Translational/Rotational Energy of Molecules

I am creating rigid bodies from groups of spheres using, for example,

fix 1 all rigid molecule

From doing some simple tests with two spheres setup as a diatomic molecule, it appears that LIGGGHTS will always output the total energy - translational + rotational - for ke. I want to separate this energy into rotational and translational pieces in the thermo output. LAMMPS seems to have had ways to do this with compute temp, but that does not appear to be a thing in LIGGGHTS. I tried using erotate, but that does not seem to correlate with what I need at all.

For a single molecule, I can determine angular momentum and so forth for the molecule. This becomes tricky when there is more than one molecule. I see loops are an option in an input file. Do I need to loop through all molecules to determine these values using mass, vx, ixx, etc.? I have not come across any loop examples that do anything similar to this.

I will attach my simple diatomic case. The energy should be all rotational under the current input data file conditions.

Thanks!