shiny - R闪亮预测ARMA

Question

我想学习 ARMA 预测，我正在尝试使用从比特币下载的比特币数据集进行预测

我正在努力使用以下代码。图书馆（Quandl）

 library(forecast)
    library(tseries)
    library(shiny)


    ui <- shinyUI(fluidPage(
      titlePanel("Simple ARMA Bitcoin Forecasting"),
      sidebarLayout(
         sidebarPanel("Please choose the number of periods",
                     numericInput("num",h3("Numeric input"), value = 1)),

      mainPanel(plotOutput("plot")))
    )
    )

    server <- shinyServer(function(input,output) {

      output$plot <- plotOutput({

        bitcoin <- Quandl("BITSTAMP/USD",type = "xts")

        bitcoin_price <- bitcoin[,3]

        barplottest <- diff(log(bitcoin_price))

        fit <- auto.arima(dataInput)

        fit %>% forecast(h="input$num") %>% autoplot()
          })

    })

    shinyApp(ui,server)

这是我收到的错误：

 Warning: Error in as.ts: object 'dataInput' not found
Stack trace (innermost first):
    44: as.ts
    43: auto.arima
    42: imageOutput [#11]
    41: plotOutput
    40: server [#3]
     4: <Anonymous>
     3: do.call
     2: print.shiny.appobj
     1: <Promise>
Error in as.ts(x) : object 'dataInput' not found

Answer 1

好吧，我不确定是dataInput什么

这是highcharter包装的示例：

library(shiny)
library(Quandl)
library(forecast)
library(highcharter)

bitcoin <- Quandl("BITSTAMP/USD",type = "xts")
bitcoin_price <- bitcoin[,3]
barplottest <- diff(log(bitcoin_price))
fit <- auto.arima(barplottest )

ui <- shinyUI(fluidPage(
  titlePanel("Simple ARMA Bitcoin Forecasting"),
  sidebarLayout(
    sidebarPanel("Please choose the number of periods",
                 numericInput("num",h3("Numeric input"), value = 1)),

    mainPanel(highchartOutput("hcontainer",height = "500px")))
)
)

server <- shinyServer(function(input,output) {

  output$hcontainer <- renderHighchart({
    hchart(forecast(fit,h=input$num))
  })
})

shinyApp(ui,server)