Best way to convert from a 1D matrix to a 3D matrix with isotropic distribution

I have 1D data which has values at equally spaced intervals from r=0 to r=r, where r is a distance i.e.

 r         x
0.0  1.16000
0.2  1.21000
0.5  1.24000
0.7  1.25000
0.9  1.24000
1.2  1.22000
1.4  1.21000
1.6  1.18000 
etc

This data is isotropic (from r=0) in three dimensions, so I want to go from a 1D dataset to 3D. I thought that the best option may be to convert to spherical polar coordinates using the function cart2pol and writing a For loop to loop around all values of theta and phi, filling in the values, then convert back to cartesian coodinates. However, I know that for loops can be time consuming computationally in Matlab and wondered if there was a better way of doing it? I am using Matlab version R2011B with the image processing toolbox. Thanks!

[Azzi Abdelmalek]

The conversion is based on what? How can use cart2pol with one 1-D array?

[Catherine Scott]

A good point, I don't actually need to convert to spherical polar coordinates because I already have r (which is rho in the Matlab documentation) and I can assume that the data I have is at theta=phi=0 (where theta= azimuth and phi=elevation). The next bit still stands I believe, so I would then loop over all possible values of theta and phi, then convert to cartesian (using sph2cart). So is there a more efficient method for this part? Apologies for the mix up and thanks for the comment.