Closed MichaelClerx closed 6 years ago
I see sampling as one bit of inference (inference also including the process of using data, prior to assemble the distribution we want to sample from).
The trouble with inference as a name, for me, is that you could also think of optimisation as a type of inference. You infer the best fit parameters in optimisation (albeit without uncertainty quantification).
Changed this in the examples & docs
Will mostly be in documentation I think, probably not in the API much