Use dplyr conditional filter in reactive function in Shiny

Question

I'm using dplyr in a Shiny app reactive function. I have an interactive widget in the UI that users can use to select status and data will be shown based on selected status. However I also want to give the option to show everything, which is a status that I have called 'All'.

I know how to achieve this without dplyr:

library(tibble)
library(shiny)

# Test Data -----
test_data <- tribble(
  ~id, ~status,
  1, 'active',
  2, 'inactive',
  3, 'test'
)

# Ui ----
ui = fluidPage(
  selectInput("status", "Choose a status:",
              choices = list("All","active","inactive","test")
  ),
  tableOutput('result')
)

# Server ----
server = function(input, output) {

  table <- reactive({
    if(input$status != 'All')
    {
      test_data <- test_data[test_data$status == input$status,]
    } else {
      test_data
    }
  })

  output$result <- renderTable(table(), rownames = FALSE)

}

# App ---
shinyApp(ui = ui, server = server)

However, I would like to be consistent with the rest of my code and use dplyr. Is there a way to do something like the below?

library(tibble)
library(dplyr)
library(shiny)

# Test Data -----
test_data <- tribble(
  ~id, ~status,
  1, 'active',
  2, 'inactive',
  3, 'test'
)

# Ui ----
ui = fluidPage(
  selectInput("status", "Choose a status:",
              choices = list("All","active","inactive","test")
  ),
  tableOutput('result')
)

# Server ----
server = function(input, output) {

  table <- reactive({
    test_data %>%
      filter(ifelse(input$status != 'All', status == input$status, status == ***pass***))
  })

  output$result <- renderTable(table(), rownames = FALSE)

}

# App ---
shinyApp(ui = ui, server = server)

In other words, is there a way to use ifelse inside filter function to not filtering under a certain condition?

Thanks!

Konrad · Answer 1 · 2019-09-23T08:29:17.310

6

If your desire is to conditionally apply filter depending on the external value, you could attempt to use syntax:

`TRUE`

table <- reactive({
  test_data %>%
    filter(if(input$status != 'All')  (status == input$status) else TRUE)
})

Passing TRUE as condition would then not filter any rows.

`{}`

Alternative approach uses {} and returns untouched data set: else ..

table <- reactive({
  test_data %>%
  {if(input$status != 'All') filter(status == input$status) else .}
})

`%<>%`

Finally you could consider breaking the pipe and using %<>% operator available in magrittr package:

if(input$status != 'All') test_data %<>% filter(status == input$status)

edited Sep 23 '19 at 08:29

answered Mar 21 '18 at 16:19

Konrad

17,740
16
106
167

1

Hi @Konrad, nice solution! I was unaware of the fact that you filter all rows by passing a a single `TRUE` in the `filter` statement. – Florian Mar 21 '18 at 20:59
1

Excellent answer! – Jochem Sep 23 '19 at 08:28

score 3 · Accepted Answer · edited Aug 24 '17 at 09:50

3

I do not think that is possible. You could do as follows:

if(input$status!="All")
  test_data %>% filter(status %in% input$status)
else
  test_data

So you only filter if the input is not "All". If you have a very long dplyr chain and you do not want to break it into pieces, you could also do:

if(input$status=="All")
  my_filter=unique(test_data$status)
else
  my_filter = input$status

test_data %>%
  filter(status %in% my_filter) %>%
  ..... %>%
  .....

Hope this helps.

edited Aug 24 '17 at 09:50

Giacomo

1,796
1
24
35

answered Aug 24 '17 at 07:49

Florian

24,425
4
49
80

1

Thanks. The first example you posted needed an else otherwise it was print an empty table with option 'All' - I have edited your post to reflect that. I like the second example you posted. – Giacomo Aug 24 '17 at 09:49
@Florian Why, maybe I'm missing something but the way I see it is that the only challenge is for dplyr's filter function to evaluate to `TRUE` depending `input$status == "All"` and return all rows. – Konrad Mar 21 '18 at 16:29

score 1 · Answer 3 · answered Dec 12 '19 at 08:47

1

How about:

table <- reactive({
test_data %>%
filter(status == input$status | input$status == 'All')
})

answered Dec 12 '19 at 08:47

Algomas

61
7

Use dplyr conditional filter in reactive function in Shiny

3 Answers3

TRUE

{}

%<>%

`TRUE`

`{}`

`%<>%`