python中rbf神经网络包_径向基（RBF）神经网络python实现

1 from numpy importarray, append, vstack, transpose, reshape,2 dot, true_divide, mean, exp, sqrt, log,3 loadtxt, savetxt, zeros, frombuffer4 from numpy.linalg importnorm, lstsq5 from multiprocessing importProcess, Array6 from random importsample7 from time importtime8 from sys importstdout9 from ctypes importc_double10 from h5py importFile11 12 13 defmetrics(a, b):14 return norm(a -b)15 16 17 defgaussian (x, mu, sigma):18 return exp(- metrics(mu, x)**2 / (2 * sigma**2))19 20 21 defmultiQuadric (x, mu, sigma):22 return pow(metrics(mu,x)**2 + sigma**2, 0.5)23 24 25 definvMultiQuadric (x, mu, sigma):26 return pow(metrics(mu,x)**2 + sigma**2, -0.5)27 28 29 defplateSpine (x,mu):30 r =metrics(mu,x)31 return (r**2) *log(r)32 33 34 classRbf:35 def __init__(self, prefix = 'rbf', workers = 4, extra_neurons = 0, from_files =None):36 self.prefix =prefix37 self.workers =workers38 self.extra_neurons =extra_neurons39 40 #Import partial model41 if from_files is notNone:42 w_handle = self.w_handle = File(from_files['w'], 'r')43 mu_handle = self.mu_handle = File(from_files['mu'], 'r')44 sigma_handle = self.sigma_handle = File(from_files['sigma'], 'r')45 46 self.w = w_handle['w']47 self.mu = mu_handle['mu']48 self.sigmas = sigma_handle['sigmas']49 50 self.neurons =self.sigmas.shape[0]51 52 def_calculate_error(self, y):53 self.error = mean(abs(self.os -y))54 self.relative_error =true_divide(self.error, mean(y))55 56 def_generate_mu(self, x):57 n =self.n58 extra_neurons =self.extra_neurons59 60 #TODO: Make reusable61 mu_clusters = loadtxt('clusters100.txt', delimiter='')62 63 mu_indices =sample(range(n), extra_neurons)64 mu_new =x[mu_indices, :]65 mu =vstack((mu_clusters, mu_new))66 67 returnmu68 69 def_calculate_sigmas(self):70 neurons =self.neurons71 mu =self.mu72 73 sigmas =zeros((neurons, ))74 for i inxrange(neurons):75 dists = [0 for _ inxrange(neurons)]76 for j inxrange(neurons):77 if i !=j:78 dists[j] =metrics(mu[i], mu[j])79 sigmas[i] = mean(dists)* 280 #max(dists) / sqrt(neurons * 2))81 returnsigmas82 83 def_calculate_phi(self, x):84 C =self.workers85 neurons =self.neurons86 mu =self.mu87 sigmas =self.sigmas88 phi = self.phi =None89 n =self.n90 91 92 defheavy_lifting(c, phi):93 s = jobs[c][1] -jobs[c][0]94 for k, i in enumerate(xrange(jobs[c][0], jobs[c][1])):95 for j inxrange(neurons):96 #phi[i, j] = metrics(x[i,:], mu[j])**3)97 #phi[i, j] = plateSpine(x[i,:], mu[j]))98 #phi[i, j] = invMultiQuadric(x[i,:], mu[j], sigmas[j]))99 phi[i, j] =multiQuadric(x[i,:], mu[j], sigmas[j])100 #phi[i, j] = gaussian(x[i,:], mu[j], sigmas[j]))101 if k % 1000 ==0:102 percent = true_divide(k, s)*100103 print(c, ': {:2.2f}%'.format(percent))104 print(c, ': Done')105 106 #distributing the work between 4 workers107 shared_array = Array(c_double, n *neurons)108 phi =frombuffer(shared_array.get_obj())109 phi =phi.reshape((n, neurons))110 111 jobs =[]112 workers =[]113 114 p = n /C115 m = n %C116 for c inrange(C):117 jobs.append((c*p, (c+1)*p + (m if c == C-1 else0)))118 worker = Process(target = heavy_lifting, args =(c, phi))119 workers.append(worker)120 worker.start()121 122 for worker inworkers:123 worker.join()124 125 returnphi126 127 def_do_algebra(self, y):128 phi =self.phi129 130 w =lstsq(phi, y)[0]131 os =dot(w, transpose(phi))132 returnw, os133 #Saving to HDF5134 os_h5 = os_handle.create_dataset('os', data =os)135 136 deftrain(self, x, y):137 self.n =x.shape[0]138 139 ## Initialize HDF5 caches140 prefix =self.prefix141 postfix = str(self.n) + '-' + str(self.extra_neurons) + '.hdf5'142 name_template = prefix + '-{}-' +postfix143 phi_handle = self.phi_handle = File(name_template.format('phi'), 'w')144 os_handle = self.w_handle = File(name_template.format('os'), 'w')145 w_handle = self.w_handle = File(name_template.format('w'), 'w')146 mu_handle = self.mu_handle = File(name_template.format('mu'), 'w')147 sigma_handle = self.sigma_handle = File(name_template.format('sigma'), 'w')148 149 ## Mu generation150 mu = self.mu =self._generate_mu(x)151 self.neurons =mu.shape[0]152 print('({} neurons)'.format(self.neurons))153 #Save to HDF5154 mu_h5 = mu_handle.create_dataset('mu', data =mu)155 156 ## Sigma calculation157 print('Calculating Sigma...')158 sigmas = self.sigmas =self._calculate_sigmas()159 #Save to HDF5160 sigmas_h5 = sigma_handle.create_dataset('sigmas', data =sigmas)161 print('Done')162 163 ## Phi calculation164 print('Calculating Phi...')165 phi = self.phi =self._calculate_phi(x)166 print('Done')167 #Saving to HDF5168 print('Serializing...')169 phi_h5 = phi_handle.create_dataset('phi', data =phi)170 delphi171 self.phi =phi_h5172 print('Done')173 174 ## Algebra175 print('Doing final algebra...')176 w, os = self.w, _ =self._do_algebra(y)177 #Saving to HDF5178 w_h5 = w_handle.create_dataset('w', data =w)179 os_h5 = os_handle.create_dataset('os', data =os)180 181 ## Calculate error182 self._calculate_error(y)183 print('Done')184 185 defpredict(self, test_data):186 mu = self.mu =self.mu.value187 sigmas = self.sigmas =self.sigmas.value188 w = self.w =self.w.value189 190 print('Calculating phi for test data...')191 phi =self._calculate_phi(test_data)192 os =dot(w, transpose(phi))193 savetxt('iok3834.txt', os, delimiter='')194 returnos195 196 @property197 defsummary(self):198 return ''.join(199 ['-----------------',200 'Training set size: {}'.format(self.n),201 'Hidden layer size: {}'.format(self.neurons),202 '-----------------',203 'Absolute error : {:02.2f}'.format(self.error),204 'Relative error : {:02.2f}%'.format(self.relative_error * 100)])205 206 207 defpredict(test_data):208 mu = File('rbf-mu-21222020年12月05日00.hdf5', 'r')['mu'].value209 sigmas = File('rbf-sigma-21222020年12月05日00.hdf5', 'r')['sigmas'].value210 w = File('rbf-w-21222020年12月05日00.hdf5', 'r')['w'].value211 212 n =test_data.shape[0]213 neur =mu.shape[0]214 215 mu =transpose(mu)216 mu.reshape((n, neur))217 218 phi =zeros((n, neur))219 for i inrange(n):220 for j inrange(neur):221 phi[i, j] =multiQuadric(test_data[i,:], mu[j], sigmas[j])222 223 os =dot(w, transpose(phi))224 savetxt('iok3834.txt', os, delimiter='')225 return os