R Shiny：dygraphs系列对`date_window`作出反应

Question

在这个miniapp中，目标是显示原始系列及其在所选范围内的平均值：

library(dygraphs)
library(datasets)

server <- function(input, output) {

  reacteddata <- reactive({

    dt = cbind(as.xts(ldeaths),ave=NA)
    if (!is.null(input$dygraph_date_window)){
      start=strftime(input$dygraph_date_window[[1]])
      end=strftime(input$dygraph_date_window[[2]])
      subset = window(as.xts(ldeaths), start=start, end=end)
      ave = rep(mean(subset), length(subset))
      dt[index(as.xts(subset)),"ave"] = ave
      dt = dt[index(as.xts(subset))]
    } else {
      dt[,"ave"] = rep(mean(ldeaths), length(ldeaths))
    }
    dt
  })

  output$dygraph <- renderDygraph({
    dygraph(reacteddata(), main = "Predicted Deaths/Month")
  })
}

ui <- fluidPage(

  sidebarLayout(
    mainPanel(
      dygraphOutput("dygraph")
    )
  )
)

shinyApp(ui = ui, server = server)

它可以工作，甚至可以重绘缩放上的平均线（使用鼠标选择缩放日期范围）：

然而，问题是它在每次重绘时都会丢失数据，因此无法缩小。任何想法如何重做？

Answer 1

有助于在dt元素之外保留完整的reactive数据集，并根据所选的被动ave更新dygraph_date_window（平均值）列。 此外，retainDateWindow需要设置为TRUE。

library(dygraphs); library(shiny); library(datasets); library(xts)

server <- function(input, output) {
  dt = setNames(as.xts(ldeaths), "ldeaths")
  dt = cbind(dt,ave=NA)

  reacteddata <- reactive({
    if (!is.null(input$dygraph_date_window)){
      start=strftime(input$dygraph_date_window[[1]])
      end=strftime(input$dygraph_date_window[[2]])
      subset = window(dt, start=start, end=end)
      ave = rep(mean(subset$ldeaths), nrow(subset))
      dt[index(as.xts(subset)),"ave"] = ave
    } else {
      dt[,"ave"] = rep(mean(ldeaths), length(ldeaths))
    }
    dt
  })

  output$dygraph <- renderDygraph({
    dygraph(reacteddata(), main = "Predicted Deaths/Month") %>% 
      dyOptions(retainDateWindow = TRUE)
  })
}

ui <- fluidPage(
      dygraphOutput("dygraph")
)

shinyApp(ui = ui, server = server)