R: print multiple types of plots on one page

Question

I am trying to plot multiple plots in one page. I know functions like gridExtra::grid.arrange that can plot graphs generated by ggplot2 package. The problem I am facing is that I have two plots (bar.plot and density.plot below) that are generated by ggplot2 package and one plot generated using limma::vennDiagram function. I have tried the below but it is not working:

output <- paste('summary.pdf')
pdf(output,width = 25,height = 20)
par(mfrow = c(3, 3))
plot(bar.plot)
plot(density.plot)
print(vennDiagram(dat.venn, circle.col = col,cex = c(3,3,3)))
invisible(dev.off())

dat.venn is a data of type VennCounts:

 I-H-10003-T1-D1 I-H-10003-T2-D1 I-H-10003-T3-D1 Counts
               0               0               0      0
               0               0               1     41
               0               1               0     81
               0               1               1     66
               1               0               0     10
               1               0               1      2
               1               1               0      4
               1               1               1     56
attr(,"class")
[1] "VennCounts"

I am unable to find a venn diagram package that is compatible with the grid.arrange function. I don't think that VennCounts cant be printed out with grid.arrange function and ggplot2 plots can be printed out with par function.

UPDATE: I tried using pushViewport but it is still printing the venn diagram on the next page:

pdf(output,width = 25,height = 20)

# Create layout : nrow = 2, ncol = 2
pushViewport(viewport(layout = grid.layout(2, 2)))

# A helper function to define a region on the layout
define_region <- function(row, col){
  viewport(layout.pos.row = row, layout.pos.col = col)
} 

# Arrange the plots
print(bar.plot, vp = define_region(1, 1:2))
print(density.plot, vp = define_region(2, 1))
print(vennDiagram(dat.venn, circle.col = col,cex = c(3,3,3)), vp = define_region(2, 2))
dev.off()

Any help would be much appreciated!

Answer 1

The gridGraphics package, like gridBase, makes the mixing of grid, ggplot, lattice and base graphics fairly straightforward. The grid.echo() function converts a base graphic into a grid graphic. (The advantage of using gridGraphics is that base graphics become amenable to editing and modification via grid; see gridGraphics.pdf.)

In the following, I draw two ggplots and a venn diagram (taken for the vennDiagram help page in the limma package), and I position the three plots within viewports, making sure the venn diagram plot is subjected to the grid.echo treatment.

UPDATE to ggplot2 2.0.0 (Default stat for geom_bar is stat = "count")

  library(ggplot2)
  library(gridGraphics)
  library(grid)
  library(limma)  # From bioC software repository

  grid.newpage()

  # Position 1st ggplot
  pushViewport(viewport(y = 1, height = 1/3, just = "top"))
  gg1 <- ggplot(mtcars, aes(x = mpg)) + geom_density()  
  print(gg1, newpage = FALSE)
  upViewport()

  # Position 2nd ggplot
  pushViewport(viewport(y = .5, height = 1/3, just = "centre"))
  gg2 <- ggplot(mtcars, aes(x = factor(carb, levels = 1:8))) + 
          geom_bar() + scale_x_discrete(drop = FALSE)
  print(gg2, newpage = FALSE)
  upViewport()

  # Function to draw venn diagram - the venn diagram is taken from ?vennDiagram
  plotfun <- function() {   
   Y <- matrix(rnorm(100*6),100,6)
   Y[1:10,3:4] <- Y[1:10,3:4]+3
   Y[1:20,5:6] <- Y[1:20,5:6]+3
   design <- cbind(1,c(0,0,1,1,0,0),c(0,0,0,0,1,1))
   fit <- eBayes(lmFit(Y,design))
   results <- decideTests(fit)
   a <- vennCounts(results)
   print(a)
   mfrow.old <- par()$mfrow
   par(mfrow=c(1,2))
   vennDiagram(a)
   vennDiagram(results, 
       include=c("up", "down"),
       counts.col=c("red", "blue"),
       circle.col = c("red", "blue", "green3"))
  }

  # Position the venn diagram
  pushViewport(viewport(y = 0, height = 1/3, just = "bottom"))
  grid.echo(plotfun, newpage = FALSE)
  upViewport(0)

Answer 2

First, save each of the files as .png files, such as save(yourfile, file = "yourname.png").

Second, use the read.PNG function of the grid package to load them, such as yours.png <- readPNG("yourfile.PNG")

After that, convert them with the rasterGrob function as in g1 <- rasterGrob(yours.png, interpolate=TRUE).

Once all the plots are comparable in format, grid.arrange() does the trick. It might look like this:

grid.arrange(g1, g2, g3, nrow=1)  # one row
dev.copy(png,'threeplots.png')   # to save the array of plots     
dev.off()                         # to close the device