How to convert a model from NetLogo to Mesa

This is a short list for quick conversion of the most important instructions from NetLogo to Python, with the Mesa implementation as a target.

Create conversion patterns

conversion patterns for the big three instructions: ask, of, with.

Let us assume that the agents are represented by the agentset turtles in NetLogo and by the list 'agents' in Python.

1. ask pattern: ask turtles [ believe ] becomes:

for x in self.schedule.agents: x.believe()

For the next patterns we use list comprehension (https://en.wikipedia.org/wiki/List_comprehension).

2. of pattern: let a [ shell ] of turtles becomes:

[x.shell for x in self.schedule.agents]

3. usage of with will depend on the main form (ask, of) and will just turn in a if construct, in one of the two following ways:

#ask turtles with [ miscredent? ] [ believe ]
for x in self.schedule.agents if x.miscredent: x.believe()

#let a [ shell ] of turtles with [ into-soup? ]
a = [x.shell for x in self.schedule.agents if x.into-soup]

extractions

1. n-of, the first extraction pattern that is learned in NetLogo:

# let a n-of 10 turtles
import random
a = random.sample(a, 10)
# conversion accurate from NetLogo version 6.1.1 on; code before 6.1.1 would fail if len(a) < 10
a = a if len(a) < 10 else random.sample(a,10)

conversion of advanced structures

There are some patterns we use a lot. Random extraction is one of them; in NetLogo, there's the really useful rnd:weighted_one_of construct. A possible conversion pattern is as follows:

#let chosen rnd:weighted-one-of turtles [ holiness-weight ]
import numpy as np
chosen = np.random.choice(self.schedule.agents,
    p = [x.holiness_weight for x in self.schedule.agents],
	size.self.schedule.agents=1,
	replace=False)[0]

Note that the replace=False makes little sense here, but helps when turning this into a pattern when we move from one to n, that is, to convert rnd:weighted_n_of, which is left as an exercise for the reader.

Next: the tutorial

Once you have understood the basic transfomation rules, you can continue with the Tutorial, that will drive you step by step in the trasformation of a NetLogo model in Mesa-based python.