locale for data formatting, provide a vector of BCP 47 language tags in the locales argument. See a list of common BCP 47 language tags for reference.

Note: Column formatters change how data is displayed without affecting the underlying data. Sorting, filtering, and grouping will still work on the original data.

data <- data.frame(
  price_USD = c(123456.56, 132, 5650.12),
  price_INR = c(350, 23208.552, 1773156.4),
  number_FR = c(123456.56, 132, 5650.12),
  temp = c(22, NA, 31),
  percent = c(0.9525556, 0.5, 0.112),
  date = as.Date(c("2019-01-02", "2019-03-15", "2019-09-22"))
)

reactable(data, columns = list(
  price_USD = colDef(format = colFormat(prefix = "$", separators = TRUE, digits = 2)),
  price_INR = colDef(format = colFormat(currency = "INR", separators = TRUE, locales = "hi-IN")),
  number_FR = colDef(format = colFormat(locales = "fr-FR")),
  temp = colDef(format = colFormat(suffix = " °C")),
  percent = colDef(format = colFormat(percent = TRUE, digits = 1)),
  date = colDef(format = colFormat(date = TRUE, locales = "en-GB"))
))

Date formatting

datetimes <- as.POSIXct(c("2019-01-02 3:22:15", "2019-03-15 09:15:55", "2019-09-22 14:20:00"),
                        tz = "America/New_York")
data <- data.frame(
  datetime = datetimes,
  date = datetimes,
  time = datetimes,
  time_24h = datetimes,
  datetime_pt_BR = datetimes
)

reactable(data, columns = list(
  datetime = colDef(format = colFormat(datetime = TRUE)),
  date = colDef(format = colFormat(date = TRUE)),
  time = colDef(format = colFormat(time = TRUE)),
  time_24h = colDef(format = colFormat(time = TRUE, hour12 = FALSE)),
  datetime_pt_BR = colDef(format = colFormat(datetime = TRUE, locales = "pt-BR"))
))

Currency formatting

data <- data.frame(
  USD = c(12.12, 2141.213, 0.42, 1.55, 34414),
  EUR = c(10.68, 1884.27, 0.37, 1.36, 30284.32),
  INR = c(861.07, 152122.48, 29.84, 110, 2444942.63),
  JPY = c(1280, 226144, 44.36, 164, 3634634.61),
  MAD = c(115.78, 20453.94, 4.01, 15, 328739.73)
)

reactable(data, columns = list(
  USD = colDef(
    format = colFormat(currency = "USD", separators = TRUE, locales = "en-US")
  ),
  EUR = colDef(
    format = colFormat(currency = "EUR", separators = TRUE, locales = "de-DE")
  ),
  INR = colDef(
    format = colFormat(currency = "INR", separators = TRUE, locales = "hi-IN")
  ),
  JPY = colDef(
    format = colFormat(currency = "JPY", separators = TRUE, locales = "ja-JP")
  ),
  MAD = colDef(
    format = colFormat(currency = "MAD", separators = TRUE, locales = "ar-MA")
  )
))

Formatting aggregated cells

Column formatters apply to both standard and aggregated cells by default. If you want to format aggregated cells separately, provide a named list of cell and aggregated options:

colDef(
  format = list(
    cell = colFormat(...),       # Standard cells
    aggregated = colFormat(...)  # Aggregated cells
  )
)

For example, only the aggregated States are formatted here:

data <- data.frame(
  States = state.name,
  Region = state.region,
  Area = state.area
)

reactable(
  data,
  groupBy = "Region",
  columns = list(
    States = colDef(
      aggregate = "count",
      format = list(
        aggregated = colFormat(suffix = " states")
      )
    ),
    Area = colDef(
      aggregate = "sum",
      format = colFormat(suffix = " mi²", separators = TRUE)
    )
  )
)

Displaying missing values

Missing values are ignored by formatters and shown as empty cells by default. You can customize their display text by setting na on a column:

reactable(
  data.frame(
    n = c(1, 2, NA, 4, 5),
    x = c(55, 27, NA, NaN, 19),
    y = c(1, NA, 0.25, 0.55, NA)
  ),
  columns = list(
    x = colDef(na = "–", format = colFormat(prefix = "$")),
    y = colDef(na = "NA", format = colFormat(percent = TRUE))
  )
)

Custom data formatting

If none of the built-in formatters apply to your data, you can use a custom cell renderer instead.

Custom Rendering

You can customize how data is displayed using an R or JavaScript function that returns custom content. R render functions support Shiny HTML tags (or htmltools) and HTML widgets, while JavaScript render functions allow for more dynamic behavior.

You can also render content as HTML using colDef(html = TRUE). Note that all raw HTML is escaped by default.

See Custom Rendering for details on how to use render functions, and the Demo Cookbook for even more examples of custom rendering.

Note: Custom rendering changes how data is displayed without affecting the underlying data. Sorting, filtering, and grouping will still work on the original data.

Cell rendering

R render function

data <- MASS::Cars93[1:5, c("Manufacturer", "Model", "Type", "AirBags", "Price")]

reactable(data, columns = list(
  Model = colDef(cell = function(value, index) {
    # Render as a link
    url <- sprintf("https://wikipedia.org/wiki/%s_%s", data[index, "Manufacturer"], value)
    htmltools::tags$a(href = url, target = "_blank", as.character(value))
  }),
  AirBags = colDef(cell = function(value) {
    # Render as an X mark or check mark
    if (value == "None") "\u274c No" else "\u2714\ufe0f Yes"
  }),
  Price = colDef(cell = function(value) {
    # Render as currency
    paste0("$", format(value * 1000, big.mark = ","))
  })
))

JavaScript render function

data <- MASS::Cars93[1:5, c("Manufacturer", "Model", "Type", "AirBags", "Price")]

reactable(data, columns = list(
  Model = colDef(html = TRUE, cell = JS('
    function(cellInfo) {
      // Render as a link
      const url = `https://wikipedia.org/wiki/${cellInfo.row["Manufacturer"]}_${cellInfo.value}`
      return `<a href="${url}" target="_blank">${cellInfo.value}</a>`
    }
  ')),
  AirBags = colDef(cell = JS("
    function(cellInfo) {
      // Render as an X mark or check mark
      return cellInfo.value === 'None' ? '\u274c No' : '\u2714\ufe0f Yes'
    }
  ")),
  Price = colDef(cell = JS("
    function(cellInfo) {
      // Render as currency
      return '$' + (cellInfo.value * 1000).toLocaleString()
    }
  "))
))

