How to use buttons in a Reactable widget for navigation in a Shiny application

[This article was first published on R | Discindo, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

How to render and use buttons in tables is a relatively common task faced by {shiny} developers. Commonly, buttons in tables are used for getting more details about a row of data, for opening modal panels for user input, for displaying charts, and of course, for navigation. There are many ways to achieve each of these tasks and many tools to do it with. We can create an HTML table containing buttons from scratch (e.g. here), or we can use table widgets like the ones provided by the {reactable}, {DT} and many other packages. In this post, we’ll introduce a very straightforward way to implement buttons and use them for navigation (and alters) with the {reactable} package. Most of what you’ll see below is also documented in or derived from the reactable docs.

Putting buttons in a {reactable}

To render buttons inside a reactable widget we need to make an HTML button tag for each row of the table, and add this column of buttons as column in the data frame we wish to render. We also need to let reactable know that the columns containing our buttons should be interpreted as HTML.

colA <- 1:3
colB <- c("A", "B", "C")
makeButton <- function(label) {
as.character(htmltools::tags$div(htmltools::tags$button(paste(label, "button"))))
}
colC <- as.character(lapply(c("A", "B", "C"), makeButton))
dataFrame <- data.frame(A = colA, B = colB, C = colC)
reactable::reactable(dataFrame,
columns = list(C = reactable::colDef(sortable = FALSE,
html = TRUE)))

Making the buttons alive

For the buttons to respond to our clicks, we need to add an onclick or ondbclick event to each button. To do this, we add the onclick argument to the tags$button tag in our makeButton function. Below, we create an event to each button that will show the id of the button clicked:

labels <- LETTERS[1:3]
buttonAlert <- function(label) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste(label, "button"),
onClick = sprintf("alert('Button value is: %s')", label)
)))
}
alertButtonsHTML <- as.character(lapply(labels, buttonAlert))
tableWithButtons <- reactable::reactable(
data.frame(
Names = labels,
Alert = alertButtonsHTML
),
columns = list(
Alert = reactable::colDef(sortable = FALSE,
html = TRUE)
)
)
tableWithButtons

We also have the option to use reactable’s onClick argument, however this requires writing a JavaScript function to figure our which row was clicked and then trigger the appropriate action. In some complex scenarios, such an approach might be necessary, but for this use case, its simple to attach an onclick event to each button when we create the HTML in R.

Using the buttons to toggle tabs

So far, we rendered a {reactable} with buttons from R, and added events to the buttons such that they trigger an alert in the browser showing the button id. This got us part of the way towards our goal to use the button clicks for navigation in {shiny} To be able to change tabs in a shiny application, we need to send the button id from the browser to the server. We do this with the Shiny.setInputValue JS function as described in the shiny docs on JS to R communication.

Next, we’ll define two columns of buttons, one that trigger an alert like before, and another that sets in input in {shiny}’s session that can be accessed from the server. To show this example we need a full {shiny} application with a tabset panel that we can toggle by clicking the {reactable} buttons. Toggling between tabs happens by the usual updateTabsetPanel in an observer listening to the input updated by the buttons.

