teal.picks without teal (standalone Shiny)

Introduction

You can use picks_ui() and picks_srv() in a plain Shiny app: pass a reactive teal.data::teal_data() object to picks_srv() and combine results with merge_srv() when you need merged analysis data. This mirrors what tm_merge() does inside teal, without teal::init().

Run the shinyApp chunk interactively.

library(shiny)
library(teal.data)
library(teal.picks)

data <- teal_data()
data <- within(data, {
  ADSL <- data.frame(
    USUBJID = sprintf("S%03d", 1:8),
    AGE = sample(35:70, 8, replace = TRUE),
    stringsAsFactors = FALSE
  )
  ADLB <- data.frame(
    USUBJID = rep(sprintf("S%03d", 1:8), each = 3),
    PARAM = rep(c("ALT", "AST", "BILI"), 8),
    AVAL = round(rnorm(24, 42, 6), 1),
    stringsAsFactors = FALSE
  )
})

join_keys(data) <- join_keys(teal.data::join_key("ADSL", "ADLB", keys = "USUBJID"))

selector_default <- picks(
  datasets(choices = c("ADSL", "ADLB"), selected = "ADLB"),
  variables(
    choices = tidyselect::everything(),
    selected = c(1L, 2L),
    multiple = TRUE
  )
)

Minimal Shiny app

ui <- fluidPage(
  titlePanel("Standalone picks + merge"),
  fluidRow(
    column(
      width = 4,
      picks_ui("sel", picks = selector_default)
    ),
    column(
      width = 8,
      tags$h4("Mapped variables"),
      verbatimTextOutput("mapped"),
      tags$h4("Merge preview"),
      tableOutput("merged")
    )
  )
)

server <- function(input, output, session) {
  data_r <- reactive(data)

  selectors <- list(sel = picks_srv("sel", picks = selector_default, data = data_r))

  merged <- merge_srv(
    id = "merge",
    data = data_r,
    selectors = selectors,
    output_name = "anl",
    join_fun = "dplyr::left_join"
  )

  output$mapped <- renderPrint({
    yaml::as.yaml(merged$variables())
  })

  output$merged <- renderTable({
    merged$data()[["anl"]]
  })
}

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

Notes

merge_srv() expects selectors to be a named list of reactives (as returned by picks_srv() for each selector).
Define join_keys() on your teal_data before merging across datasets. One relationship between two datasets is enough: join_keys(join_key("ADSL", "ADLB", keys = "USUBJID")) is expanded by teal.data into a symmetric map so both names exist. Extra join_key("DS", "DS", …) self-keys are optional; they record primary-key / row grain (for example USUBJID + PARAM on long lab rows), which matters in full CDISC-style setups more than in this minimal example.
For bookmarking, picks_srv() stores resolved picks when enableBookmarking = "server" is used on shinyApp().
values() filters the column(s) content chosen in variables(). If multiple = TRUE variables are selected, values are derived from a combined representation of those columns—so do not pair PARAM-only level choices with a selection that also includes AVAL. Use values() with a single categorical column, or omit values() when taking several columns (as in this example).

NEST CoreDev Team

2026-05-19

Introduction

Minimal Shiny app

Notes