def set_parameter(age,lambda_set,alpha_set,min_cluster,if_delete):
"""Initilization of SOINN, calling this function after training will reset the
neural network for further training. age, lambda_set, alpha_set,min_cluster,if_delete
are the SOINN parameters meaning max age, learning step, neuron clustering coefficient,
minimum desired clusters and a choice whether to delete the neurons without neighbors
in the final round of training."""
global age_max
global nn_lambda
global alpha
global setN
global accumulated
global numbers
global density
global big_clusters
global t
global minimum_cluster
global gr
global delete_noise
delete_noise = if_delete
t = 1
setN = []
accumulated = [1,1]
numbers = [1,1]
setLabel = [0,0]
big_clusters = []
density = []
nn_lambda = lambda_set
age_max = age
alpha = alpha_set
minimum_cluster = min_cluster
gr = graph()
return
def set_parameter(age, lambda_set, ann_set, bnn_set, eb_, en_):
"""Initilization of GNG, calling this function after training will reset the
neural network for further training. age, lambda_set, ann_set, bnn_set, eb_,
en_ are the GNG parameters meaning max age, learning step, winner adaptation size,
neighbor of winner adaptation size, and error reduction sizes."""
global age_max #:max age parameter of GNG
global nn_lambda #:training step of GNG
global alpha
global setN #:set of neuron weights
global accumulated #:winning times of each neuron
global numbers
global density
global big_clusters #:a dictionary stating which cluster of a neuron belong to
global t
global minimum_cluster
global gr
global delete_noise
global eb
global en
global ann
global bnn
global accumulated_error
ann = ann_set
bnn = bnn_set
eb = eb_
en = en_
t = 1
setN = []
accumulated = [1, 1]
accumulated_error = [0, 0]
numbers = [1, 1]
big_clusters = []
density = []
nn_lambda = lambda_set
age_max = age
gr = graph()
return
"""The single layer SOINN algorithm procedures, for developers only."""
from numpy import *
from . import __grouping2 as grouping2
from math import isnan
from oldgraph.classes.graph import graph
winner_index = 0
winner_2nd_index = 0
setN = [] #:neuron weights
gr = graph() #:topology structure implemented by python-graph
T_winner = 0
T_2nd_winner = 0
accumulated = [1,1]
big_clusters = []
density = []
age_max = 0
nn_lambda = 0
alpha = 0
t = 1
numbers = [1,1]
setLabel = [0,0]
miminum_cluster = 2
delete_noise = False
def set_parameter(age,lambda_set,alpha_set,min_cluster,if_delete):
"""Initilization of SOINN, calling this function after training will reset the
neural network for further training. age, lambda_set, alpha_set,min_cluster,if_delete
are the SOINN parameters meaning max age, learning step, neuron clustering coefficient,
minimum desired clusters and a choice whether to delete the neurons without neighbors
in the final round of training."""
## print 'cut tree',yourcut.values(),grade
yourset = list(yourcut.values())
for i in range(grade):
print(min(yourset))
yourset.remove(min(yourset))
max_degree = min(yourset)
## print max_degree,yourset
for edge_name, cut_degree in list(yourcut.items()):
if (cut_degree < max_degree + 1) and (gr.has_edge(edge_name)):
## print edge_name
gr.del_edge(edge_name)
print('cluster break x 1', edge_name)
if __name__ == '__main__':
gr = graph()
setN = [
array([0.0, 0.0]),
array([1.0, 0.0]),
array([0.5, 0.5]),
array([0.5, 1.0]),
array([-0.1, -0.1]),
array([2.0, 1.0]),
array([2.0, 0.0]),
array([1.5, 0.5])
]
gr.add_nodes(list(range(8)))
gr.add_edge((0, 1))
gr.add_edge((0, 2))
gr.add_edge((2, 1))
gr.add_edge((3, 2))