更新数据框后重绘光泽图？

Question

我用数据库后端构建了一个Shiny应用程序-

我试图重新加载数据帧current_data_frame并在按下input$draw_plot （操作按钮）时使用它重绘该图。

调用数据框后，我在重新绘制绘图时遇到麻烦吗？

任何想法，我缺少什么？

来自server.R的片段

  current_data_frame = data.frame(matrix(ncol = 4, nrow = 0))
  names( current_data_frame ) <- c("sample_id", "call", "intensity_A" , "intensity_B")


  # OBSERVE BUTTON PRESS & UPDATE DATA FRAME
  observeEvent( input$draw_plot, {
                current_data_frame <- get_data_frame( input$probeset_id , input$study_id , input$batch_id)

                }) 

  vals <- reactiveValues(
                        keeprows = rep(TRUE, nrow( current_data_frame ))
                        )

  output$call_plot <- renderPlot({
                        # Lists for holding unactive_points                      
                        keep    <- current_data_frame[ vals$keeprows, , drop = FALSE]
                        exclude <- current_data_frame[ !vals$keeprows, , drop = FALSE]
                        # Le plot 
                        ggplot(keep, aes( intensity_A , intensity_B)) +
                          geom_point(aes(colour = factor(call), shape = factor(call)) )  #+ 
                          #geom_point(data = exclude, shape = 21 , fill = NA, colour = "black", alpha = 0.25)

                        })

                        # Toggle click points 
                        observeEvent( input$call_plot_click, {
                          res <- nearPoints(current_data_frame, input$call_plot_click, allRows = TRUE)

                        vals$keeprows <- xor(vals$keeprows, res$selected_)
                        })

                        # Toggle points that are brushed when clicked 

                        observeEvent(input$exclude_toggle, {
                          res <- brushedPoints(current_data_frame, input$call_plot_brush, allRows = TRUE)

                          vals$keeprows <- xor(vals$keeprows, res$selected_)
                        })

                        # Reset all points 

                        observeEvent( input$exclude_reset, {
                          vals$keeprows <- rep( TRUE, nrow(current_data_frame))
                        })

})  

Answer 1

您可能需要使current_data_frame为反应性值。 你可以做到这一点通过返回它的内部reactive或将其添加到您的vals对象，包括使用vals$current_data_frame无处不在，你目前使用的current_data_frame和不断变化的vals看起来是这样的：

vals <- reactiveValues(keeprows = rep(TRUE, nrow( current_data_frame )),
                       current_data_frame = current_data_frame
                        )

它将vals$current_data_frame设置为您第一次在代码开头定义的默认值，然后允许您在每次触发observeEvent时对其进行更改。

Answer 2

自己回答了这个问题-

将整个输出$ call_plot包装在响应函数中。 现在，每按一次按钮，它就会重新绘制并获取新数据。

下面的代码...

observeEvent( input$draw_plot, {
                current_data_frame <- get_data_frame( input$probeset_id , input$study_id , input$batch_id)



  vals <- reactiveValues(
                        keeprows = rep(TRUE, nrow( current_data_frame ))
                        )

  output$call_plot <- renderPlot({
                        # Lists for holding unactive_points                      
                        keep    <- current_data_frame[ vals$keeprows, , drop = FALSE]
                        exclude <- current_data_frame[ !vals$keeprows, , drop = FALSE]
                        # Le plot 
                        ggplot(keep, aes( intensity_A , intensity_B)) +
                          geom_point(aes(colour = factor(call), shape = factor(call)) )  #+ 
                          #geom_point(data = exclude, shape = 21 , fill = NA, colour = "black", alpha = 0.25)

                        })

                        # Toggle click points 
                        observeEvent( input$call_plot_click, {
                          res <- nearPoints(current_data_frame, input$call_plot_click, allRows = TRUE)

                        vals$keeprows <- xor(vals$keeprows, res$selected_)
                        })

                        # Toggle points that are brushed when clicked 

                        observeEvent(input$exclude_toggle, {
                          res <- brushedPoints(current_data_frame, input$call_plot_brush, allRows = TRUE)

                          vals$keeprows <- xor(vals$keeprows, res$selected_)
                        })

                        # Reset all points 

                        observeEvent( input$exclude_reset, {
                          vals$keeprows <- rep( TRUE, nrow(current_data_frame))
                        })

                    })  

})