闪亮:基于输入的SelectInput子集 [英] Shiny: SelectInput subset based on input

问题描述

感谢您的帮助，我的代码存在多个问题，但主要问题是我缺少一个Observe语句，以下内容解决了该问题:

Thanks for the help, there were multiple issues with my code but the main issue was that I was missing an Observe statement, the following solved the problem:

 get_ddf <- reactive({
      filter(poskick, Name == input$player)
    })

  observe({
    updateSelectInput(session, 'fixture', choices   =levels(droplevels(get_ddf()$Event)) )
  })

我希望我的Shiny应用程序具有两个selectInput下拉菜单，第一个选择一个名称，第二个从一个人参与的事件中选择一个，该选择基于第一个输入创建的子集.样本数据:

I want my Shiny app to have two selectInput dropdowns, the first one selecting a name and the second selcting from the events a person took part in, based on a subset created by the first input. sample data:

 PLID   Name        x_coord y_coord x_coord_end y_coord_end action  Event
 7046   Sample Name1    35    37      34          25          4    23/07/11
 7046   Sample Name1    21    11       0           0          4    23/07/11
 7046   Sample Name1    49    60      56           8          4    23/07/11
 7046   Sample Name1    46    56      72          34          4    23/07/11
 7046   Sample Name1    58    49      24          58          4    23/07/11
 7046   Sample Name1    87    57      42          52          4    23/07/11
 7046   Sample Name1    14    58      18          37          4    23/07/11
 7140   Sample Name2    38    14      11          11          4    23/07/11
 7140   Sample Name2    11    11      11          11          4    23/07/11
 7140   Sample Name2    56    8       56           8          4    23/07/11

我的代码用户界面:

library(shiny)
library(ggplot2)


poskick<-read.csv('poskicks.csv')


shinyUI(pageWithSidebar(

  headerPanel("position map"),

  sidebarPanel(



selectInput('player', 'Player', choices= attributes(poskick$Name)),
selectInput('fixture', 'Match', choices= attributes(firstsub()$Fixtu))

),

mainPanel(
  plotOutput('plot')
)
  ))

服务器代码:

library(shiny)
library(ggplot2)

poskick<-read.csv('poskicks.csv')



shinyServer(function(input, output) {


  firstsub <- reactive({
     subset(poskick, poskick$Name %in% input$player)
     })

  secondsub <- reactive({
     subset(poskick, poskick$Fixtu %in% input$fixture & poskick$Name %in%    input$player )
    })


  output$plot <- renderPlot({
    p <- ggplot(data = secondsub()) + geom_segment(aes(x = x_coord, y =   y_coord, xend = x_coord_end, yend = y_coord_end))
    print(p)    }, height=700)

})

任何建议，谢谢.

推荐答案

您的代码中存在一些问题，例如使用Fixtu，但未涉及任何内容.此外，我认为，对于要获取因子变量中唯一值的情况，level()可能比attribute()更好.

There were some issues in your code, like using Fixtu, which did not refer to anything. Also, I think that levels() is probably a better choice than attributes() for getting the unique values in a factor variable.

我发现，当您希望一个窗口小部件中的输入控制另一个窗口小部件的输入时，在server.R文件中使用renderUI会很有帮助.然后，您可以放入return语句，以防止小部件在知道要提供的选项之前甚至无法显示.为此，我添加了一个选择一个"选项，该选项使下一个小部件甚至都不会显示.如果可以将selectInput默认设置为NULL，那会更好，但这不是一个选择.

I find that it's helpful to use renderUI in the server.R file when you want the input into one widget to control the input into another. You can then put in return statements to prevent the widget from even showing up before it knows what options to offer. I do this by adding a "pick one" option that causes the next widget to not even show up. It would be better if you could make the selectInput default to NULL, but that is not an option.

这是我所做的:

server.R:

library(shiny)
library(ggplot2)

poskick<-read.csv('poskicks.csv')

shinyServer(function(input, output) {

  output$Box1 = renderUI(selectInput('player', 
                                     'Player', 
                                     c(levels(poskick$Name),"pick one"),
                                     "pick one")
  )

  output$Box2 = renderUI(
    if (is.null(input$player) || input$player == "pick one"){return() 
    }else selectInput('fixture', 
                      'Match', 
                      c(levels(poskick$Event[which(poskick$Name == input$player)]),"pick one"),
                      "pick one")
    )

  subdata1 = reactive(poskick[which(poskick$Name == input$player),])
  subdata2 = reactive(subdata1()[which(subdata1()$Event == input$fixture),])

  output$plot <- renderPlot({
    if (is.null(input$player) || is.null(input$fixture)){return()
    } else if(input$player == "pick one" || input$fixture == "pick one") { return()
    } else p <- ggplot(data = subdata2()) + geom_segment(aes(x = x_coord, y =   y_coord, xend = x_coord_end, yend = y_coord_end))
    print(p)    })

})

ui.R:

library(shiny)
library(ggplot2)
shinyUI(pageWithSidebar(
  headerPanel("position map"),
  sidebarPanel(uiOutput("Box1"),uiOutput("Box2")),
  mainPanel(plotOutput('plot')
  )
))

这篇关于闪亮:基于输入的SelectInput子集的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！