Automatic model weighting

[perrette]

Add an automatic model democracy/ensemble democracy weighting option to the emulator function call, either as True or False or as weights for a dict with weights for each model or ensemble member in the dataset. Keep the possibility to weight individual simulations in csv file, but give the possibility to rename the weighting column. Like this one only needs one dataset for different weighting options according to different criteria, e.g. if tipping points are triggered in a simulation.

[originally posterd in https://github.com//issues/23#issue-2473505220]