Embedding HTML widgets

library(dplyr)
library(sparkline)

data <- chickwts %>%
  group_by(feed) %>%
  summarise(weight = list(weight)) %>%
  mutate(boxplot = NA, sparkline = NA)

reactable(data, columns = list(
  weight = colDef(cell = function(values) {
    sparkline(values, type = "bar", chartRangeMin = 0, chartRangeMax = max(chickwts$weight))
  }),
  boxplot = colDef(cell = function(value, index) {
    sparkline(data$weight[[index]], type = "box")
  }),
  sparkline = colDef(cell = function(value, index) {
    sparkline(data$weight[[index]])
  })
))

Grouped cell rendering

data <- MASS::Cars93[10:22, c("Manufacturer", "Model", "Type", "Price", "MPG.city")]

reactable(
  data,
  groupBy = c("Manufacturer", "Type"),
  columns = list(
    Manufacturer = colDef(
      # Render grouped cells without the row count
      grouped = JS("function(cellInfo) {
        return cellInfo.value
      }")
    ),
    Type = colDef(
      # Render grouped cells with the row count, only if there are multiple sub rows
      grouped = JS("function(cellInfo) {
        if (cellInfo.subRows.length > 1) {
          return cellInfo.value + ' (' + cellInfo.subRows.length + ')'
        }
        return cellInfo.value
      }")
    )
  )
)

Aggregated cell rendering

library(dplyr)

set.seed(10)

data <- sample_n(tail(MASS::Cars93, 9), 30, replace = TRUE) %>%
  select(Manufacturer, Model, Type, Sales = Price)

reactable(
  data,
  groupBy = "Manufacturer",
  searchable = TRUE,
  columns = list(
    Model = colDef(aggregate = "unique"),
    Type = colDef(
      # Render aggregated value as a comma-separated list of unique values
      aggregated = JS("function(cellInfo) {
        const values = cellInfo.subRows.map(function(row) { return row['Type'] })
        const unique = values.reduce(function(obj, v) { obj[v] = true; return obj }, {})
        return Object.keys(unique).join(', ')
      }")
    ),
    Sales = colDef(
      aggregate = "sum",
      # Render aggregated cell as currency
      aggregated = JS("function(cellInfo) {
        return '$' + cellInfo.value.toFixed(2)
      }")
    )
  )
)

Header rendering

This example requires reactable v0.3.0 or above.

library(htmltools)

reactable(
  iris[1:5, ],
  defaultColDef = colDef(header = function(value) {
    units <- div(style = "color: #737373", "cm")
    div(title = value, value, units)
  }),
  columns = list(
    Petal.Width = colDef(
      name = "Petal Width",
      html = TRUE,
      align = "left",
      header = JS('function(column) {
        return column.name + `<div style="color: #737373">cm</div>`
      }')
    ),
    Species = colDef(header = function(value) {
      tags$a(href = "https://wikipedia.org/wiki/List_of_Iris_species", value)
    })
  )
)

Custom metadata

New in v0.4.0

You can pass arbitrary data from R to JavaScript render functions using the meta argument in reactable().

meta should be a named list of values that can also be JS() expressions or functions. Custom metadata can be accessed from JavaScript using the state.meta property, and updated using updateReactable() in Shiny or Reactable.setMeta() in the JavaScript API.

Use custom metadata to:

• Simplify JavaScript render functions that need access to data outside of the table
• Dynamically change how data is formatted without rerendering the table
• Share JavaScript code or data between different render functions

library(htmltools)

data <- MASS::Cars93[1:6, c("Manufacturer", "Model", "Type", "Price", "MPG.city")]

exchange_rates <- list(
  USD = 1,
  CAD = 1.30,
  JPY = 137.56
)

tbl <- reactable(
  data,
  columns = list(
    Price = colDef(
      cell = JS("function(cellInfo, state) {
        const { currency, exchangeRates } = state.meta
        const converted = cellInfo.value * exchangeRates[currency]
        return converted.toLocaleString(undefined, { style: 'currency', currency: currency })
      }")
    )
  ),
  meta = list(
    currency = "USD",
    exchangeRates = exchange_rates
  ),
  elementId = "cars-currency-table"
)

browsable(
  tagList(
    tags$label(
      "Currency",
      tags$select(
        onchange = "Reactable.setMeta('cars-currency-table', { currency: event.target.value })",
        lapply(names(exchange_rates), tags$option)
      )
    ),

    tags$hr("aria-hidden" = "true"),

    tbl
  )
)

Currency USD CAD JPY

---

Footers

You can add column footers using the footer argument in colDef().

footer can either be custom content to render (e.g., a character string or HTML tag), or a custom render function. See Custom Rendering to learn more about using custom render functions.

R render function

library(dplyr)
library(htmltools)

data <- MASS::Cars93[18:47, ] %>%
  select(Manufacturer, Model, Type, Sales = Price)

reactable(
  data,
  defaultPageSize = 5,
  columns = list(
    Manufacturer = colDef(footer = "Total"),
    Sales = colDef(footer = function(values) sprintf("$%.2f", sum(values)))
  ),
  defaultColDef = colDef(footerStyle = list(fontWeight = "bold"))
)

JavaScript render function

This example requires reactable v0.3.0 or above.

reactable(
  data,
  searchable = TRUE,
  defaultPageSize = 5,
  minRows = 5,
  columns = list(
    Manufacturer = colDef(footer = "Total"),
    Sales = colDef(
      footer = JS("function(column, state) {
        let total = 0
        state.sortedData.forEach(function(row) {
          total += row[column.id]
        })
        return '$' + total.toFixed(2)
      }")
    )
  ),
  defaultColDef = colDef(footerStyle = list(fontWeight = "bold"))
)

Embedding HTML widgets

library(sparkline)

reactable(
  iris[1:20, ],
  defaultPageSize = 5,
  bordered = TRUE,
  defaultColDef = colDef(footer = function(values) {
    if (!is.numeric(values)) return()
    sparkline(values, type = "box", width = 100, height = 30)
  })
)

Expandable Row Details

This example requires reactable v0.3.0 or above.

