The question is whether there is a difference in x when the final state parton is a gluon or a quark. This was done with the same set up and simulated data as my previous x,Z analyis. 

The trees did not retain the pythia particle ID of the final state partons, but it did have the pythia process ID. All the events in the trees came from 1 of 6 different pythia event types. These are:

Events from processes 11,12, and 53 were used to find x from pion pairs that came from a quark and events from processes 13 and 68 were used to find x from pion pairs coming from gluons. Events from process 28 were thrown out because it was imposible to tell whether the pair came from a quark or gluon. 

I plotted X vs Pt for Eta > 0 and Eta < 0 for pairs coming from quarks and gluons. In both Eta bins, x for pairs coming from quarks was slightly larger than pairs coming from gluons.  

For Eta > 0:

For Eta < 0:

I also did the same thing for Z:

Eta > 0:

Eta < 0:

The total number of events from process id 11,12,53 = 1161331  (quark)
The total number of events from process id 13, 68     =     49815  (gluon)