Experimenter defines spawning rules: 8pts

[skleinfeldt]

User story:

As an experimenter I can define overall spawning rules for the map: whether objects are randomly placed or clumped.

[Note: this may have elements in common with the wall-building algorithms. Natalia will take the lead on algorithm development.]

• [x] Apply configurable spawning algorithms to item types

[alecpm]

On further investigation GU already handles this pretty well, and we're going to be generalizing the existing spawning rules as part of the definition of objects in #232 in order to replicate the current functionality for "Food" objects. The following optional properties will be available on all objects:

• initial_count: how many to put on the map initially (formerly num_food)
• respawn: Should a consumed object be immediately replaced in another location (formerly respawn_food)
• public_good_multiplier: Should the calories of this object be split among all players when consumed (formerly food_pg_multiplier)
• growth_rate: The rate at which the number of objects increases (formerly food_growth_rate)
• seasonal_growth_rate: A rate multiplier that changes per game round. In odd rounds the growth_rate is multiplied by this value, and in even rounds it's divided by this value.
• probability_distribution: A named probability distribution used for placing objects, see distributions.py (formerly food_probability_distribution defaults to random).

Somewhat related are:

• maturation_speed: The rate at which objects "mature" and eventually "spoil" or disappear (formerly food_maturation_speed).
• maturation_threshold: The maturity level at which an object is removed from the map (formerly food_maturation_threshold)

[alecpm]

This was 95% done when implementing #232, but I did a small refactor to support configuring the initial player distribution. I've included that change into PR #248. Both aspects seem to be working well.