You can make rows expandable with additional content through details, which takes an R or JavaScript render function. See Custom Rendering for details on how to use render functions.

reactable(iris[1:5, ], details = function(index) {
  htmltools::div(
    "Details for row: ", index,
    htmltools::tags$pre(paste(capture.output(iris[index, ]), collapse = "\n"))
  )
})

The details column can be customized by providing a colDef() instead. This can be used to add a column name, render HTML content, or change the column width:

reactable(iris[1:5, ], details = colDef(
  name = "More",
  details = JS("function(rowInfo) {
    return `Details for row: ${rowInfo.index}` +
      `<pre>${JSON.stringify(rowInfo.values, null, 2)}</pre>`
  }"),
  html = TRUE,
  width = 60
))

Nested tables

With R render functions, you can render HTML tags, HTML widgets, and even nested tables:

data <- unique(CO2[, c("Plant", "Type")])

reactable(data, details = function(index) {
  plant_data <- CO2[CO2$Plant == data$Plant[index], ]
  htmltools::div(style = "padding: 1rem",
    reactable(plant_data, outlined = TRUE)
  )
})

Conditional row details

R render functions support conditional rendering. If a render function returns NULL, the row won’t be expandable:

reactable(iris[1:5, ], details = function(index) {
  if (index %in% c(3, 5)) {
    reactable(data.frame(x = c(1, 2, 3), y = c("a", "b", "c")), fullWidth = FALSE)
  }
})

Multiple row details

This example requires reactable v0.3.0 or above.

You can add details to individual columns, and even show multiple details for a row:

reactable(iris[1:5, ],
  details = function(index) {
    if (index %in% c(3, 5)) {
      reactable(data.frame(x = c(1, 2, 3), y = c("a", "b", "c")), fullWidth = FALSE)
    }
  },
  columns = list(
    Petal.Length = colDef(details = function(index) {
      paste("Petal.Length: ", iris[index, "Petal.Length"])
    }),
    Sepal.Length = colDef(format = colFormat(digits = 1), details = JS("
      function(rowInfo) {
        return 'Sepal.Length: ' + rowInfo.values['Sepal.Length']
      }
    "))
  )
)

Default expanded rows

You can expand all rows by default by setting defaultExpanded to TRUE:

reactable(
  iris[1:12, ],
  defaultPageSize = 4,
  details = function(index) paste("Details for row:", index),
  defaultExpanded = TRUE
)

Conditional Styling

You can conditionally style a table using functions that return inline styles or CSS classes. Just like with custom rendering, style functions can either be in R or JavaScript.

See Conditional Styling for details on how to use style functions, and the Demo Cookbook for even more examples of conditional styling.

Cell styling

R style function

reactable(sleep[1:6, ], columns = list(
  extra = colDef(
    style = function(value) {
      if (value > 0) {
        color <- "#008000"
      } else if (value < 0) {
        color <- "#e00000"
      } else {
        color <- "#777"
      }
      list(color = color, fontWeight = "bold")
    }
  )
))

JavaScript style function

This example requires reactable v0.3.0 or above.

reactable(sleep[1:6, ], columns = list(
  extra = colDef(
    style = JS("function(rowInfo) {
      const value = rowInfo.values['extra']
      let color
      if (value > 0) {
        color = '#008000'
      } else if (value < 0) {
        color = '#e00000'
      } else {
        color = '#777'
      }
      return { color: color, fontWeight: 'bold' }
    }")
  )
))

Row styling

R style function

reactable(sleep[1:6, ], 
  rowStyle = function(index) {
    if (sleep[index, "extra"] < -1) {
      list(background = "rgba(0, 0, 0, 0.05)")
    }
  },
  rowClass = function(index) {
    if (sleep[index, "extra"] < -1) {
      "bold"
    }
  }
)

.bold {
  font-weight: bold;
}

JavaScript style function

This example requires reactable v0.3.0 or above.

reactable(sleep[1:6, ],
  rowStyle = JS("function(rowInfo) {
    if (rowInfo.values['extra'] < -1) {
      return { background: 'rgba(0, 0, 0, 0.05)' }
    }
  }"),
  rowClass = JS("function(rowInfo) {
    if (rowInfo.values['extra'] < -1) {
      return 'bold'
    }
  }")
)

Custom metadata

New in v0.4.0

You can pass arbitrary data from R to JavaScript style functions using the meta argument in reactable().

meta should be a named list of values that can also be JS() expressions or functions. Custom metadata can be accessed from JavaScript using the state.meta property, and updated using updateReactable() in Shiny or Reactable.setMeta() in the JavaScript API.

Use custom metadata to:

• Simplify JavaScript style functions that need access to data outside of the table
• Dynamically change how data is styled without rerendering the table
• Share JavaScript code or data between different style functions

library(htmltools)

data <- MASS::Cars93[1:6, c("Manufacturer", "Model", "Type", "Price", "MPG.city")]

mpg_normalized <- (data$MPG.city - min(data$MPG.city)) / (max(data$MPG.city) - min(data$MPG.city))
mpg_colors <-  rgb(colorRamp(c("#ffe4cc", "#ff9f1a"))(mpg_normalized), maxColorValue = 255)

tbl <- reactable(
  data,
  columns = list(
    MPG.city = colDef(
      style = JS("function(rowInfo, column, state) {
        const { showColors, mpgColors } = state.meta
        if (showColors) {
          return { backgroundColor: mpgColors[rowInfo.index] }
        }
      }")
    )
  ),
  meta = list(
    mpgColors = mpg_colors,
    showColors = TRUE
  ),
  elementId = "cars-colors-table"
)

browsable(
  tagList(
    tags$label(
      tags$input(
        type = "checkbox",
        checked = NA,
        onclick = "Reactable.setMeta('cars-colors-table', function(prevMeta) {
          return { showColors: !prevMeta.showColors }
        })"
      ),
      "Show color scale"
    ),

    tags$hr("aria-hidden" = "true"),

    tbl
  )
)

Show color scale

---

Table Styling

You can customize table styling using several options, which can all be combined:

Highlight rows on hover

reactable(iris[1:5, ], highlight = TRUE)

Bordered

reactable(iris[1:5, ], bordered = TRUE)

Borderless

reactable(iris[1:5, ], borderless = TRUE)

Outlined

reactable(iris[1:5, ], outlined = TRUE)

Striped

reactable(iris[1:5, ], striped = TRUE)

Bordered + striped + highlighting

reactable(iris[1:5, ], bordered = TRUE, striped = TRUE, highlight = TRUE)

Outlined + borderless

reactable(iris[1:5, ], outlined = TRUE, borderless = TRUE)

Compact

reactable(iris[1:5, ], compact = TRUE)

No text wrapping

Long text is wrapped by default, but you can force text to fit on a single line by setting wrap to FALSE:

data <- aggregate(. ~ Species, iris, toString)

reactable(
  data,
  wrap = FALSE,
  resizable = TRUE,
  bordered = TRUE,
  columns = list(Petal.Length = colDef(name = "Petal Length (cm)", minWidth = 50))
)

Fixed height + sticky header/footer

You can make tables scrollable by setting a fixed height or width. Headers and footers are sticky by default, so they stay in place when scrolling.

