# Which packages or functions can I use in R to plot 3D data like this?

There're many data points, each of which is associated with two coordinates and a numeral value, or three coordinates. And I wish it is coloured. I checked packages "scatterplot3d" and "plot3D" but I couldn't find one like the example I give. It is like it has a fitting surface.

My data is basically like the following. In this way I think this kind of plot is gonna be perfectly suitble for this data:

    ki,kt,Top10AverageF1Score
360,41,0.09371256716549396
324,41,0.09539634212851525
360,123,0.09473510831594467
36,164,0.09773486852645874
...


But I also may have one more additional variable, which makes it like:

    NeighborhoodSize,ki,kt,Top10AverageF1Score
10,360,41,0.09371256716549396
15,324,41,0.09539634212851525
15,360,123,0.09473510831594467
20,36,164,0.09773486852645874
...


Do you also have any good idea for visualizing the second case? What kind of plot and which packages and functions, etc.

• Be clearer about the structure of your data. Is it already gridded? Or do you need to interpolate it to a grid first? Because that's a whole other big question. Then (once you have a grid) can we talk about 3d graphics. Show us a chunk of your data. – Spacedman Oct 21 '14 at 11:44

You could use the wireframe function from the lattice package:
library("lattice")