Fitting a 2D correlated dataset

Hello,

I've been trying to work out a 2D fitting model, but as I started to learn RooFit only today I'm not quite there yet where I want to be. So basically I have two variables, that are correlated: mass and sumpt. Individually both can be fitted with the crystal ball function quite perfectly, but together I'm not able to figure out how to make the fit work.

Have a look at here:
http://neptune.hep.kbfi.ee/mario/roofit.png

The script used to generate it is here:
http://neptune.hep.kbfi.ee/mario/test2DRooFit.py

The top two plots show the individual variables and the relevant fits (already from the 2D model). On the lower right you have the original sample drawn from the tree directly and you can see the correlation between the variables. On the lower left is the fitted model that's drawn to show the shape in 2D. As can be clearly seen simple production of the two 2D shapes for p.d.f. doesn't seem to work (even putting one as conditional of the other). However I'm not sure how to build a complex 2D p.d.f. that would then be usable with the data. Any suggestions are welcome.

Thanks in advance,

Mario Kadastik, PhD
Researcher

---
  "Physics is like sex, sure it may have practical reasons, but that's not why we do it" 
     -- Richard P. Feynman