automatically store summary stats on every trial without having to ask

EDIT: Is this to be done in badapted?

Following on from #51 , I want to make Model classes automatically calculate and build up a set of summary stats over all trials (including trial 0 which corresponds to the prior beliefs).

Will be useful as you can use this info after a real or simulated experiment. We get a DataFrame in this general form: screen shot 2018-11-28 at 16 19 00

This is currently done manually in simulated_experiment_trial_loop(). I want the same effect, but done automatically.

Things to do:

[ ] this should apply to all free parameters
[ ] create a new summary_stats property of the Model class.
[ ] automatically call model.get_θ_summary_stats() upon model creation, after the initial samples are generated from the prior. This will represent the summary for before any data was observed.
[ ] automatically call model.get_θ_summary_stats() within the model.update_beliefs() method

drbenvincent / darc_toolbox

automatically store summary stats on every trial without having to ask #54