# Flocking + Clustering

# original Flocking Sketch by Daniel Shiffman.

#

# In this adaptation of Dan Shiffman's Flocking example, the boids are

# repeatedly clustered using the KMeans algorithm from scikit-learn. Each Boid

# will ignore Boids in other clusters, but each Boid will also change cluster

# membership over time. The cluster labels are used to give the boids in each

# cluster a unique color.

#

# An implementation of Craig Reynold's Boids program to simulate

# the flocking behavior of birds. Each boid steers itself based on

# rules of avoidance, alignment, and coherence.

#

# Click the mouse to add a new boid.

import traceback

from jpype import JClass

from sklearn.cluster import KMeans

import py5_tools

import py5

NUMBER_OF_CLUSTERS = 8

cluster_centers = None

def cluster_count():

# the number of clusters should be set in one place for both the Python and Java code

return NUMBER_OF_CLUSTERS

def cluster_boids(boid_locations):

global cluster_centers

try:

init_arg = 'k-means++' if cluster_centers is None else cluster_centers

kmeans = KMeans(n_clusters=NUMBER_OF_CLUSTERS, random_state=0, init=init_arg, n_init=1).fit(boid_locations)

# save the cluster centers to initialize clustering in the next invocation

# this keeps cluster labels stable from one call to the next

cluster_centers = kmeans.cluster_centers_.copy()

return kmeans.labels_

except Exception as e:

traceback.print_exc()

return JClass('java.lang.RuntimeException')(str(e))

# py5_tools.register_processing_mode_key("cluster_boids", cluster_boids)

# py5_tools.register_processing_mode_key("cluster_count", cluster_count)

# run the sketch in processing mode, specifying the Java class to instantiate

py5.run_sketch(jclassname='test.Example3Sketch')