Scrollable tables are automatically made focusable when navigating using a keyboard to ensure that they’re always accessible for keyboard users.

reactable(
  iris[1:20, ],
  height = 270,
  striped = TRUE,
  defaultColDef = colDef(
    footer = function(values, name) {
      htmltools::div(name, style = list(fontWeight = 600))
    }
  )
)

Column widths

By default, columns have a minimum width of 100px and stretch to fill the table. You can control the width of a column using the following arguments in colDef():

• minWidth - minimum width of the column in pixels (defaults to 100)
• maxWidth - maximum width of the column in pixels
• width - fixed width of the column in pixels (overrides minWidth and maxWidth)

When columns stretch, minWidth also controls the ratio at which columns grow. For example, if a table consists of 3 columns having minWidth = 100 each, the columns will stretch at a ratio of 100:100:100. Each column will take up 1/3 of the table’s width and not shrink below 100px.

Another example: if a table consists of three columns having minimum widths of 200px, 100px, and 100px, the columns will take up 50%, 25%, and 25% of the table’s width respectively:

reactable(
  MASS::Cars93[1:6, c("Make", "Type", "Weight")],
  columns = list(
    Make = colDef(minWidth = 200),   # 50% width, 200px minimum
    Type = colDef(minWidth = 100),   # 25% width, 100px minimum
    Weight = colDef(minWidth = 100)  # 25% width, 100px minimum
  ),
  bordered = TRUE
)

No full width

Tables are full width by default, but you can shrink the table to fit its contents by setting fullWidth to FALSE:

reactable(
  MASS::Cars93[1:5, 1:5],
  fullWidth = FALSE,
  bordered = TRUE,
  defaultColDef = colDef(minWidth = 120)
)

You can also set a maximum or fixed width on the table:

reactable(
  MASS::Cars93[1:5, 1:5],
  bordered = TRUE,
  defaultColDef = colDef(minWidth = 120),
  # Set a maximum width on the table:
  style = list(maxWidth = 650),
  # Or a fixed width:
  width = 650
)

Vertical alignment

You can change the vertical alignment of cell content using the vAlign or headerVAlign arguments in colDef() and colGroup(). vAlign controls the alignment of data cells, while headerVAlign controls the alignment of header cells. Possible options are "top" (the default), "center", and "bottom".

library(dplyr)
library(htmltools)

data <- starwars[1:6, ] %>%
  select(character = name, height, mass, gender, homeworld, species)

reactable(
  data,
  columns = list(
    character = colDef(
      name = "Character / Species",
      # Show species under character names
      cell = function(value, index) {
        species <- data$species[index]
        species <- if (!is.na(species)) species else "Unknown"
        div(
          div(style = list(fontWeight = 600), value),
          div(style = list(fontSize = "0.75rem"), species)
        )
      }
    ),
    species = colDef(show = FALSE)
  ),
  # Vertically center cells and bottom-align headers
  defaultColDef = colDef(vAlign = "center", headerVAlign = "bottom"),
  bordered = TRUE
)

Custom CSS

For more control over styling, you can add custom class names to the table and apply your own CSS:

reactable(
  iris[1:18, ],
  defaultPageSize = 6,
  borderless = TRUE,
  class = "my-tbl",
  defaultColDef = colDef(headerClass = "my-header"),
  columns = list(
    Sepal.Width = colDef(class = "my-col"),
    Petal.Width = colDef(class = "my-col")
  ),
  rowClass = "my-row"
)

In R Markdown documents, you can embed CSS using a css language chunk:

```{css, echo=FALSE}
.my-tbl {
  border: 1px solid rgba(0, 0, 0, 0.1);
}

.my-header {
  border-width: 1px;
}

.my-col {
  border-right: 1px solid rgba(0, 0, 0, 0.05);
}

.my-row:hover {
  background-color: #f5f8ff;
}
```

The examples here embed CSS for demonstration, but it’s sometimes better to add CSS through an external style sheet. To learn more about adding custom CSS through an external style sheet:

• Using custom CSS in R Markdown documents
• Using custom CSS in Shiny apps

Note: If you inspect a table’s HTML, you might find CSS classes like .rt-table on different elements of the table. These CSS classes are undocumented and subject to change, so we recommend adding your own custom class names, or using themes to customize parts of the table that aren’t covered by the custom class names.

Theming

Themes provide a powerful way to customize table styling that can be reused across tables. You can either set theme variables to change the default styles (e.g., row stripe color), or add your own custom CSS to specific elements of the table.

To apply a theme, provide a reactableTheme() to theme:

reactable(
  iris[1:30, ],
  searchable = TRUE,
  striped = TRUE,
  highlight = TRUE,
  bordered = TRUE,
  theme = reactableTheme(
    borderColor = "#dfe2e5",
    stripedColor = "#f6f8fa",
    highlightColor = "#f0f5f9",
    cellPadding = "8px 12px",
    style = list(fontFamily = "-apple-system, BlinkMacSystemFont, Segoe UI, Helvetica, Arial, sans-serif"),
    searchInputStyle = list(width = "100%")
  )
)

Global theme

To set the default theme for all tables, use the global reactable.theme option:

options(reactable.theme = reactableTheme(
  color = "hsl(233, 9%, 87%)",
  backgroundColor = "hsl(233, 9%, 19%)",
  borderColor = "hsl(233, 9%, 22%)",
  stripedColor = "hsl(233, 12%, 22%)",
  highlightColor = "hsl(233, 12%, 24%)",
  inputStyle = list(backgroundColor = "hsl(233, 9%, 25%)"),
  selectStyle = list(backgroundColor = "hsl(233, 9%, 25%)"),
  pageButtonHoverStyle = list(backgroundColor = "hsl(233, 9%, 25%)"),
  pageButtonActiveStyle = list(backgroundColor = "hsl(233, 9%, 28%)")
))

reactable(
  iris[1:30, ],
  filterable = TRUE,
  showPageSizeOptions = TRUE,
  striped = TRUE,
  highlight = TRUE,
  details = function(index) paste("Details for row", index)
)

Nested selectors

You can use nested CSS selectors in theme styles to target the current element, using & as the selector, or other child elements (just like in Sass). This is useful for adding pseudo-classes like &:hover, or adding styles in a certain context like .outer-container &.

