validate_inputs: Send input validation messages to output

In teal: Exploratory Web Apps for Analyzing Clinical Trials Data

Send input validation messages to output

Description

Captures messages from InputValidator objects and collates them into one message passed to validate.

Usage

validate_inputs(..., header = "Some inputs require attention")

Arguments

...	either any number of InputValidator objects or an optionally named, possibly nested list of InputValidator objects, see Details
header	(character(1)) generic validation message; set to NULL to omit

Details

shiny::validate is used to withhold rendering of an output element until certain conditions are met and to print a validation message in place of the output element. shinyvalidate::InputValidator allows to validate input elements and to display specific messages in their respective input widgets. validate_inputs provides a hybrid solution. Given an InputValidator object, messages corresponding to inputs that fail validation are extracted and placed in one validation message that is passed to a validate/need call. This way the input validator messages are repeated in the output.

The ... argument accepts any number of InputValidator objects or a nested list of such objects. If validators are passed directly, all their messages are printed together under one (optional) header message specified by header. If a list is passed, messages are grouped by validator. The list's names are used as headers for their respective message groups. If neither of the nested list elements is named, a header message is taken from header.

Value

Returns NULL if the final validation call passes and a shiny.silent.error if it fails.

See Also

shinyvalidate::InputValidator, shiny::validate

Examples

library(shiny)
library(shinyvalidate)

ui <- fluidPage(
  selectInput("method", "validation method", c("sequential", "combined", "grouped")),
  sidebarLayout(
    sidebarPanel(
      selectInput("letter", "select a letter:", c(letters[1:3], LETTERS[4:6])),
      selectInput("number", "select a number:", 1:6),
      br(),
      selectInput("color", "select a color:",
        c("black", "indianred2", "springgreen2", "cornflowerblue"),
        multiple = TRUE
      ),
      sliderInput("size", "select point size:",
        min = 0.1, max = 4, value = 0.25
      )
    ),
    mainPanel(plotOutput("plot"))
  )
)

server <- function(input, output) {
  # set up input validation
  iv <- InputValidator$new()
  iv$add_rule("letter", sv_in_set(LETTERS, "choose a capital letter"))
  iv$add_rule("number", function(x) {
    if (as.integer(x) %% 2L == 1L) "choose an even number"
  })
  iv$enable()
  # more input validation
  iv_par <- InputValidator$new()
  iv_par$add_rule("color", sv_required(message = "choose a color"))
  iv_par$add_rule("color", function(x) {
    if (length(x) > 1L) "choose only one color"
  })
  iv_par$add_rule(
    "size",
    sv_between(
      left = 0.5, right = 3,
      message_fmt = "choose a value between {left} and {right}"
    )
  )
  iv_par$enable()

  output$plot <- renderPlot({
    # validate output
    switch(input[["method"]],
      "sequential" = {
        validate_inputs(iv)
        validate_inputs(iv_par, header = "Set proper graphical parameters")
      },
      "combined" = validate_inputs(iv, iv_par),
      "grouped" = validate_inputs(list(
        "Some inputs require attention" = iv,
        "Set proper graphical parameters" = iv_par
      ))
    )

    plot(faithful$eruptions ~ faithful$waiting,
      las = 1, pch = 16,
      col = input[["color"]], cex = input[["size"]]
    )
  })
}

if (interactive()) {
  shinyApp(ui, server)
}

teal documentation built on May 29, 2024, 10:11 a.m.