server.R

#
# This is the server logic of a Shiny web application. You can run the 
# application by clicking 'Run App' above.
#
# Find out more about building applications with Shiny here:
# 
#    http://shiny.rstudio.com/
#

library(shiny)
# library(devtools)
# install_github('MoBiodiv/mobsim')    # downloads the latest version of the package
library(mobsim, lib.loc="./Library")

# Define server logic for slider examples
shinyServer(function(input, output, session) {
  
  # if(keep){
  #   ...
  # }
  
  # update range for species richness, an observed species has minimum one individual
  observe({
    updateSliderInput(session, "S", min=5, max=input$N, value=5, step=5)
  })
  
  output$ui <- renderUI({
    if (is.null(input$sad_type))
      return()
    switch(input$sad_type,
           "lnorm"=sliderInput("coef", label="CV(abundance), i.e. standard deviation of abundances divided by the mean abundance",value=1, min=0, max=5, step=0.1, ticks=F),
           "geom"=sliderInput("coef",label="Probability of success in each trial. 0 < prob <= 1",value=0.5,min=0,max=1,step=0.1, ticks=F),
            "ls"=textInput("coef",label="Fisher's alpha parameter",value=1)
    )
  })
   
  ## plot theme
  output$InteractivePlot <- renderPlot({
    input$Restart
    
    isolate({
      set.seed(229376)
    
      n.mother <- ifelse(input$spatdist=="n.mother", as.numeric(input$spatcoef), NA)
      n.cluster <- ifelse(input$spatdist=="n.cluster", as.numeric(input$spatcoef), NA)
      
      sim.com <- switch(input$sad_type,
                        "lnorm"=sim_thomas_community(s_pool = input$S, n_sim = input$N, 
                                                     sigma=input$spatagg, mother_points=n.mother, cluster_points=n.cluster,
                                                     sad_type = input$sad_type, sad_coef=list(cv_abund=input$coef),
                                                     fix_s_sim = T),
                        "geom"=sim_thomas_community(s_pool = input$S, n_sim = input$N,                                                    sigma=input$spatagg, mother_points=n.mother, cluster_points=n.cluster, sad_type = input$sad_type, 
                                                    sad_coef=list(prob=input$coef),
                                                    fix_s_sim = T),
                        "ls"=sim_thomas_community(s_pool = input$S, n_sim = input$N,                                                    sigma=input$spatagg, mother_points=n.mother, cluster_points=n.cluster,
                                                  sad_type = input$sad_type, 
                                                  sad_coef=list(N=input$N,alpha=as.numeric(input$coef)),
                                                  fix_s_sim = T)
                        
      )
      
      layout(matrix(c(1,2,3,
                      4,5,6), byrow = T, nrow = 2, ncol = 3),
             heights = c(1,1), widths=c(1,1,1))
      
      sad1 <- community_to_sad(sim.com)
      sac1 <- spec_sample_curve(sim.com)
      divar1 <- divar(sim.com)
      dist1 <- dist_decay(sim.com)
      
      plot(sad1, method = "octave")
      plot(sad1, method = "rank")
      
      plot(sim.com)
      
      plot(sac1)
      plot(divar1)
      plot(dist1)
    })
    
  })
  
})