CODEWITHSUNDEEP
rggplot2shiny
TheComeOnMan
TheComeOnMan

Reputation: 12905

Overlay Two ggplot in Shiny

I have a very large dataset that I'm plotting on Shiny using ggplot. I have a a slider tied to the values on x axis with which I would like to recolor the selected subset of the data and let the rest of the data remain as is.

The simplest option is to recreate the entire plot but since it's a large dataset it's a very slow process. As an alternative, I was considering letting the original plot stay as is, and putting another plot on top of it with the plot.background = 'none' and just the coloured points. I'm unable to achieve this with the renderPlot in the server.r and plotOutput in the ui.r -

server.r:

output$Plot1= renderPlot({
   ggplot(mtcars) + geom_point(aes(x = mpg, y = wt))
})

output$Plot2= renderPlot({
       ggplot(mtcars[mtcars$mpg > 15,]) + geom_point(aes(x = mpg, y = wt), color = 'Efficient')
})

ui.r (I don't know how to fit Plot2 in this):

fluidRow(
               class = 'Plot1Row',
               column(
                  plotOutput(
                     "Plot1", 
                     width = '100%'
                  ),
                  width = 12
               )
            )

Any tips?

Upvotes: 3

Views: 1410

Answers (1)

Roland
Roland

Reputation: 132969

Here is an example explaining my comment:

library(ggplot2)
p <- ggplot(mtcars[mtcars$mpg > 15,]) + geom_point(aes(x = mpg, y = wt), color = 'green')
p    

#produce an otherwise identical gtable with a subset of points with different color 
p1 <- ggplot(mtcars[mtcars$mpg > 15,]) + geom_blank(aes(x = mpg, y = wt), color = 'green') +
  geom_point(data = mtcars[mtcars$mpg > 25,], aes(x = mpg, y = wt), color = 'blue')
g1 <- ggplot_gtable(ggplot_build(p1))

library(gtable)
#find the correct viewport
seekViewport(grid.grep("panel", grep = TRUE)$name)
#draw just the points 
grid.draw(gtable_filter(g1, "panel")$grobs[[1]]$children$geom_point.points)

resulting plot

This doesn't redraw the complete plot. Since I'm not a shiny user, you'll have to test yourself how to do this in shiny.

Upvotes: 1

Related Questions