tab_names <- LETTERS[1:3]
buttonSetInput <- function(nav_id, nav_value) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste("Go to tab", nav_value),
onClick = sprintf(
"Shiny.setInputValue('%s', '%s', {priority: 'event'})",
nav_id,
nav_value
)
)))
}
buttonWithAlert <- function(nav_id, nav_value) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste("Alert", nav_value),
onClick = sprintf("alert('Nav id is: %s, and nav value is: %s')", nav_id, nav_value)
)))
}
alertButtonsHTML <- as.character(lapply(tab_names, buttonWithAlert, nav_id = "myNav"))
setInputButtonsHTML <- as.character(lapply(tab_names, buttonSetInput, nav_id = "myNav"))
tableWithButtons <- reactable::reactable(
data.frame(
Names = LETTERS[1:3],
Alert = alertButtonsHTML,
SetInput = setInputButtonsHTML
),
columns = list(
Alert = reactable::colDef(sortable = FALSE,html = TRUE),
SetInput = reactable::colDef(sortable = FALSE, html = TRUE)
)
)
ui <- bslib::page_fluid(
theme = bslib::bs_theme(version = 5, bootswatch = "flatly"),
title = "Reactable buttons navigation",
shiny::titlePanel("App navigation with buttons in reactable widget"),
shiny::div(
class = "row",
shiny::div(
class = "col-4",
reactableOutput("myTab"),
helpText("Input from reactable button is:"),
verbatimTextOutput("inputFromTableButton"),
bslib::navs_tab_card(
id = "myTabs",
bslib::nav(title = "Tab A", value = "A", "Contents of tab A"),
bslib::nav(title = "Tab B", value = "B", "Contents of tab B"),
bslib::nav(title = "Tab C", value = "C", "Contents of tab C")
)
)
)
)
server <- function(input, output, session) {
output$myTab <- renderReactable({
tableWithButtons
})
output$inputFromTableButton <- renderPrint({
input$myNav
})
shiny::observeEvent(input$myNav, {
shiny::updateTabsetPanel(session = session,
inputId = "myTabs",
selected = input$myNav)
})
}
shiny::shinyApp(ui, server)

But this is not really navigation

Yes, so far we’ve only toggled tabs on the same page. For this to be ‘navigation’, when we click the buttons in the table we should ‘jump’ from one to another tab. The mechanism to do something like this is essentially the same, except we should render the table within one of our tabs, and then have buttons that lead to other tabs. The example app below is a minor tweak of what we saw earlier to achieve exactly that.

tab_names <- LETTERS[1:3]
buttonSetInput <- function(nav_id, nav_value) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste("Go to tab", nav_value),
onClick = sprintf(
"Shiny.setInputValue('%s', '%s', {priority: 'event'})",
nav_id,
nav_value
)
)))
}
buttonWithAlert <- function(nav_id, nav_value) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste("Alert", nav_value),
onClick = sprintf("alert('Nav id is: %s, and nav value is: %s')", nav_id, nav_value)
)))
}
alertButtonsHTML <- as.character(lapply(tab_names, buttonWithAlert, nav_id = "myNav"))
setInputButtonsHTML <- as.character(lapply(tab_names, buttonSetInput, nav_id = "myNav"))
tableWithButtons <- reactable::reactable(
data.frame(
Names = LETTERS[1:3],
Alert = alertButtonsHTML,
SetInput = setInputButtonsHTML
),
columns = list(
Alert = reactable::colDef(sortable = FALSE,html = TRUE),
SetInput = reactable::colDef(sortable = FALSE, html = TRUE)
)
)
ui <- bslib::page_fluid(
theme = bslib::bs_theme(version = 5, bootswatch = "flatly"),
title = "Reactable buttons navigation",
shiny::titlePanel("App navigation with buttons in reactable widget"),
shiny::div(
class = "row",
shiny::div(
class = "col-4",
helpText("Input from reactable button is:"),
verbatimTextOutput("inputFromTableButton"),
bslib::navs_tab_card(
id = "myTabs",
bslib::nav(title = "Tab X", value = "X", reactableOutput("myTab")),
bslib::nav(title = "Tab A", value = "A", "Contents of tab A"),
bslib::nav(title = "Tab B", value = "B", "Contents of tab B"),
bslib::nav(title = "Tab C", value = "C", "Contents of tab C")
)
)
)
)
server <- function(input, output, session) {
output$myTab <- renderReactable({
tableWithButtons
})
output$inputFromTableButton <- renderPrint({
input$myNav
})
shiny::observeEvent(input$myNav, {
shiny::updateTabsetPanel(session = session,
inputId = "myTabs",
selected = input$myNav)
})
}
shiny::shinyApp(ui, server)

What if we used modules?

