Got 'argument 1 is empty error' with selectInput in shinydashboard - r

Question

I have a shiny app like below using diamonds dataset as an example. I'm using two selectInput as data filter. The first one select a variable. The second one shows the values, depending on the variable selected in the first selectInput. After selection, click the actionButton to trigger the filter. Without any variable and value selected, I want it output the whole dataset.
What I found is after I clear the two selectInput and then click the actionButton, I got error: argument 1 is empty. I do not understand why that happens. Spent hours but unable to find the solution. Do anyone know how I can fix it? Thanks a lot!
library(shiny)
library(shinydashboard)
library(dplyr)
library(rlang)
data(diamonds)
df = diamonds[1:1000,]
subset_data = function(data,
var=NULL,
value=NULL){
if (!is.null(var)) {
if(!is.null(value)) {
data = data %>% filter(!!sym(var) == value)
}
}
return(data)
}
ui <- dashboardPage(
dashboardHeader(title = "example"),
dashboardSidebar(),
dashboardBody(
fluidRow(
selectizeInput(inputId = 'var',
label='Select variable',
choices = c('cut', 'color'),
multiple=FALSE,
options = list(
maxItems = 1,
placeholder = '',
onInitialize = I("function() { this.setValue(''); }"))),
uiOutput("valueUI"),
actionButton('go', 'apply filter'),
tableOutput('table')
)
)
)
server <- function(input, output, session) {
output$valueUI = renderUI({
if (input$var == '') {
vals = ''
}
if (input$var == 'cut') {
vals = c('Premium', 'Good', 'Very Good', 'Fair')
}
if (input$var == 'color'){
vals = c('E', 'J', 'I', 'H')
}
selectizeInput(inputId = 'value',
label='Select values',
choices = c('',vals),
multiple=FALSE,
options = list(
maxItems = 1,
placeholder = '',
onInitialize = I("function() { this.setValue(''); }")))
})
dat = reactive({
input$go
isolate( subset_data(data=df, var=input$var, value=input$value) )
})
output$table <- renderTable({
dat()
})
}
shinyApp(ui, server)

Answer 1

if input$var not given, it will be handled with '' as you implemented.
However, in same time it will also give parameter var value in subset_data function as '' not NULL, so it will be not recognized with !is.null(var) and trigger filter for data .
You can see these explanation if you change subset_data like below code;
subset_data = function(data,
var=NULL,
value=NULL){
print('var:')
print(var) # var is given as ''
print("value:")
print(value)
if (!is.null(var)) {
if(!is.null(value)) {
print('hi')
data = data %>% filter(!!sym(var) == value)
print('hi2') # this will not printed, since filter makes error.
}
}
return(data)
}
so in this case, you can fix them with adding just 1 line on subset_data to check if var == ''.
if( var == '' ) return(data)
Regards.