Visualizing randomized four dimensional data set

Question

I have a four dimensional data set. None of the four variables are equally spaced. Right now, I visualize the data using 3D scatter (with the color of the dots indicating the fourth dimension). But this makes it extremely unwieldy while it is printed. Had the variables been evenly spaced,a series of pcolors would have been an option. Is there some way, wherein I can represent such a data using a series of 2D plots? My data set looks something like this:

   x = [3.67, 3.89, 25.6]
   y = [4.88, 4.88, 322.9]
   z = [1.0, 2.0, 3.0]
   b = [300.0,411.0,414.5]

Answer 1

A scatter plot matrix is a common way to plot multiple dimensions. Here's a plot of four continuous variables colored by a fifth categorical variable.

To deal with the uneven spacing, it depends on the nature of the unevenness.

• You might plot it as-is if the unevenness is significant.
• You might make a second plot with the extreme values excluded.
• You might apply a transformation (such as log or quantile) if the data justifies it.