问题
If I have a situation in which about a 1000 black turtles disperse at random angles and steps throughout the netlogo world for a given duration of ticks. Each turtle is assigned a random probability at each timestep during dispersal, and if this number exceeds a given threshold for any given turtle it changes it's color to red and stops moving. Additionally, black turtles (still moving) that happen to move within a patch of red turtles (stopped/settled), change their color to grey and settle (stop moving) as well. Finally, other black turtles (still moving) that happen to be move within a patch of either grey or red turtles (stopped/settled), also change their color to grey and settle (stop moving)
My question is a post-processing question for when the simulation duration is reached. How do I determine the number of clusters of red-grey turtles in the sea of black turtles? Also, how do I determine the size (radial extent) of each cluster? And finally, how do I determine the number of turtles in each cluster?
回答1:
Jen is right: you need a clear idea of what constitutes a cluster before being able to truly answer that question.
That being said, one possible option is to use a clustering algorithm. I'd suggest taking a look at Christopher Frantz's dbscan extension.
Here is a quickly thrown together example:
extensions [ dbscan ]
to setup
clear-all
ask patches [ set pcolor white ]
create-turtles 1000 [
set color black
set label-color blue
setxy random-xcor random-ycor
]
ask n-of 5 turtles [
ask turtles in-radius 3 [
set color one-of [red grey]
]
]
end
to find-clusters
let red-grey-turtles turtles with [ member? color [red grey] ]
let clusters dbscan:cluster-by-location red-grey-turtles 3 3
(foreach clusters range length clusters [ [c i] ->
foreach c [ t ->
ask t [ set label i ]
]
])
end
Sorry for the lack of further explanations: I have a plane to catch...
来源：https://stackoverflow.com/questions/46475893/determining-the-radius-of-turtle-clusters-and-number-of-turtles-in-them-postpr