add viviane's bfs implementation
This commit is contained in:
parent
9fae37347b
commit
ef7646f556
53
main.py
53
main.py
@ -172,6 +172,59 @@ def check_ants_follow_gradient():
|
|||||||
print(20*"#")
|
print(20*"#")
|
||||||
model.step()
|
model.step()
|
||||||
|
|
||||||
|
def viviane_bfs_example_run():
|
||||||
|
# Breadth-first-search algorithm for connectivity
|
||||||
|
def bfs(graph, start_node, threshold): #graph=grid, start_node=nest, threshold=TBD?
|
||||||
|
from collections import deque
|
||||||
|
visited = set()
|
||||||
|
queue = deque([(start_node, [])])
|
||||||
|
paths = {}
|
||||||
|
connected_food_sources = set()
|
||||||
|
|
||||||
|
while queue:
|
||||||
|
current_node, path = queue.popleft()
|
||||||
|
#current_node = tuple(current_node)
|
||||||
|
visited.add(current_node)
|
||||||
|
|
||||||
|
if current_node in graph:
|
||||||
|
for neighbor, m.grid.fields["A"] in graph[current_node].items():
|
||||||
|
if neighbor not in visited and m.grid.fields["A"] >= threshold:
|
||||||
|
new_path = path + [neighbor]
|
||||||
|
queue.append((neighbor, new_path))
|
||||||
|
|
||||||
|
# Check if the neighbor is a food source
|
||||||
|
if neighbor in self.grid_food:
|
||||||
|
if neighbor not in paths:
|
||||||
|
paths[neighbor] = new_path
|
||||||
|
connected_food_sources.add(neighbor)
|
||||||
|
|
||||||
|
connectivity = len(connected_food_sources)
|
||||||
|
|
||||||
|
return connectivity
|
||||||
|
|
||||||
|
|
||||||
|
# Calculate connectivity through BFS
|
||||||
|
|
||||||
|
current_paths = bfs(self.grid, self.grid.fields["nests"], 0.000001)
|
||||||
|
|
||||||
|
import numpy as np
|
||||||
|
|
||||||
|
N = 121
|
||||||
|
N_X = int(np.sqrt(N))
|
||||||
|
N_Y = N // N_X
|
||||||
|
# fancy way of saying absolutely nothing but 11
|
||||||
|
|
||||||
|
xv, yv = np.meshgrid(np.arange(N_X), np.arange(N_Y), sparse=False, indexing='xy')
|
||||||
|
|
||||||
|
|
||||||
|
print(f"{N_X=}")
|
||||||
|
|
||||||
|
print(f"{N_Y=}")
|
||||||
|
|
||||||
|
print(f"{(xv, yv)=}")
|
||||||
|
|
||||||
|
print(f"{xv=}")
|
||||||
|
|
||||||
|
|
||||||
from model import kwargs_paper_setup1 as kwargs
|
from model import kwargs_paper_setup1 as kwargs
|
||||||
if __name__ == "__main__":
|
if __name__ == "__main__":
|
||||||
|
57
model.py
57
model.py
@ -15,7 +15,6 @@ from multihex import MultiHexGridScalarFields
|
|||||||
from mesa.time import SimultaneousActivation
|
from mesa.time import SimultaneousActivation
|
||||||
from mesa.datacollection import DataCollector
|
from mesa.datacollection import DataCollector
|
||||||
from agent import RandomWalkerAnt
|
from agent import RandomWalkerAnt
|
||||||
from collections import deque
|
|
||||||
|
|
||||||
kwargs_paper_setup1 = {
|
kwargs_paper_setup1 = {
|
||||||
"width": 100,
|
"width": 100,
|
||||||
@ -101,6 +100,7 @@ class ActiveWalkerModel(Model):
|
|||||||
self.e_min : float = e_min # energy at which walker dies
|
self.e_min : float = e_min # energy at which walker dies
|
||||||
self.N_f : int = N_f #num food sources
|
self.N_f : int = N_f #num food sources
|
||||||
self.successful_ants = 0 # for viviane's graph
|
self.successful_ants = 0 # for viviane's graph
|
||||||
|
self.connectivity = 0 # for viviane's persistence
|
||||||
|
|
||||||
fields=["A", "B", "nests", "food", "res"]
|
fields=["A", "B", "nests", "food", "res"]
|
||||||
self.schedule = SimultaneousActivation(self)
|
self.schedule = SimultaneousActivation(self)
|
||||||
@ -141,39 +141,42 @@ class ActiveWalkerModel(Model):
|
|||||||
|
|
||||||
|
|
||||||
# Breadth-first-search algorithm for connectivity
|
# Breadth-first-search algorithm for connectivity
|
||||||
#def bfs(graph, start_node, threshold): #graph=grid, start_node=nest, threshold=TBD?
