selecting points on 3D scatterplots

It looks like Rcmdr may be able to select points on 3D scatterplots
however when I try to use it's 3dscatter plot function I get the error
message: ?could not find function "nice"

If I copy the code:

scatter3d(data$X, data$Z, data$Y, surface=FALSE, residuals=TRUE, bg="white",
+ axis.scales=TRUE, grid=TRUE, ellipsoid=FALSE, xlab="X", ylab="Z", zlab="Y")

into the R console I get the same error message. Sorry I'm new - does
anyone know where this missing "nice" function can be found?


I tried using scatterplot3d but it doesn't rotate or zoom - which I
need to be able to do to select the data... but thanks for the
suggestion!


2009/5/12 Uwe Ligges <ligges at statistik.tu-dortmund.de>:

Or with scatterplot3d:

The regular identify() usage with the iris data:

?s3d <- scatterplot3d(iris[,1:3], color=as.numeric(iris[,5]))
?identify(s3d$xyz.convert(iris[,1:3]))

or for locator: the values you get are the 2D representation of the current
coordinate system and you can use locator directly for putting text in such
as in:

text(locator(1), "The iris data")

If you want to convert back to 3D: This is not possible, because there is a
while line behind that point you are clicking on ....

Best,
Uwe Ligges




Duncan Murdoch wrote:

On 5/12/2009 3:08 PM, Abby Drake wrote:

Hello Everyone,

I am new to R and need some help.

I have a matrix of x,y,z coordinates that I would like to
interactively plot in 3D and then using the cursor select points on
the plot and have the coordinates sent to a matrix. I am using the rgl
package to plot the data at the moment because it allows me to rotate
and zoom. I also tried cloud and scatterplot3D.

I am looking for a function like 'locator' which is used to select
points on 2D scatterplots.

There's no locator3d, but select3d is somewhat similar.

You could write a locator3d function using the rgl.setMouseCallbacks,
rgl.user2window and rgl.window2user functions, if you can figure out how to
indicate depth with a mouse click. ?identify3d might be a little easier.

Duncan Murdoch