Two different plots from different libraries on the same page

Question

I have one plot (levelplot) from the lattice library, and a stick plot from the oce library, and I'd like to stack them on top of each other. I'm having problems stacking these plots as I would normally using base R, and I assume it is because these are both specialized plots. Is it possible to stack the two below plots on the same page so that the timestamps (x axis) line up with each other?

library(oce)
library(lattice)
library(grid)

First Plot (example from help docs of OCE)

# Oceanographic example
data(met)
t <- met[["time"]]
u <- met[["u"]]
v <- met[["v"]]
p <- met[["pressure"]]

dev.new(width=15, height=4)
plotSticks(t, 99, u, v, yscale=25)

Second plot:

AllT<-NULL
for(i in seq(1,length(t))){

    myTime<-t[i]
    myTime2<-rep(myTime,10)
    AllT<-append(AllT,myTime2)      
}

myY<-NULL
for(j in seq(1,length(t))){
    mySeq<-seq(1,10)
    myY<-append(myY,mySeq)    
}

Temp<-seq(1,7200)


MyDF<-data.frame(Temp,myY,AllT)



#Plot code
dev.new(width=15, height=6)
p1<-levelplot(Temp ~ AllT * myY,
    data = MyDF,ylim=c(10,1),
    xlab = "Time", ylab = "y]",
    aspect=0.4,
    )

p1

trellis.focus("legend", side="right", clipp.off=TRUE, highlight=FALSE)
grid.text('[Temp]', 1.9, .5, hjust=0.5, vjust=1,rot=270)
trellis.unfocus()

Ideally, the final plot would have the stick plot on top (no need for tick labels on the stick plot if x axis are lined up), with the level plot below with minimal space in between

I was hoping `ggplotify` would be able to just convert the plot to `grobs`, but I don't think it can. `OCE` tends to pretty unique because of the data it handles, and possibly because of all its `S4` objects. Getting the plots as `grobs` would be easy to do what you're looking for. — Anonymous coward, Sep 17 '18 at 17:39
I have yet to find a way to plot this kind of wind speed/ direction other than using oce, so I'm tethered to using this library for now — Vint, Sep 17 '18 at 18:13

score 1 · Answer 1 · answered Sep 17 '18 at 19:58

This is as good as I can figure out for now. There's got to be a way to actually modify the grobs like I'm aware of, but it doesn't like the gTrees when I try. So this gets you close.

These links may be of some use.

https://www.andrewheiss.com/blog/2016/12/08/save-base-graphics-as-pseudo-objects-in-r/

force a regular plot object into a Grob for use in grid.arrange

#libraries
library(gridGraphics)
library(grid)
library(gridExtra)
library(oce)
library(gggplot2)
library(ggplotify)

#plots
data(met)
t <- met[["time"]]
u <- met[["u"]]
v <- met[["v"]]
p <- met[["pressure"]]

grab_grob <- function(){
  grid.echo()
  grid.grab()
  }

plotSticks(t, 99, u, v, yscale=25)
p1 <- grab_grob() #grab the last plot as a grob

AllT <- NULL
for(i in seq(1, length(t))){
  myTime <- t[i]
  myTime2 <- rep(myTime, 10)
  AllT <- append(AllT, myTime2)      
}

myY <- NULL
for(j in seq(1, length(t))){
  mySeq <- seq(1, 10)
  myY <- append(myY, mySeq)    
}

Temp <- seq(1, 7200)

MyDF <- data.frame(Temp, myY, AllT)

#Plot code
dev.new(width = 15, height = 6)
p2 <- levelplot(Temp ~ AllT * myY,
          data = MyDF, ylim=c(10, 1),
          xlab = "Time", ylab = "y]",
          aspect = 0.4,
)

p2

trellis.focus("legend", side="right", clipp.off = TRUE, highlight = FALSE)
grid.text('[Temp]', 1.9, .5, hjust = 0.5, vjust = 1, rot = 270)
trellis.unfocus()

p2a <- as.grob(p2)

grid.arrange(p1, p2a, ncol = 1)

Two different plots from different libraries on the same page

1 Answers1