Add logging in propensity calculations and a method that takes a model and evaluates its performance at time 0

[artgoldberg]

See email to Yin Hoon 2020/05/11:

Based on the traceback your model is here in the stack:

wc_sim/submodels/nrm.py:initialize_propensities()
wc_sim/submodels/dynamic_submodel.py:calc_reaction_rates()

and then in some eval() method(s) in wc_sim/dynamic_components.py

First, that code doesn’t generate logging messages, so I’ll add some. Second, it’s possible that a small number of the rate laws are taking most of the time. In that case we need to identify them and determine how to reduce their compute time. Third, we’d like to solve this problem without waiting many hours or days for results. Ideally we’d make a test that iterates through all rate law calculations, skipping ones that take too long, and creates a list of those. There are ways to do that, but it needs to be built into wc-sim. Perhaps it needs an initial_performance_eval method that takes a model and evaluates its performance at time 0.

[artgoldberg]

Did "Add logging in propensity calculations"