For example, to highlight headers when sorting:

reactable(
  iris[1:5, ],
  columns = list(Sepal.Length = colDef(sortable = FALSE)),
  showSortable = TRUE,
  theme = reactableTheme(
    headerStyle = list(
      "&:hover[aria-sort]" = list(background = "hsl(0, 0%, 96%)"),
      "&[aria-sort='ascending'], &[aria-sort='descending']" = list(background = "hsl(0, 0%, 96%)"),
      borderColor = "#555"
    )
  )
)

Or to apply a dark theme when a parent element has a certain class, like .dark:

theme <- reactableTheme(
  style = list(".dark &" = list(color = "#fff", background = "#282a36")),
  cellStyle = list(".dark &" = list(borderColor = "rgba(255, 255, 255, 0.15)")),
  headerStyle = list(".dark &" = list(borderColor = "rgba(255, 255, 255, 0.15)")),
  paginationStyle = list(".dark &" = list(borderColor = "rgba(255, 255, 255, 0.15)")),
  rowHighlightStyle = list(".dark &" = list(background = "rgba(255, 255, 255, 0.04)")),
  pageButtonHoverStyle = list(".dark &" = list(background = "rgba(255, 255, 255, 0.08)")),
  pageButtonActiveStyle = list(".dark &" = list(background = "rgba(255, 255, 255, 0.1)"))
)

tbl <- reactable(iris[1:12, ], highlight = TRUE, defaultPageSize = 6, theme = theme)

# Simple theme toggle button
tags$button(onclick = "document.querySelector('.themeable-tbl').classList.toggle('dark')",
            "Toggle light/dark")

# Start with the dark theme enabled
div(class = "themeable-tbl dark", tbl)

Toggle light/dark

Dynamic theming

Themes can also be functions that return a reactableTheme() for context-specific styling.

For example, to style tables in RStudio R Notebooks only when a dark editor theme is active:

options(reactable.theme = function() {
  theme <- reactableTheme(
    color = "hsl(233, 9%, 85%)",
    backgroundColor = "hsl(233, 9%, 19%)",
    borderColor = "hsl(233, 9%, 22%)",
    stripedColor = "hsl(233, 12%, 22%)",
    highlightColor = "hsl(233, 12%, 24%)",
    inputStyle = list(backgroundColor = "hsl(233, 9%, 25%)"),
    selectStyle = list(backgroundColor = "hsl(233, 9%, 25%)"),
    pageButtonHoverStyle = list(backgroundColor = "hsl(233, 9%, 25%)"),
    pageButtonActiveStyle = list(backgroundColor = "hsl(233, 9%, 28%)")
  )

  if (isTRUE(getOption("rstudio.notebook.executing"))) {
    if (requireNamespace("rstudioapi", quietly = TRUE) && rstudioapi::getThemeInfo()$dark) {
      return(theme)
    }
  }
})

Column Groups

You can create column groups by passing a list of colGroup() definitions to columnGroups:

reactable(
  iris[1:5, ],
  columns = list(
    Sepal.Length = colDef(name = "Length"),
    Sepal.Width = colDef(name = "Width"),
    Petal.Length = colDef(name = "Length"),
    Petal.Width = colDef(name = "Width")
  ),
  columnGroups = list(
    colGroup(name = "Sepal", columns = c("Sepal.Length", "Sepal.Width")),
    colGroup(name = "Petal", columns = c("Petal.Length", "Petal.Width"))
  )
)

Column Resizing

You can make columns resizable by setting resizable to TRUE:

reactable(MASS::Cars93[1:5, ], resizable = TRUE, wrap = FALSE, bordered = TRUE)

Sticky Columns

You can make columns sticky when scrolling horizontally using the sticky argument in colDef() or colGroup(). Set sticky to either "left" or "right" to make the column stick to the left or right side.

reactable(
  MASS::Cars93[1:5, ],
  columns = list(
    Manufacturer = colDef(
      sticky = "left",
      # Add a right border style to visually distinguish the sticky column
      style = list(borderRight = "1px solid #eee"),
      headerStyle = list(borderRight = "1px solid #eee")
    ),
    Make = colDef(
      sticky = "right",
      # Add a left border style to visually distinguish the sticky column
      style = list(borderLeft = "1px solid #eee"),
      headerStyle = list(borderLeft = "1px solid #eee")
    )
  ),
  defaultColDef = colDef(minWidth = 150)
)

Multiple sticky columns

# Background style to visually distinguish sticky columns
sticky_style <- list(backgroundColor = "#f7f7f7")

reactable(
  MASS::Cars93[1:5, ],
  columns = list(
    Manufacturer = colDef(
      sticky = "left",
      style = sticky_style,
      headerStyle = sticky_style
    ),
    Model = colDef(
      sticky = "left",
      style = sticky_style,
      headerStyle = sticky_style
    ),
    Type = colDef(
      sticky = "left",
      style = sticky_style,
      headerStyle = sticky_style
    )
  ),
  resizable = TRUE,
  wrap = FALSE,
  bordered = TRUE
)

Sticky column groups

If a column group is sticky, all columns in the group will automatically be made sticky.

reactable(
  MASS::Cars93[1:5, ],
  columnGroups = list(
    colGroup("Make", columns = c("Manufacturer", "Model"), sticky = "left"),
    colGroup("Price", columns = c("Min.Price", "Price", "Max.Price"), sticky = "left")
  ),
  defaultColDef = colDef(footer = "Footer"),
  resizable = TRUE,
  wrap = FALSE,
  bordered = TRUE
)

Row Names and Row Headers

Row names

Row names are shown by default if present. You can customize the row names column using ".rownames" as the column name:

reactable(
  USPersonalExpenditure,
  columns = list(
    .rownames = colDef(name = "Category", sortable = TRUE)
  )
)

If row names haven’t been set explicitly, you can force them to show by setting rownames to TRUE:

reactable(iris[1:5, ], rownames = TRUE)

Row headers

You can mark up cells in a column as row headers by setting rowHeader to TRUE in colDef().

