Much of simulation class's conerganceFilter() should move to GSFC-GRASP-Python-Interface

[ReedEspinosa]

The majority of the conerganceFilter() method in the simulation class within simulateRetrieval.py is code to recalculate χ^2 which is not specific to retrieval simulations and should be moved to the GSFC-GRASP-Python-Interface repository. The method should probably be kept for the purposes of purging rsltBck and rsltFwd of none converging cases, but mostly it would just be a wrapper for the accompanying χ^2 calculation with the (probably) graspDB class.

Also, we should fix the name to correct the spelling of "convergence".

[ReedEspinosa]

Οkay, so this is a little more complicated than I thought...

It seems that the conerganceFilter() method takes the difference between rsltBck['fit_...'] and rsltFwd['fit_...'] (instead of rsltBck['meas_...']). I think the only implication here is that the χ^2 then ignores any random noise that will have been included rsltBck['meas_...'] but not rsltFwd['fit_...']. If we really want to exclude that noise, then we should keep things as they are but I'm not sure we care...