R/module_tsne.R

In chasemc/IDBac_App: MALDI-TOF MS Protien and Small Molecule Analysis

#' UI module for creating absolute panel popup
#'
#' @param id shiny namespace
#'
#' @return PCA UI
#' 
#'
popupPlotTsne_UI <- function(id){
  ns <- shiny::NS(id)
  tagList(
    actionButton(ns("openAbsPanel"), "Open t-SNE plot"),
    uiOutput(ns("absPanel"))
  )
}


#' popupPlot_server
#' 
#' @param input shiny nput
#' @param output shiny output
#' @param session shiny session
#' @param namedColors named colors to have same colors as dend
#' @param data data
#' @param plotTitle title of plot
#'
#' @return NA
#' 
#'
popupPlotTsne_server <- function(input,
                                 output,
                                 session,
                                 data,
                                 namedColors,
                                 plotTitle = "t-SNE"){ 
  
  
  dataFrame <- reactive({
    
    validate(need(nrow(data()) > 1, "t-SNE requires more samples"))
    
    validate(need(input$tsnePerplexity > 0, "Perplexity must be greater than 0"))
    validate(need(input$tsnePerplexity < 101, "Perplexity must be less than 101"))
    
    validate(need(input$tsneTheta > 0, "Theta must be greater than 0"))
    validate(need(input$tsneTheta < 1, "Theta should be less than 1"))
    
    validate(need(input$tsneIterations > 0, "Iterations must be greater than 0"))
    validate(need(input$tsneIterations < 10000, "Iterations must be less than 1000"))
    
    #tsneCalculation(dataMatrix = dataframe(),
    tsneCalculation(dataMatrix = data(),
                              perplexity = input$tsnePerplexity,
                              theta = input$tsneTheta,
                              iterations = input$tsneIterations)
    
  })
  
  
  
  
  output$plot <- plotly::renderPlotly({
    req(nrow(dataFrame()) > 2,
        ncol(dataFrame()) > 2)
    
    
    if (is.null(namedColors())) {
      colorsToUse <- cbind.data.frame(fac = rep("#000000", nrow(dataFrame())), 
                                      dataFrame())
    } else {
      
      colorsToUse <- cbind.data.frame(fac = as.vector(namedColors()), 
                                      nam = (names(namedColors())))
      
      
      colorsToUse <- merge(dataFrame(),
                           colorsToUse, 
                           by = "nam")
    }
    
    plotly_3d_scatter(data = colorsToUse,
                                plotTitle = plotTitle)
    
  })
  
  observeEvent(input$closeAbsPanel, {
    output$absPanel <- renderUI({
      # Intentionally Blank
    })
  })  
  
  observeEvent(input$openAbsPanel,
               ignoreInit = T,
               ignoreNULL = T,
               {
                 output$absPanel <- renderUI(

                tsneUiPop(session$ns, 
                                    plotTitle = plotTitle)
                 )
                 
               })
}





#' tSNE UI
#'
#' @param ns shiny namespace
#' @param plotTitle plot title
#'
#' @return shiny ui
#' 
#'
tsneUiPop <- function(ns, plotTitle){
  
shiny::fixedPanel(
  class = "popup_Plots",
  top = "20%",
  bottom = "20%",
  width = "60%",
  draggable = F,
  style = "z-index:1002;",
  p(plotTitle),
  absolutePanel(
    top = "0%",
    bottom = "95%",
    right = "5%",
    left = "95%",
    fixed = F,
    shiny::actionButton(ns("closeAbsPanel"),
                        class = "closeX",
                        label = "",
                        icon = icon("far fa-window-close"))
  ),
  
  fluidRow(
    tagList(
      shiny::numericInput(inputId = ns("tsnePerplexity"),
                          label = "Perplexity",
                          value = 10,
                          min = 1,
                          max = 100,
                          step = 1,
                          width = NULL),
      shiny::numericInput(inputId = ns("tsneTheta"),
                          label = "Theta",
                          value = .5,
                          min = 0,
                          max = 1,
                          step = 0.1,
                          width = NULL),
      shiny::numericInput(inputId = ns("tsneIterations"),
                          label = "Iterations",
                          value = 1000,
                          min = 1,
                          max = 10000,
                          step = 1,
                          width = NULL)
      
    )
  ),
  
  fluidRow(                     
    plotly::plotlyOutput(ns("plot"),
                         width = "100%", 
                         height = "100%")
  )
)
}