def experiment1(): l = CMAES(fitnessFunction, myNetwork.params) l.minimize = True l.verbose = True l.maxLearningSteps = 500 params, fitness = l.learn() myNetwork._setParameters(params) logNet()
def experiment1(self):
l = CMAES(self.fitnessFunction, self.myNetwork.params[self.indices])
l.minimize = True
l.verbose = True
l.maxLearningSteps = 500
params, fitness = l.learn()
self.myNetwork.params[self.indices] = params
self.metaInfo["numsteps"] = l.maxLearningSteps
self.metaInfo["fitness"] = fitness
# self.myNetwork._setParameters(self.originalWeights)
self.logNet()
tri_0.append(module.rec_number(main,0)) tri_1.append(module.rec_number(main,1)) tri_2.append(module.rec_number(main,2)) count = count + 1 print count return pow((mean(tri_0) - tri_count[0]),2) + pow((mean(tri_1) - tri_count[1]),2) + pow((mean(tri_2) - tri_count[2]),2) def objF(p) : return graph_function(p) p0 = [0.333,0.3333,0.3333,0,0,0,0] #p0 = [1/7,1/7,1/7,1/7,1/7,1/7,1/7] l = CMAES(objF, p0) l.verbose = True l.minimize = True l._notify() l.desiredEvaluation = 3 g = l.learn() if(g[0][0]<0): g[0][0] = g[0][0]*(-1) if(g[0][1]<0): g[0][1] = g[0][1]*(-1) if(g[0][2]<0): g[0][2] = g[0][2]*(-1) summ = g[0][0] + g[0][1] + g[0][2] print g[0][0]/summ, g[0][1]/summ, g[0][2]/summ print g[1] end_time = time.time() print "The optimization took ", end_time - start_time, " seconds"
