How do I programmatically generate a data.frame from many Shiny inputs? - r

Question

I'm trying to create a Shiny application that has ~100 user inputs, some of which may go untouched. I think that I could make a reactive dataframe within the app by writing input$user_input_1, input$user_input_2, etc. 100 times to call each input individually. Instead, I'm wondering if there is a way to programmatically generate that list of inputs using a concatenating function and then evaluate those inputs as if I had typed input$user_input_1. When I run the following code, however, I get a message telling me that Error: object 'input' not found.
My end goal is to be able to take these user inputs and use them in some behind-the-scenes transformations. If my current set up isn't the most efficient way to capture a large number of inputs, please let me know, but at the very least I'd like to understand what isn't working here.
I'd appreciate some help here. Thanks.
library(shiny)
library(tidyverse)
input_tbl <- tibble(inputs = c("x", "y", "z"))
ui <- fluidPage(
sidebarLayout(
sidebarPanel(
numericInput(
inputId = "number_x",
label = "Input X",
value = 30
),
numericInput(
inputId = "number_y",
label = "Input Y",
value = 60
),
numericInput(
inputId = "number_z",
label = "Input Z",
value = 90
)
),
mainPanel(
verbatimTextOutput("table")
)
)
)
server <- function(input, output) {
value_tbl <- eventReactive(
eventExpr = input$number_x,
valueExpr = {
tibble(
inputs = input_tbl %>% pull(),
values = input_tbl %>%
pull() %>%
modify(~ str_c("input$number_", .x)) %>%
map(~ parse(text = .x)) %>%
map(eval)
)
})
output$table <- renderPrint(value_tbl())
}
shinyApp(ui = ui, server = server)

Answer 1

If I got you right you could achieve your desired result with purrr::map_df in one line of code like so:
server <- function(input, output) {
value_tbl <- eventReactive(
eventExpr = input$number_x,
valueExpr = {
map_df(input_tbl$inputs, ~ data.frame(inputs = .x, values = input[[paste0("number_", .x)]]))
})
output$table <- renderPrint(value_tbl())
}

Answer 2

Turns out I was using the wrong evaluation function.
Instead of map(eval), I need to use map(eval.parent).