[R] How to generate a smoothed surface for a three dimensional dataset?

[Duncan Murdoch]

On 04/12/2013 11:36 AM, Jun Shen wrote:
> Hi,
>
> I have a dataset with two independent variables (x, y) and a response
> variable (z). I was hoping to generate a response surface by plotting x, y,
> z on a three dimensional plot. I can plot the data with rgl.points(x, y,
> z). I understand I may not have enough data to generate a surface. Is there
> a way to smooth out the data points to generate a surface? Thanks a lot.

There are many ways to do that.  You need to fit a model that predicts z 
from (x, y), and then plot the predictions from that model.
An example below follows yours.
>
> Jun
>
> ===========================
>
> An example:
>
> x<-runif(20)
> y<-runif(20)
> z<-runif(20)
>
> library(rgl)
> rgl.points(x,y,z)

Don't use rgl.points, use points3d() or plot3d().  Here's the full script:


x<-runif(20)
y<-runif(20)
z<-runif(20)

library(rgl)
plot3d(x,y,z)

fit <- lm(z ~ x + y + x*y + x^2 + y^2)

xnew <- seq(min(x), max(x), len=20)
ynew <- seq(min(y), max(y), len=20)
df <- expand.grid(x = xnew,
                   y = ynew)

df$z <- predict(fit, newdata=df)

surface3d(xnew, ynew, df$z, col="red")


Duncan Murdoch
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.