inst/shinyAPP/server.R

In marzione00/marzioneunimi: TempClim2: a smart way to plot EPICA DOME C data

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

  matrix  <- read.csv(system.file("extdata", "Epica-tpt-co2x.csv", package="TempClim2"),header=TRUE,skip=4)
  colnames(matrix) <- c("BP","Temperature","CO2","MIL")


  colour <-"red" #set of reactive variables that allows the user to choose what he would analyse

  type<-reactive({input$"Plot_input"


  })


  cor<-reactive({input$"Color_input"


  })


  lag_1<-reactive({input$"lag1"


  })

  lag_2<-reactive({input$"lag2"

  })

  AR_p<-reactive({input$"ARp"

  })

  MA_q<-reactive({input$"MAq"

  })

  I_O<-reactive({input$"I"

  })



  Fit_set<-reactive({input$"Fit_data"

    if (input$"Fit_data"=="CO2_FIT"){
      CO2_FIT<-arima(matrix[,"CO2"],order=c(AR_p(),I_O(),MA_q()))}

    else if (input$"Fit_data"=="Temperature_FIT"){
      Temperature_FIT <- arima(matrix[,"Temperature"],order=c(AR_p(),I_O(),MA_q()))}



  })

  G_1<-reactive({input$"GR1"

  })

  G_2<-reactive({input$"GR2"

  #  granger<-coint.test(matrix[,G_1()], matrix[,G_2()], d = 0, nlag = NULL, output = TRUE) #granger causality test, to be done

  })


  output$coolplot <- renderPlot( #plot of the data choosen by the user on the panels

#    plot(matrix[,"BP"],matrix[,input$"Plot_input"],col = cor(),type = "l",xlab="Years BP",ylab=input$"Plot_input",cex.lab=1.3,lwd=size()),

    ggplot2::ggplot(matrix,ggplot2::aes_string("BP",input$"Plot_input"))+ ggplot2::geom_point(color=cor(),size=2)

  )

  output$coolplot2 <- renderPlot(


    pacf(matrix[,input$"Plot_input"],lag.max = lag_1(),na.action = na.pass,main="partial Autocorrelation"),

  )

  output$coolplot3 <- renderPlot(


    acf(matrix[,input$"Plot_input"],lag.max = lag_2(),na.action = na.pass,main="Autocorrelation"),

  )

  output$coolplot4<- renderPlot(
    forecast::checkresiduals(Fit_set())
    #ggplot2::ggplot(residuals,ggplot2::aes_string(input$"Fit_data"))+ ggplot2::geom_point()
    #plot(residuals(Fit_set()),main="Residuals",ylab="Res"),

  )

  output$coolplot5<- renderPrint(

    aTSA::coint.test(matrix[,G_1()], matrix[,G_2()],d = 0, nlag = NULL) #granger causality test, to be done

  )

}