library(shiny)
library(markdown)
#install.packages("kst")
library("kst")
#install.packages("pks")
library("pks")
#install.packages("DAKS")
library("DAKS")
#install.packages("Rgraphviz")
library(Rgraphviz)
#install.packages("kstMatrix")
library(kstMatrix)
#install.packages('shinyBS')
library(shinyBS)
#install.packages("plotrix")
library(plotrix)

PAGE_TITLE <- "Validating simulated data in knowledge space theory"

legend.text = c("dpot = average difference between power set and knowledge structure", "dpot")


spaceselection <- c(
  "Density" = "density97",
  "Matter" = "matter97",
  "DoignonFalmagne7" = "DoignonFalmagne7"
)


data("DoignonFalmagne7")
data("Taagepera")

# Define UI ----
ui <- fluidPage(
    includeScript("../../../Matomo-tquant.js"),
  
  theme = "bootstrap.css",
  titlePanel(windowTitle = PAGE_TITLE,
    tags$a(imageOutput("tquant.png", width = "5px", height = "5px"), href="https://tquant.eu/", PAGE_TITLE
                    ))
  ,
  

  fluidRow(column(width = 3, 
                          selectInput("spacesel", "Select your knowledge structure", spaceselection),
                          sliderInput(
                            "beta",
                            "Careless error rate",
                            min = 0,
                            max = 0.5,
                            step = 0.01,
                            value = 0.1
                          ),
                          sliderInput(
                            "eta",
                            "Lucky guess rate",
                            min = 0,
                            max = 0.5,
                            step = 0.01,
                            value = 0.25
                          ),
                          numericInput(
                            "n",
                            "Number of respondents",
                            min = 10,
                            max = 5000,
                            value = 500
                          ),
                          actionButton("new", "New Simulation")
                        ),
                        column(9, plotOutput("ex1"))),
  fluidRow(column(width = 6,
                  plotOutput("bar_plot")),
                  
                  column(3, plotOutput("hist")),
           column(3, tableOutput("tab"))),
  
  fluidRow(column(6,plotOutput("legend")),
           column(6, textOutput("txt1"),
                  textOutput("txt2"),
                  textOutput("txt3"),
                  textOutput("txt4"))

))
                  
  

# Define server logic ----
server <- function(input, output) { 
  	output$legend <- renderPlot({
  	 plot.new()
   	 legend(x = "top",
           legend = c(
             "States",
             "Non-States"
           ),
           col = c("green", "red"),
           lwd = 7
    )
  })


  selections <- reactiveValues(eta = 0.1, beta = 0.25, n = 500, data = density97$K)
   pkst <- reactiveValues(pkst = kmsimulate(density97$K, 500, .25, .1))
    
  
  observeEvent(input$new, {
    output$bar_plot <- renderPlot({
      kst <- selections$data
      
   pkst$pkst <- kmsimulate(selections$data, selections$n, selections$beta, selections$eta)
   pnr <- as.pattern(pkst$pkst, freq = TRUE)
    
      
      cols <- vector(mode = "character", length = length(pnr))
      names(cols) <- names(pnr)
      cols[] <- "red"
      cols[rownames(kst)] <- "green"
      
      barplot(pnr,
              horiz = TRUE,
              space = 0.4,
              names.arg = names(pnr),
              cex.names = 0.75,
              las = 1,
              col=cols,
              main = "Simulated response pattern frequencies")
    })
    
  })
  
  observeEvent(input$beta, {
    selections$beta <- input$beta
  })
  
  observeEvent(input$eta, {
    selections$eta <- input$eta
  })
  
  observeEvent(input$n, {
    selections$n <- input$n
  })
  
  observeEvent(input$spacesel, {
    if (input$spacesel == "density97")
      selections$data <- density97$K
    else if (input$spacesel == "matter97")
      selections$data <- matter97$K
    else if (input$spacesel == "DoignonFalmagne7")
      selections$data <- DoignonFalmagne7$K
  })
  
  output$bar_plot <- renderPlot({
    kst <- selections$data
    
   pkst$pkst <- kmsimulate(selections$data, selections$n, selections$beta, selections$eta)
   pnr <- as.pattern(pkst$pkst, freq = TRUE)
    
    cols <- vector(mode = "character", length = length(pnr))
    names(cols) <- names(pnr)
    cols[] <- "red"
    cols[rownames(kst)] <- "green"
    
    barplot(pnr,
            horiz = TRUE,
            space = 0.4,
            names.arg = names(pnr),
            cex.names = 0.75,
            las = 1,
            col=cols,
            main = "Simulated responses",
            xlab = "Frequency")
  })

 	output$ex1 <- renderPlot({
		ex1 <- selections$data
		plot(as.famset(ex1), main = "Hasse Diagram")
	}) 
  	
 	output$hist <- renderPlot({  
 	  validateD <- kmvalidate(pkst$pkst, selections$data)
 	  dpot <- validateD$DI / validateD$DA  #dpot --> DI / DA
 	  ddat <- validateD$DI
 	  barplot(c(dpot, ddat), width = 1, names.arg = c("dpot","ddat"), col = c("#6699FF","#3366CC"), ylim = c(0,1), las = 1,
 	           border = NA, main = "Validation")
 	  })
 	
 	output$tab <- renderTable({
 	  validateD <- kmvalidate(pkst$pkst, selections$data)
 	  #data.frame(row.names = names(validateD$dist), frequency = as.character(validateD$dist)))
 	  data.frame(dist = names(validateD$dist), frequency = as.character(validateD$dist))
 	})
 	output$example <- renderUI({
 	  tags$a(imageOutput("tquant.png"),href="https://tquant.eu/")
 	})
 	
 	output$txt1 <- renderText({
 	  c(paste("DA = ddat / dpot = ",collapse = "\n"), round(print((kmvalidate(pkst$pkst, selections$data))$DA), digits = 3))
 	})
 	
 	output$txt2 <- renderText({
 	  "ddat: average minimal distance between each response pattern and the nearest state in the knowledge structure"
 	})
 	
output$txt3 <- renderText({
  "dpot: average minimal distance between the power set of all potential response patterns and the nearest state in the knowledge structure"
})
output$txt4 <- renderText({
  "DA: distance agreement coefficient"
})

}

# Run the app ----
shinyApp(ui = ui, server = server)