def onInterest_StartDE(self, prefix, interest, face, interestFilterId, filter):
interestName = interest.getName()
print "Interest Name: %s" %interestName
interest_name_components = interestName.toUri().split("/")
if "start_de" in interest_name_components:
#print 'Query database'
print 'Call decision engine algorithm'
parent_dir = os.path.split(self.script_dir)[0]
monitor_path = os.path.join(self.script_dir, parent_dir, 'Monitoring', 'Monitoring_DB')
print monitor_path
myDE = de(monitor_path)
json_lst_dict = myDE.get_lst_of_dictionaries()
json_server_Spec = self.json_server_Spec_default
node_name = myDE.selectHost_to_deploy_firstInstance(json_lst_dict, json_server_Spec)
print 'Selected Host Name %s' %node_name
### User will add this parameter via trigger
service_name = 'uhttpd.tar'
print 'Start service deployment'
deployService = self.prefix_deployService + node_name + '/' + service_name
config_prefix_deployService = Name(deployService)
interest = Interest(config_prefix_deployService)
interest.setInterestLifetimeMilliseconds(self.interestLifetime)
interest.setMustBeFresh(True)
self.face.expressInterest(interest, None, None) ## set None --> sent out only, don't wait for Data and Timeout
print "Sent Push Interest to SEG %s" % config_prefix_deployService
else:
print "Interest name mismatch"
def func(Dimension_number):
Time = []
for i in range(Dimension_number):
Fn = 20 # Function to be evaluated
Dimension = i + 1
popsize = 20
mut = 0.9
crossp = 0.9
iter_max = 1500
func_eval = fobj # func to be evaluated
ca = 512
# Execution
start = time.time()
target, fitness = de(func_eval, mut, crossp, popsize, iter_max, Fn, ca,
Dimension)
end = time.time()
Time.append(end - start)
print('Time taken to Execute this code = {} seconds'.format(end -
start))
return Time
num_input = 15
num_hidden = 60
num_output = 6
weight_interval = 2000
pop_size = 200
gen_limit = 500
inps = loader.load('paru.xlsx')
inps = loader.stringifyVar(inps,loader.normalizeVar(loader.getVar(inps)))
# inps = loader.loadDat('titanic.dat')
start = time.time()
weights = ann.decodeWeights(de.de((num_output*num_hidden)+(num_hidden*num_input),weight_interval,pop_size,gen_limit))
end = time.time()
print end - start
<<<<<<< HEAD
# match = 0
# for i in range(len(inps)):
# if(ann.correct(inps[i],weights)):
# match += 1
# print match,len(inps)
# print num_hidden,weight_interval,pop_size,gen_limit,'ret = 1/(1+np.exp((-0.001*x)))'
=======
num_input = 3
num_hidden = 10
import time
from de import de
from func import fobj
from figure_plot import figure_plot
import matplotlib.pyplot as plt
# Inputs
Fn = 3 # Function to be evaluated
popsize = 50
mut = 0.9
crossp = 0.9
iter_max = 25
func_eval = fobj # func to be evaluated
ca = 512
# Execution
start = time.time()
target, fitness = de(func_eval, mut, crossp, popsize, iter_max, Fn, ca)
end = time.time()
print('Time taken to Execute this code = {} seconds'.format(end - start))
if len(target[0]) == 2:
plt.title('Target vector after {} iteration'.format(
iter_max)) # plot for reference
figure_plot(target, popsize, Fn, ca) # comment if unnecessary
import time
from de import de
from func import fobj
from figure_plot import figure_plot
import matplotlib.pyplot as plt
# Inputs
Fn = 11 # Function to be evaluated
popsize = 20
mut = 0.9
crossp = 0.9
iter_max = 50
func_eval = fobj # func to be evaluated
ca =512
# Execution
start = time.time()
a , b = de(func_eval, mut, crossp, popsize, iter_max, Fn,ca)
end = time.time()
print('Time taken to Execute this code = {} seconds'.format(end - start))
if len(a[0]) == 2:
plt.title('Target vector after {} iteration'.format(iter_max)) # plot for reference
figure_plot(a , popsize ,Fn,ca ) # comment if unnecessary
SBX_parents[j][1],
c=parent_color,
s=parent_size)
plt.xticks([])
plt.yticks([])
plt.title('Simulated Binary Crossover')
# create DE plot
plt.subplot(232)
for j in range(len(DE_parents)):
plt.scatter(DE_parents[j][0],
DE_parents[j][1],
c=parent_color,
s=parent_size)
for i in range(10):
(cx, cy) = de.de(DE_parents, DE_crossover_rate, DE_scaling_factor)
plt.scatter(cx, cy, c=child_color, s=child_size)
plt.xticks([])
plt.yticks([])
plt.title('Differential Evolution')
# create UM plot
plt.subplot(233)
for i in range(num_samples):
child = um.um(UM_parent, bounds, UM_probability)
plt.scatter(child[0], child[1], c=child_color, s=child_size)
plt.scatter(UM_parent[0], UM_parent[1], c=parent_color, s=parent_size)
plt.axhline(y=UM_parent[1], xmin=bounds[0][0], xmax=bounds[0][1], c='black')
plt.axvline(x=UM_parent[0], ymin=bounds[1][0], ymax=bounds[1][1], c='black')
plt.xticks([])
plt.yticks([])
import loader
import ann
import de
import time
num_input = 15
num_hidden = 60
num_output = 6
inps = loader.load('paru.xlsx')
inps = loader.stringifyVar(inps,loader.normalizeVar(loader.getVar(inps)))
# inps = loader.loadDat('titanic.dat')
# print inps
start = time.time()
weights = ann.decodeWeights(de.de((num_output*num_hidden)+(num_hidden*num_input),2000,200,3,0.5,0.5,10))
end = time.time()
print end - start
# # print weights
# # print inps[0]
# match = 0
# for i in range(len(inps)):
# if(ann.correct(inps[i],weights)):
# match += 1
# print match,len(inps)
# print num_hidden,2000,500,'ret = 1/(1+np.exp((-0.001*x)))'
import loader import jst import de import sys import numpy as np num_input = int(sys.argv[1]) num_hidden = int(sys.argv[2]) num_output = int(sys.argv[3]) weight_interval = int(sys.argv[4]) pop_size = int(sys.argv[5]) gen_limit = int(sys.argv[6]) data_learning = sys.argv[7] format = data_learning[-4:-1]+data_learning[-1] save_weights_to = sys.argv[8] if (format == '.dat'): inps = loader.loadDat(data_learning) elif (format == '.csv'): inps = loader.loadCsv(data_learning) inps = loader.stringifyVar(inps,loader.normalizeVar(loader.getVar(inps))) weights = de.de((num_output*num_hidden)+(num_hidden*num_input),weight_interval,pop_size,gen_limit,num_input,num_hidden,num_output,data_learning) np.savetxt(save_weights_to, np.array(weights), delimiter=',')
