How to use transformModule

Introduction

There are similar modules to transform data. Below are list of them.

• Binarize
• ETC
• Log
• Replace
• Round
• Split

and in this article, we call them as transformModule.

scissor’s transformModule transforms selected variables.

In this article, we’ll use scholarship dataset of datatoys

This article is based on 0.2.0 Version of scissor.

Declare module

scissor’s module assumes that used in the Shiny application.

and You can use snippet(type shinyapp) to build very basic shiny application.

library(shiny)

ui <- fluidPage(
 
)

server <- function(input, output, session) {
  
}

shinyApp(ui, server)

This application will show nothing.

So let’s add binarizeModule in ui

ui <- fluidPage(
  mod_binarizeModule_ui(
    id = 'binModule'
  )
)

also, binarizeModule in server.

server <- function(input, output, session) {
  mod_binarizeModule_server(
    id = 'binModule', 
    inputData = reactive(datatoys::scholarship)
  )
}

So final (which is very basic) code will like this.

library(shiny)

ui <- fluidPage(
  mod_binarizeModule_ui(
    id = 'binModule'
  )
)

server <- function(input, output, session) {
  mod_binarizeModule_server(
    id = 'binModule', 
    inputData = reactive(datatoys::scholarship) # remotes::install_github('statgarten/datatoys')
  )
}

shinyApp(ui, server) # run application

You should notice 3 things.

1. Both id in ui and server should be same.
2. inputData in server should be format of reactive
3. This exampple may declare module, but it will not work.

Using multiple modules

As mentioned, there are other modules,

So let’s add ETCModule to both ui and server

ui <- fluidPage(
  mod_binarizeModule_ui(
    id = 'binModule'
  ),
  
  mod_etcModule_ui(
    id = 'etcModule'
  )
)

server <- function(input, output, session) {
  mod_binarizeModule_server(
    id = 'binModule', 
    inputData = reactive(datatoys::scholarship) # remotes::install_github('statgarten/datatoys')
  )
  
  mod_etcModule_server(
    id = 'etcModule', 
    inputData = reactive(datatoys::scholarship) 
  )
}

shinyApp(ui, server) # run application

Here is three big problems.

1. Each module’s UI will simultaneously added to below. (which looks bad)
2. Each module’s transform will be independent. That is, even binModule change column A, etcModule will not recognize it.
3. Module’s transform will not effect to inputData

So here’s one usage of door that utilize this scissor module.

  library(shiny)

ui <- fluidPage(
  reactable::reactableOutput('table'), # view data
  
  tabsetPanel( # Modules as tabset
    id = "transformPanel",
    tabPanel(
      title = "Round",
      mod_roundModule_ui("roundModule_1")
    ),
    tabPanel( # Log2 / Log / Log10
      title = "Log",
      mod_logModule_ui("logModule_1")
    ),
    tabPanel(
      title = "Replace",
      mod_replaceModule_ui("replaceModule_1")
    ),
    tabPanel(
      title = "ETC",
      mod_etcModule_ui("etcModule_1")
    ),
    tabPanel(
      title = "Binarize",
      mod_binarizeModule_ui("binModule_1")
    ),
    tabPanel(
      title = "Split",
      mod_splitModule_ui(id = "splitModule_1")
    )
  ),
  actionButton( # actionbutton to apply transform
    inputId = "applyTransform",
    label = tagList(
      phosphoricons::ph("arrow-circle-right"),
      "Apply changes"
    ),
    width = "100%"
  )
)

server <- function(input, output, session) {
  
  # Data handling
  inputData <- reactive(datatoys::scholarship)
  data_rv <- reactiveValues(data = NULL)
  
  observeEvent(inputData(), {
    data_rv$data <- inputData()
  })
  
  # Data view
  output$table <- reactable::renderReactable({
    reactable::reactable(data_rv$data)
  })
  
  
  # Data transform
  res_round <- mod_roundModule_server(
    id = "roundModule_1",
    inputData = reactive(data_rv$data)
  )
  
  ## log Module
  
  res_log <- mod_logModule_server(
    id = "logModule_1",
    inputData = reactive(data_rv$data)
  )
  
  ## replace Module
  
  res_replace <- mod_replaceModule_server(
    id = "replaceModule_1",
    inputData = reactive(data_rv$data)
  )
  
  ## binarize Module
  
  res_binary <- mod_binarizeModule_server(
    id = "binModule_1",
    inputData = reactive(data_rv$data)
  )
  
  ## etc Module
  
  res_trans <- mod_etcModule_server(
    id = "etcModule_1",
    inputData = reactive(data_rv$data)
  )
  
  ## Split Module
  
  res_split <- mod_splitModule_server(
    id = "splitModule_1",
    inputData = reactive(data_rv$data)
  )
  
  # update data
  observeEvent(input$applyTransform, {
    if (input$transformPanel == "Round") {
      data_rv$data <- res_round()
    }
    if (input$transformPanel == "Log") {
      data_rv$data <- res_log()
    }
    if (input$transformPanel == "Replace") {
      data_rv$data <- res_replace()
    }
    if (input$transformPanel == "Binarize") {
      data_rv$data <- res_binary()
    }
    if (input$transformPanel == "ETC") {
      data_rv$data <- res_trans()
    }
    
    if (input$transformPanel == "Split") {
      data_rv$data <- res_split()
    }
    
    inputData(data_rv$data) # then use isolated
  })
}

shinyApp(ui, server) # run application

Note that, module consist or layout may change by user’s goal.

However, we strongly recommend utilize these modules as tabsetpanel or modal

And finally, Action button for transform must be added (it is not included in module now. but we consider it in further update version)

Structure of transformModule

transformModule is consisted with Control Area and Description Area (no Result Area)

Description area explains expected transform result with example control options.

and below is using flow.

1. Declare module (we did already)
2. Select function and column to transform, and apply button
3. Finish (result should be linked to view like table)

For any issue or suggestion, please make issue in scissor’s github.