plot3Drgl: Tools for plotting 3-D and 3-D data in openGL'
In plot3Drgl: Plotting Multi-Dimensional Data - Using 'rgl'

```{echo = FALSE, include = FALSE} output: rmarkdown::html_vignette \usepackage[T1]{fontenc}

```r
knitr::opts_chunk$set(echo = TRUE)
require(plot3Drgl)

Introduction

The R-package plot3D provides functions for plotting 2- and 3-D data, using R base graphics.

Package plot3Drgl allows to plot these functions also in openGL, as made available by package rgl.

One way to use openGL is to first create a plot in base R-graphics using plot3D functiona, and then use function plotrgl() to depict the same figure in rgl.

The main advantage of rgl over base graphics is that it allows to interactively rotate, zoom, and shift the graphics, and even select regions.

A related package that depends on plot3Drgl is R-package OceanView which contains functions for visualing oceanographic data.

Function plotrgl

Typically you can start by making a 3D plot using functions from package plot3D. Although not necessary, plotting can be postponed by setting argument plot = FALSE

persp3D(z = volcano, plot = FALSE)

The figure is then plotted in openGL by function plotrgl, whose arguments are:

args(plotrgl)

Here the ... are any parameter that would be passed to rgl functions par3d, open3d or material3d (see documentation in the rgl package).

In the code below, argument \emph{smooth} adds Gouraud shading, while \emph{lighting} adds a light source.

plotrgl(smooth = TRUE, lighting = TRUE)

knitr::include_graphics("volcanorgl.png")

Now you can use the left mouse key to rotate the plot, the middle mouse key to move it, and the right key to zoom. You may also want to try function cutrgl, which allows to cut parts of the plot. This however only works if there is no color key.

An alternative, shorter version to do the same is:

persp3Drgl(z=volcano, smooth=TRUE, lighting=TRUE, colkey=FALSE)

knitr::include_graphics("volcanocut.png")

\newpage

scatter plot example

A linear regression of the mtcars data can be easily plotted both in base graphics and using rgl:

attach(mtcars)
fit <- lm(mpg ~ wt + disp)

# predict values on regular xy grid
wt.pred <- seq(1.5, 5.5, length.out = 30)
disp.pred <- seq(71, 472, length.out = 30)
xy <- expand.grid(wt = wt.pred,
                  disp = disp.pred)

mpg.pred <- matrix (nrow = 30, ncol = 30,
    data = predict(fit, newdata = data.frame(xy),
    interval = "prediction")[,1])

# fitted points for droplines to surface
fitpoints <- predict(fit)

scatter3D(z = mpg, x = wt, y = disp, colvar = abs(mpg - fitpoints),
      pch = 18, cex = 2, theta = 20, phi = 20, ticktype = "detailed",
      xlab = "wt", ylab = "disp", zlab = "mpg", main = "mtcars",
      clab = "error", zlim = c(5, 35),
      surf = list(x = wt.pred, y = disp.pred, z = mpg.pred,
      facets = NA, border = "black", fit = fitpoints)
      )

detach(mtcars)

plotrgl()

knitr::include_graphics("plot3Drglfit_rgl.png")

\newpage

isosurfaces

Function isosurf3D from package plot3D creates surfaces of equal scalar value from a volumetric data set.

It makes use of a function from package misc3d(Feng and Tierney, 2008) .

If we depict several isosurfaces, it is best to use transparent colors by setting argument alpha smaller than 1. Plotting transparent surfaces is very slow in base graphics, but not so in openGL.

 x <- y <- z <- seq(-2, 2, length.out = 15)
 xyz <- mesh(x, y, z)
 F <- with(xyz, log(x^2 + y^2 + z^2 +
                10*(x^2 + y^2) * (y^2 + z^2) ^2))

# three levels, transparency added
 isosurf3D(x, y, z, F, level = seq(0, 4, by = 2),
   col = c("red", "blue", "yellow"),
   clab = "F", alpha = 0.2)

plotrgl()

knitr::include_graphics("plot3Drgliso_rgl.png")

\newpage

Issues

Sometimes the axes are not drawn in rgl plots. If you want axes, just type

decorate3d()

The package contains a function to visualise arrows in rgl as cones. But it has a flaw, as the arrows are distorted, if not perpendicular to the z-axis. Use with care

References

Duncan Murdoch and Daniel Adler (2021). rgl: 3D Visualization Using OpenGL. R package version 0.108.3. https://CRAN.R-project.org/package=rgl

Feng D, Tierney L (2008). "Computing and Displaying Isosurfaces in R." Journal of Statistical Software, 28(1). URL http://www.jstatsoft.org/v28/i01/.

Karline Soetaert (2021). plot3D: Plotting Multi-Dimensional Data. R package version 1.4. https://CRAN.R-project.org/package=plot3D

Karline Soetaert (2022). plot3Drgl: Plotting Multi-Dimensional Data - Using 'rgl'. R package version 1.0.3. https://CRAN.R-project.org/package=plot3Drgl

Karline Soetaert (2021). OceanView: Visualisation of Oceanographic Data and Model Output. R package version 1.0.6. https://CRAN.R-project.org/package=OceanView