Use this to help users navigate the table using assistive technologies. When cells are marked up as row headers, assistive technologies will read them aloud while navigating through cells in the table.

Cells in the row names column are automatically marked up as row headers.

data <- MASS::Cars93[1:5, c("Make", "Type", "Price", "MPG.city", "AirBags")]

reactable(
  data,
  columns = list(
    Make = colDef(rowHeader = TRUE, style = list(fontWeight = 600))
  ),
  bordered = TRUE
)

Cell Click Actions

You can add cell click actions using the onClick argument, which accepts the following values:

• "expand" to expand the row
• "select" to select the row
• A JavaScript function for a custom action, e.g., sending the click event to Shiny

Expand on click

reactable(
  iris[48:52, ],
  groupBy = "Species",
  details = function(index) paste("Details for row:", index),
  onClick = "expand",
  # Give rows a pointer cursor to indicate that they're clickable
  rowStyle = list(cursor = "pointer")
)

Select on click

reactable(iris[1:5, ], selection = "multiple", onClick = "select")

Custom action

This example requires reactable v0.3.0 or above.

This example uses a custom click action to create custom “show details” action buttons in each row of the table:

data <- cbind(
  MASS::Cars93[1:5, c("Manufacturer", "Model", "Type", "Price")],
  details = NA
)

reactable(
  data,
  columns = list(
    # Render a "show details" button in the last column of the table.
    # This button won't do anything by itself, but will trigger the custom
    # click action on the column.
    details = colDef(
      name = "",
      sortable = FALSE,
      cell = function() htmltools::tags$button("Show details")
    )
  ),
  onClick = JS("function(rowInfo, column) {
    // Only handle click events on the 'details' column
    if (column.id !== 'details') {
      return
    }

    // Display an alert dialog with details for the row
    window.alert('Details for row ' + rowInfo.index + ':\\n' + JSON.stringify(rowInfo.values, null, 2))

    // Send the click event to Shiny, which will be available in input$show_details
    // Note that the row index starts at 0 in JavaScript, so we add 1
    if (window.Shiny) {
      Shiny.setInputValue('show_details', { index: rowInfo.index + 1 }, { priority: 'event' })
    }
  }")
)

Warning: Custom click actions are currently not accessible to keyboard users, and are generally not recommended. If they must be used, ensure that they can be triggered by a keyboard through other means, such as a button in the example above.

Language Options

You can customize the language in the table by providing a set of reactableLang() options to language:

reactable(
  iris[1:30, ],
  searchable = TRUE,
  paginationType = "simple",
  language = reactableLang(
    searchPlaceholder = "Search...",
    noData = "No entries found",
    pageInfo = "{rowStart} to {rowEnd} of {rows} entries",
    pagePrevious = "\u276e",
    pageNext = "\u276f",

    # Accessible labels for assistive technologies such as screen readers.
    # These are already set by default, but don't forget to update them when
    # changing visible text.
    pagePreviousLabel = "Previous page",
    pageNextLabel = "Next page"
  )
)

Global language options

To set the default language strings for all tables, use the global reactable.language option:

options(reactable.language = reactableLang(
  pageSizeOptions = "\u663e\u793a {rows}",
  pageInfo = "{rowStart} \u81f3 {rowEnd} \u9879\u7ed3\u679c,\u5171 {rows} \u9879",
  pagePrevious = "\u4e0a\u9875",
  pageNext = "\u4e0b\u9875"
))

reactable(iris[1:12, ], defaultPageSize = 4, showPageSizeOptions = TRUE)

Shiny

To use reactable in Shiny apps, use renderReactable() and reactableOutput():

library(shiny)
library(reactable)

ui <- fluidPage(
  titlePanel("reactable example"),
  reactableOutput("table")
)

server <- function(input, output, session) {
  output$table <- renderReactable({
    reactable(iris)
  })
}

shinyApp(ui, server)

Row selection

You can enable row selection by setting selection to "single" for single selection, or "multiple" for multiple selection.

To get the selected rows in Shiny, use getReactableState(). The selected rows are given as a vector of row indices (e.g. c(1, 6, 4)) or NULL if no rows are selected.

library(shiny)
library(reactable)

ui <- fluidPage(
  titlePanel("row selection example"),
  reactableOutput("table"),
  verbatimTextOutput("selected")
)

server <- function(input, output, session) {
  selected <- reactive(getReactableState("table", "selected"))

  output$table <- renderReactable({
    reactable(iris, selection = "multiple", onClick = "select")
  })

  output$selected <- renderPrint({
    print(selected())
  })

  observe({
    print(iris[selected(), ])
  })
}

shinyApp(ui, server)

Default selected rows

You can preselect rows by specifying a vector of row indices in defaultSelected:

reactable(iris[1:4, ], selection = "multiple", defaultSelected = c(1, 3))

Style selected rows

You can style selected rows using rowSelectedStyle in reactableTheme():

reactable(
  iris[1:4, ],
  selection = "multiple",
  defaultSelected = c(1, 3),
  borderless = TRUE,
  onClick = "select",
  theme = reactableTheme(
    rowSelectedStyle = list(backgroundColor = "#eee", boxShadow = "inset 2px 0 0 0 #ffa62d")
  )
)

Or using a rowStyle or rowClass JavaScript function:

reactable(
  MASS::Cars93[10:22, c("Manufacturer", "Model", "Type", "Price", "MPG.city")],
  groupBy = "Manufacturer",
  selection = "multiple",
  defaultSelected = c(1, 2),
  borderless = TRUE,
  onClick = "select",
  rowStyle = JS("function(rowInfo) {
    if (rowInfo && rowInfo.selected) {
      return { backgroundColor: '#eee', boxShadow: 'inset 2px 0 0 0 #ffa62d' }
    }
  }")
)

Customize the selection column

You can customize the selection column using ".selection" as the column name:

reactable(
  MASS::Cars93[1:4, ],
  columns = list(
    .selection = colDef(
      width = 80,
      sticky = "left",
      style = list(cursor = "pointer"),
      headerStyle = list(cursor = "pointer")
    )
  ),
  selection = "multiple",
  onClick = "select",
  resizable = TRUE,
  wrap = FALSE,
  bordered = TRUE
)

