Need help

[BENFREDJRima-34]

Hello everyone,
I would like to implement an agent-based model, and for that, I need a function that allows me to place an agent in a grid according to a specific condition, otherwise, according to a condition (x) I decide the number of neighbors of an agent, the function must be checked for all the agents of the grid.
Thank you in advance

[rht]

For the former requirement, an example would be in this Sugarscape code

mesa/examples/sugarscape_cg/sugarscape_cg/model.py

Lines 49 to 53 in b6accc6

	max_sugar = sugar_distribution[x, y]
	sugar = Sugar(agent_id, (x, y), self, max_sugar)
	agent_id += 1
	self.grid.place_agent(sugar, (x, y))
	self.schedule.add(sugar)

, you can change it to

        for _, x, y in self.grid.coord_iter():
            max_sugar = sugar_distribution[x, y]
            if max_sugar <= 0:
                continue
            sugar = Sugar(agent_id, (x, y), self, max_sugar)
            agent_id += 1
            self.grid.place_agent(sugar, (x, y))
            self.schedule.add(sugar)

i.e. it skips the creation of the Sugar agent when max_sugar is <= 0. You can of course choose to create an agent but not place it to the grid.

For the latter requirement, it is better to solve this problem instead: generate a matrix that consists of only 0s and 1s, where each matrix element is constrained to have a certain condition for its number of neighbors. You could start by a matrix of all zeros, and set one random position in it to 1, and then set N of its neighborhood to 1, while doing a constraint check before every step to ensure that your change to the matrix satisfies the overall number of neighbors constraint. There could already be a solution for generating this matrix if you model it in a Networkx directed graph, but I wouldn't do this as a first resort.

[BENFREDJRima-34]

ok, I will check it, Thank you so much