The Wolf-Sheep-Grass model

This is a Python reimplementation of NetLogo's Wolf-Sheep-Grass model https://ccl.northwestern.edu/netlogo/models/WolfSheepPredation

Install via

pip install -e .

within this directory. (The "editable" -e flag may be omitted if you do not plan on changing the model.)

Typical usage can be viewed in wolves-sheep-grass.py. e.g. the model can be instantiated with parameters:

from wolf_sheep_grass import WolfSheepGrassModel

model = WolfSheepGrassModel(
    GRID_WIDTH=...,
    GRID_HEIGHT=...,
    INIT_WOLVES=...,
    WOLF_GAIN_FROM_FOOD=...,
    WOLF_REPRODUCE=...,
    INIT_SHEEP=...,
    SHEEP_GAIN_FROM_FOOD=...,
    SHEEP_REPRODUCE=...,
    INIT_GRASS_PROPORTION=...,
    GRASS_REGROWTH_TIME=...,
)

The model is advanced forward in time using model.time_step(). Classic usage will find model.num_wolves,model.num_sheep, and sum(model.grass) interesting. More advanced usage might look at model.grass directly which is a 2d boolean array indicating grass presence or one of the agent arrays:

• model.wolf_pos / model.sheep_pos
• model.wolf_dir / model.sheep_dir
• model.wolf_energy / model.sheep_energy

each of which should be masked by the boolean arrays model.wolf_alive or model.sheep_alive. That is, arrays such as model.wolf_pos are fixed length, (mostly) independent of the number of wolves and contain meaningless entries. To get only the meaningful entries, use model.wolf_pos[model.wolf_alive].