surface: Create a 3D surface plot for any function

In janoleko/simply3d: Convenient package for fast creation of dynamic 3D Plots

Create a 3D surface plot for any function

Description

Create a 3D surface plot for any function

Usage

surface(
  fig = NULL,
  expr,
  x = NULL,
  y = NULL,
  xlim = c(0, 40),
  ylim = c(0, 40),
  colors = c("#ff6c59", "#ffd859"),
  showscale = TRUE,
  showlegend = FALSE,
  xlab = "x",
  ylab = "y",
  zlab = "z",
  n = 50
)

Arguments

fig	Optional: A scatterplot onto which to plot the surface. If provided please also provide the x and y-variables used for the scatterplot. If not provided please specify the xlim and ylim arguments for the plot.
expr	An expression or a function containing x and y as variables. If a function is provided is does not need to be vectorizable, this will be done automatically.
x	x-variable used for the scatterplot if surface should be plotted ontop (Only needs to be provided if fig is provided).
y	y-variable used for the scatterplot if surface should be plotted ontop (Only needs to be provided if fig is provided).
xlim	A vector of limits for the x-axis.
ylim	A vector of limits for the y-axis.
colors	A vector of two colors (can also be the same) for the color gradient.
showscale	Option for showing or hiding the colorscale
showlegend	Option for showing or hiding the legend
xlab	x-axis label
ylab	y-axis label
zlab	z-axis label
n	number of points in each dimension where function is evaluated. The function will be evaluated n^2 times.

Value

Returns a interactive 3D surface plot, either only the surface or the surface plotted ontop of a scatterplot if the scatterplot is provided as the fig argument.

Examples

## Example 1: Creating a surface plot from a simple expression

surface(expr = 4*x + 3*y + 0.5*x*y)

# changing the colors
surface(expr = 4*x + 3*y + 0.5*x*y, color = c("green", "blue"))


## Example 2: Creating a surface plot from a predefined function

f = function(x,y){
-x*y*exp(-x^2-y^2)
}

# and increasing n for a higher resolution (as function has high curvature).
# Also changing xlim and ylim.
surface(expr = f, xlim = c(-5,5), ylim = c(-5,5), n = 200)


## Example 3: Adding a surface to a regression plot (e.g. to plot a linear regression)

# generating data
x = rnorm(1000, 10, 20)
y = rnorm(1000, 20, 20)
z = 4 + 0.3*x + 0.2*y + rnorm(1000, 0, 4)

# plotting scatterplot with simply_scatter
scatter = simply_scatter(x,y,z)

# fitting a linear regression model
mod = lm(z ~ x + y)

# adding model plane into scatterplot
surface(fig = scatter, expr = coef(mod)[1] + coef(mod)[2]*x + coef(mod)[3]*y, x = x, y = y)

# you can also use the pipe operator
require(dplyr)
simply_scatter(x,y,z) %>%
  surface(expr = coef(mod)[1] + coef(mod)[2]*x + coef(mod)[3]*y, x = x, y = y)

# for more complex models defining an expression as above might be tedious.
# You can also just define your model prediction as a function
f = function(x,y){
predict(mod, newdata = data.frame(x,y))
}

surface(fig = scatter, expr = f, x = x, y = y)

# or even shorter
surface(fig = scatter, expr = predict(mod, newdata = data.frame(x,y)), x = x, y = y)

janoleko/simply3d documentation built on March 1, 2023, 2:58 a.m.