In a realistic app, we would probably use shiny modules. We could easily construct a scenario where the reactable is rendered through a module, or where each tab is a separate module. The pattern we described above works in this case too, except, with modules, we have keep track of the namespaces. Modules are isolated, so an input value updated in module X will not be known to module Y. We have to implement some way of communication between modules.

Below is one of several ways to send a message from one module to another. In this, case we pass a value from a ‘submodule’ to the calling module. This is a common case, for example, we have the main app server with top-level navigation, that is calling modules for ‘pages’ within the application. The key is to return the input updated by the submodule as a reactive that can be observed in the top-level module.

tab_names <- LETTERS[1:3]
buttonSetInput <- function(nav_id, nav_value) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste("Go to tab", nav_value),
onClick = sprintf(
"Shiny.setInputValue('%s', '%s', {priority: 'event'})",
nav_id,
nav_value
)
)))
}
buttonWithAlert <- function(nav_id, nav_value) {
as.character(htmltools::tags$div(htmltools::tags$button(
paste("Alert", nav_value),
onClick = sprintf("alert('Nav id is: %s, and nav value is: %s')", nav_id, nav_value)
)))
}
rTabUI <- function(id) {
ns <- NS(id)
tagList(
reactableOutput(ns("myTab"))
)
}
rTabServer <- function(id) {
moduleServer(
id,
function(input, output, session) {
ns <- session$ns
alertButtonsHTML <- as.character(lapply(tab_names, buttonWithAlert, nav_id = ns("myNav")))
setInputButtonsHTML <- as.character(lapply(tab_names, buttonSetInput, nav_id = ns("myNav")))
tableWithButtons <- reactable::reactable(
data.frame(
Names = LETTERS[1:3],
Alert = alertButtonsHTML,
SetInput = setInputButtonsHTML
),
columns = list(
Alert = reactable::colDef(sortable = FALSE,html = TRUE),
SetInput = reactable::colDef(sortable = FALSE, html = TRUE)
)
)
output$myTab <- renderReactable({
tableWithButtons
})
return(list(
getButtonValue = shiny::reactive(input$myNav)
))
}
)
}
ui <- bslib::page_fluid(
theme = bslib::bs_theme(version = 5, bootswatch = "flatly"),
title = "Reactable buttons navigation",
shiny::titlePanel("App navigation with buttons in reactable widget"),
shiny::div(
class = "row",
shiny::div(
class = "col-4",
helpText("Input from reactable button is:"),
verbatimTextOutput("inputFromTableButton"),
bslib::navs_tab_card(
id = "myTabs",
bslib::nav(title = "Tab X", value = "X", rTabUI("rtab") ),
bslib::nav(title = "Tab A", value = "A", "Contents of tab A"),
bslib::nav(title = "Tab B", value = "B", "Contents of tab B"),
bslib::nav(title = "Tab C", value = "C", "Contents of tab C")
)
)
)
)
server <- function(input, output, session) {
rTabOut <- rTabServer("rtab")
output$inputFromTableButton <- renderPrint({
rTabOut$getButtonValue()
})
shiny::observeEvent(rTabOut$getButtonValue(), {
shiny::updateTabsetPanel(session = session,
inputId = "myTabs",
selected = rTabOut$getButtonValue())
})
}
shiny::shinyApp(ui, server)

Other ways to implement communication between modules include passing (reactive) values through session$userData (e.g. here) or updating values in an environment or R6 class passed to each module as an argument (e.g. here).

Summary

In this post we covered some examples of how one could implement navigation between ‘pages’ in a {shiny} application using buttons in a reactable. We saw how to generate some buttons, how to update input values on button click, how to listen to changes from the buttons, and how to pass the user selections between modules.

To leave a comment for the author, please follow the link and comment on their blog: R | Discindo.

R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)