Update a reactable instance

You can update the selected rows, expanded rows, current page, or data using updateReactable():

library(shiny)
library(reactable)

data <- MASS::Cars93[, 1:7]

ui <- fluidPage(
  actionButton("select_btn", "Select rows"),
  actionButton("clear_btn", "Clear selection"),
  actionButton("expand_btn", "Expand rows"),
  actionButton("collapse_btn", "Collapse rows"),
  actionButton("page_btn", "Change page"),
  selectInput("filter_type", "Filter type", unique(data$Type), multiple = TRUE),
  reactableOutput("table")
)

server <- function(input, output) {
  output$table <- renderReactable({
    reactable(
      data,
      filterable = TRUE,
      searchable = TRUE,
      selection = "multiple",
      details = function(index) paste("Details for row:", index)
    )
  })

  observeEvent(input$select_btn, {
    # Select rows
    updateReactable("table", selected = c(1, 3, 5))
  })

  observeEvent(input$clear_btn, {
    # Clear row selection
    updateReactable("table", selected = NA)
  })

  observeEvent(input$expand_btn, {
    # Expand all rows
    updateReactable("table", expanded = TRUE)
  })

  observeEvent(input$collapse_btn, {
    # Collapse all rows
    updateReactable("table", expanded = FALSE)
  })

  observeEvent(input$page_btn, {
    # Change current page
    updateReactable("table", page = 3)
  })

  observe({
    # Filter data
    filtered <- if (length(input$filter_type) > 0) {
      data[data$Type %in% input$filter_type, ]
    } else {
      data
    }
    updateReactable("table", data = filtered)
  })
}

shinyApp(ui, server)

Get the state of a reactable instance

You can get the current state of a table using getReactableState().

By default, getReactableState() returns a named list with the following values:

• page: the current page
• pageSize: the page size
• pages: the number of pages
• sorted: the sorted columns - a named list of columns with values of "asc" for ascending order or "desc" for descending order, or NULL if no columns are sorted
• selected: the selected rows - a numeric vector of row indices, or NULL if no rows are selected

To only watch for changes on a specific value, you can use the optional name argument, like getReactableState(outputId, "selected").

library(shiny)
library(reactable)
library(htmltools)

ui <- fluidPage(
  actionButton("prev_page_btn", "Previous page"),
  actionButton("next_page_btn", "Next page"),
  reactableOutput("table"),
  verbatimTextOutput("table_state"),
  uiOutput("selected_row_details")
)

server <- function(input, output) {
  output$table <- renderReactable({
    reactable(
      MASS::Cars93[, 1:5],
      showPageSizeOptions = TRUE,
      selection = "multiple",
      onClick = "select"
    )
  })

  output$table_state <- renderPrint({
    state <- req(getReactableState("table"))
    print(state)
  })

  observeEvent(input$prev_page_btn, {
    # Change to the previous page
    page <- getReactableState("table", "page")
    if (page > 1) {
      updateReactable("table", page = page - 1)
    }
  })

  observeEvent(input$next_page_btn, {
    # Change to the next page
    state <- getReactableState("table")
    if (state$page < state$pages) {
      updateReactable("table", page = state$page + 1)
    }
  })
  
  output$selected_row_details <- renderUI({
    selected <- getReactableState("table", "selected")
    req(selected)
    details <- MASS::Cars93[selected, -c(1:5)]
    tagList(
      h2("Selected row details"),
      tags$pre(
        paste(capture.output(print(details, width = 1200)), collapse = "\n")
      )
    )
  })
}

shinyApp(ui, server)

Cross-Widget Interactions

You can link selection and filtering with other HTML widgets in an R Markdown document or Shiny app using Crosstalk. To get started, install the crosstalk package and wrap your data frame in a crosstalk::SharedData object:

install.packages("crosstalk")

library(crosstalk)

data <- SharedData$new(iris)

Then, pass the shared data to reactable() and any other Crosstalk-compatible HTML widget or filter input:

reactable(data)

filter_slider("sepal_length", "Sepal Length", data, ~Sepal.Length)

For more examples and a list of Crosstalk-compatible widgets, check out Using Crosstalk in the Crosstalk package documentation.

Filtering

Tables can be filtered by widgets that support Crosstalk’s filtering API, such as Crosstalk’s filter_checkbox(), filter_slider(), and filter_select() inputs:

library(crosstalk)

cars <- MASS::Cars93[1:20, c("Manufacturer", "Model", "Type", "Price")]
data <- SharedData$new(cars)

shiny::fluidRow(
  shiny::column(
    4,
    filter_checkbox("type", "Type", data, ~Type),
    filter_slider("price", "Price", data, ~Price, width = "100%"),
    filter_select("mfr", "Manufacturer", data, ~Manufacturer)
  ),
  shiny::column(
    8,
    reactable(data, minRows = 10)
  )
)

Type

Compact

Large

Midsize

Small

Sporty

Van

Price

Manufacturer

Note: This example uses shiny::fluidRow() and shiny::column() to create a Bootstrap grid layout, which works with all Bootstrap versions. crosstalk::bscols() can also create a grid, but is only compatible with Bootstrap 3. If you’re not using Bootstrap, here’s an alternative way to create a responsive grid using CSS grid.

Example: grid layout using CSS grid

library(crosstalk)

cars <- MASS::Cars93[1:20, c("Manufacturer", "Model", "Type", "Price")]
data <- SharedData$new(cars)

div(
  style = "display: grid; grid-template-columns: repeat(auto-fit, minmax(200px, 1fr)); gap: 0.75rem;",
  div(
    filter_checkbox("type", "Type", data, ~Type),
    filter_slider("price", "Price", data, ~Price, width = "100%"),
    filter_select("mfr", "Manufacturer", data, ~Manufacturer)
  ),
  div(
    style = "grid-column: span 2;",
    reactable(data, minRows = 10)
  )
)

Linked selection

Table selection state can be linked with other widgets that support Crosstalk’s linked selection (or linked brushing) API.