|
# TODO: Implement pheromone B (take max of the two or sum?)
|
||||||
# visited = set()
|
# alex: what's to say against max?
|
||||||
# queue = deque([(start_node, [])])
|
def bfs(self):
|
||||||
# paths = {}
|
threshold = 0.0000001 #the value of A
|
||||||
# connected_food_sources = set()
|
connectivity = 0 #initial value of connectivity
|
||||||
|
connected_food_sources = list() #empty list of connected food sources
|
||||||
|
visited = list() #empty list of visited (by the algorithm) nodes
|
||||||
|
|
||||||
# while queue:
|
nest = np.argwhere(self.grid.fields["nests"] == 1) #get nest location
|
||||||
# current_node, path = queue.popleft()
|
nest = nest[0].tolist() #transforming not to have type errors
|
||||||
#current_node = tuple(current_node)
|
nest = tuple(nest) #transforming not to have type errors
|
||||||
# visited.add(current_node)
|
start_node = nest #rename
|
||||||
|
|
||||||
# if current_node in graph:
|
neighbours_to_check = list([start_node]) #start node gets checked first
|
||||||
# for neighbor, m.grid.fields["A"] in graph[current_node].items():
|
neighbours_to_check = neighbours_to_check + self.grid.get_neighborhood(start_node) #start node neighbours get added to the to check list
|
||||||
# if neighbor not in visited and m.grid.fields["A"] >= threshold:
|
|
||||||
# new_path = path + [neighbor]
|
|
||||||
# queue.append((neighbor, new_path))
|
|
||||||
|
|
||||||
# Check if the neighbor is a food source
|
while neighbours_to_check: #as long as there is something on the to check list
|
||||||
# if neighbor in self.grid_food:
|
current_node = neighbours_to_check[0] #the first list entry is taken
|
||||||
# if neighbor not in paths:
|
del neighbours_to_check[0] #and deleted on the to check list
|
||||||
# paths[neighbor] = new_path
|
|
||||||
# connected_food_sources.add(neighbor)
|
|
||||||
|
|
||||||
# connectivity = len(connected_food_sources)
|
if current_node not in visited: #if it has not previously been checked
|
||||||
|
if self.grid.fields["A"][current_node] >= threshold: #and its A value is above our threshold
|
||||||
|
new_neighbors = self.grid.get_neighborhood(current_node) #then we get its neighbours
|
||||||
|
if new_neighbors not in visited: #if they have not yet been visited
|
||||||
|
neighbours_to_check = neighbours_to_check + new_neighbors #then they are also added to our to check list
|
||||||
|
visited = visited + list([current_node]) #and the current node has now been checked
|
||||||
|
|
||||||
# return connectivity
|
neighbours_to_check = list(dict.fromkeys(neighbours_to_check)) #only check nodes once (unique values)
|
||||||
|
|
||||||
|
if self.grid.fields["food"][current_node] > 0: #in case the node we check is food
|
||||||
|
connectivity += 1 #then we have found a connected path to a food source
|
||||||
|
connected_food_sources = connected_food_sources + list([current_node]) #and it is added to the list of connected food sources
|
||||||
|
|
||||||
# Calculate connectivity through BFS
|
return connectivity #we want the connectivity (0-5)
|
||||||
|
|
||||||
# current_paths = bfs(self.grid, self.grid.fields["nests"], 0.000001)
|
|
||||||
|
|
||||||
|
|
||||||
|
self.connectivity = bfs(self)
|
||||||
|
|
||||||
|
|
||||||
self.datacollector = DataCollector(
|
self.datacollector = DataCollector(
|
||||||
@ -182,7 +185,7 @@ class ActiveWalkerModel(Model):
|
|||||||
"pheromone_b": lambda m: m.grid.fields["B"],
|
"pheromone_b": lambda m: m.grid.fields["B"],
|
||||||
"alive_ants": lambda m: m.schedule.get_agent_count(),
|
"alive_ants": lambda m: m.schedule.get_agent_count(),
|
||||||
"sucessful_walkers": lambda m: m.successful_ants,
|
"sucessful_walkers": lambda m: m.successful_ants,
|
||||||
#"connectivity": lambda m: check_food_source_connectivity(self.grid_food,current_paths),
|
"connectivity": lambda m: m.connectivity,
|
||||||
},
|
},
|
||||||
agent_reporters={}
|
agent_reporters={}
|
||||||
)
|
)
|
||||||
|
Loading…
Reference in New Issue
Block a user