In this example, you can select rows to highlight points on the map, or select areas on the map to highlight rows in the table.

library(crosstalk)
library(leaflet)
library(dplyr)

# A SpatialPointsDataFrame for the map.
# Set a group name to share data points with the table.
brew_sp <- SharedData$new(breweries91, group = "breweries")

# A regular data frame (without coordinates) for the table.
# Use the same group name as the map data.
brew_data <- as_tibble(breweries91) %>%
  select(brewery, address, village, founded) %>%
  SharedData$new(group = "breweries")

map <- leaflet(brew_sp) %>%
  addTiles() %>%
  addMarkers()

tbl <- reactable(
  brew_data,
  selection = "multiple",
  onClick = "select",
  rowStyle = list(cursor = "pointer"),
  minRows = 10
)

htmltools::browsable(
  htmltools::tagList(map, tbl)
)

JavaScript API

You can use the JavaScript API to create custom interactive controls for your table without the use of Shiny, or add cross-widget interactions beyond what Crosstalk provides.

See the JavaScript API guide for details on how to use the JavaScript API in R Markdown documents or Shiny apps, and the full API reference.

CSV download button

Reactable.downloadDataCSV() downloads table data to a CSV file, including any filters that have been applied. See the JavaScript API guide for more details on usage, including how to customize the field separator, decimal separator, included columns, and more.

library(htmltools)
library(fontawesome)

data <- MASS::Cars93[1:15, c("Manufacturer", "Model", "Type", "Price")]

htmltools::browsable(
  tagList(
    tags$button(
      tagList(fontawesome::fa("download"), "Download as CSV"),
      onclick = "Reactable.downloadDataCSV('cars-download-table', 'cars.csv')"
    ),

    reactable(
      data,
      searchable = TRUE,
      defaultPageSize = 5,
      elementId = "cars-download-table"
    )
  )
)

Download as CSV

CSV download button in Shiny

While you can create download buttons in Shiny using shiny::downloadButton(), you may still prefer to use the JavaScript API, as Reactable.downloadDataCSV() automatically applies any client-side filtering that has been done to the table.

library(shiny)
library(reactable)
library(htmltools)

csvDownloadButton <- function(id, filename = "data.csv", label = "Download as CSV") {
  tags$button(
    tagList(icon("download"), label),
    onclick = sprintf("Reactable.downloadDataCSV('%s', '%s')", id, filename)
  )
}

ui <- fluidPage(
  csvDownloadButton("cars_table", filename = "cars.csv"),
  reactableOutput("cars_table")
)

server <- function(input, output) {
  output$cars_table <- renderReactable({
    reactable(
      MASS::Cars93[, c("Manufacturer", "Model", "Type", "Price")],
      searchable = TRUE
    )
  })
}

shinyApp(ui, server)

Custom column filter

library(htmltools)

data <- MASS::Cars93[1:15, c("Manufacturer", "Model", "Type", "Price")]

htmltools::browsable(
  tagList(
    div(
      div(tags$label("Filter Type", `for` = "cars-type-filter")),
      tags$select(
        id = "cars-type-filter",
        onchange = "Reactable.setFilter('cars-filter-table', 'Type', this.value)",
        tags$option("All", value = ""),
        lapply(unique(data$Type), tags$option)
      )
    ),
    
    tags$hr("aria-hidden" = "true"),

    reactable(data, defaultPageSize = 5, elementId = "cars-filter-table")
  )
)

Filter Type

All Small Midsize Compact Large Sporty

---

Custom search input

library(htmltools)

data <- MASS::Cars93[1:15, c("Manufacturer", "Model", "Type", "Price")]

htmltools::browsable(
  tagList(
    div(
      style = "margin-bottom: 0.75rem",
      tags$input(
        type = "text",
        placeholder = "Search for cars...",
        style = "padding: 0.25rem 0.5rem; width: 100%",
        oninput = "Reactable.setSearch('cars-search-table', this.value)"
      )
    ),

    reactable(data, defaultPageSize = 5, elementId = "cars-search-table")
  )
)

Column grouping select

library(dplyr)
library(htmltools)

set.seed(10)

data <- sample_n(tail(MASS::Cars93, 9), 30, replace = TRUE) %>%
  select(Manufacturer, Model, Type, Sales = Price)

htmltools::browsable(
  tagList(
    div(tags$label("Group by", `for` = "cars-grouping-select")),
    tags$select(
      id = "cars-grouping-select",
      onchange = "Reactable.setGroupBy('cars-grouping-table', this.value ? [this.value] : [])",
      tags$option("None", value = ""),
      lapply(c("Manufacturer", "Model", "Type"), tags$option)
    ),

    tags$hr("aria-hidden" = "true"),

    reactable(
      data,
      columns = list(
        Manufacturer = colDef(aggregate = "unique"),
        Model = colDef(aggregate = "unique"),
        Type = colDef(aggregate = "unique"),
        Sales = colDef(aggregate = "sum", format = colFormat(currency = "USD"))
      ),
      defaultPageSize = 5,
      minRows = 5,
      elementId = "cars-grouping-table"
    )
  )
)

Group by

None Manufacturer Model Type

---

Row expansion toggle button

library(htmltools)

data <- MASS::Cars93[1:5, c("Manufacturer", "Model", "Type", "Price")]

htmltools::browsable(
  tagList(
    tags$button(
      "Expand/collapse all",
      onclick = "Reactable.toggleAllRowsExpanded('cars-expansion-table')"
    ),

    reactable(
      data,
      groupBy = "Manufacturer",
      defaultPageSize = 5,
      elementId = "cars-expansion-table"
    )
  )
)

Expand/collapse all

Column visibility toggle button

New in v0.4.0

library(htmltools)

data <- MASS::Cars93[1:5, c("Manufacturer", "Model", "Type", "Price",
                            "Passengers", "DriveTrain", "Cylinders", "EngineSize")]

htmltools::browsable(
  tagList(
    tags$button(
      "Show/hide more columns",
      onclick = "Reactable.setHiddenColumns('cars-vis-table', prevColumns => {
        return prevColumns.length === 0 ? ['Passengers', 'DriveTrain', 'Cylinders', 'EngineSize'] : []
      })"
    ),
    reactable(
      data,
      columns = list(
        Passengers = colDef(show = FALSE),
        DriveTrain = colDef(show = FALSE),
        Cylinders = colDef(show = FALSE),
        EngineSize = colDef(show = FALSE)
      ),
      elementId = "cars-vis-table"
    )
  )
)

Show/hide more columns