def run(self,burn_in=None,thinning=None,minimum_TN=None): folds_test = mask.compute_folds(self.I,self.J,self.folds,self.M) folds_training = mask.compute_Ms(folds_test) for i,(train,test) in enumerate(zip(folds_training,folds_test)): print "Fold %s." % (i+1) # Run the line search line_search = LineSearch( classifier=self.classifier, values_K=self.values_K, R=self.R, M=self.M, priors=self.priors, initUV=self.init_UV, iterations=self.iterations, restarts=self.restarts) line_search.search(burn_in=burn_in,thinning=thinning,minimum_TN=minimum_TN) # Store the model fits, and find the best one according to the metric all_performances = line_search.all_values(metric=self.quality_metric) self.fout.write("All model fits for fold %s, metric %s: %s.\n" % (i+1,self.quality_metric,all_performances)) self.fout.flush() best_K = line_search.best_value(metric=self.quality_metric) self.fout.write("Best K for fold %s: %s.\n" % (i+1,best_K)) # Train a model with this K and measure performance on the test set performance = self.run_model(train,test,best_K,burn_in=burn_in,thinning=thinning,minimum_TN=minimum_TN) self.fout.write("Performance: %s.\n\n" % performance) self.fout.flush()
def test_search(): # Check whether we get no exceptions... I,J = 10,9 values_K = [1,2,4,5] R = 2*numpy.ones((I,J)) R[0,0] = 1 M = numpy.ones((I,J)) priors = { 'alpha':3, 'beta':4, 'lambdaU':5, 'lambdaV':6 } initUV = 'exp' iterations = 1 linesearch = LineSearch(classifier,values_K,R,M,priors,initUV,iterations) linesearch.search()
def test_search(): # Check whether we get no exceptions... I, J = 10, 9 values_K = [1, 2, 4, 5] R = 2 * numpy.ones((I, J)) R[0, 0] = 1 M = numpy.ones((I, J)) priors = {'alpha': 3, 'beta': 4, 'lambdaU': 5, 'lambdaV': 6} initUV = 'exp' iterations = 1 linesearch = LineSearch(classifier, values_K, R, M, priors, initUV, iterations) linesearch.search()
# Generate data (_, _, _, _, R) = generate_dataset(I, J, true_K, lambdaU, lambdaV, tau) M = numpy.ones((I, J)) #M = try_generate_M(I,J,fraction_unknown,attempts_M) # Run the line search. The priors lambdaU and lambdaV need to be a single value (recall K is unknown) priors = { 'alpha': alpha, 'beta': beta, 'lambdaU': lambdaU[0, 0] / 10, 'lambdaV': lambdaV[0, 0] / 10 } line_search = LineSearch(classifier, values_K, R, M, priors, initUV, iterations, restarts) line_search.search() # Plot the performances of all three metrics metrics = ['loglikelihood', 'BIC', 'AIC', 'MSE', 'ELBO'] for metric in metrics: plt.figure() plt.plot(values_K, line_search.all_values(metric), label=metric) plt.legend(loc=3) # Also print out all values in a dictionary all_values = {} for metric in metrics: all_values[metric] = line_search.all_values(metric) print "all_values = %s" % all_values '''
tau = alpha / beta lambdaU = numpy.ones((I,true_K)) lambdaV = numpy.ones((J,true_K)) classifier = bnmf_vb_optimised initUV = 'random' # Generate data (_,_,_,_,R) = generate_dataset(I,J,true_K,lambdaU,lambdaV,tau) M = numpy.ones((I,J)) #M = try_generate_M(I,J,fraction_unknown,attempts_M) # Run the line search. The priors lambdaU and lambdaV need to be a single value (recall K is unknown) priors = { 'alpha':alpha, 'beta':beta, 'lambdaU':lambdaU[0,0]/10, 'lambdaV':lambdaV[0,0]/10 } line_search = LineSearch(classifier,values_K,R,M,priors,initUV,iterations,restarts) line_search.search() # Plot the performances of all three metrics metrics = ['loglikelihood', 'BIC', 'AIC', 'MSE', 'ELBO'] for metric in metrics: plt.figure() plt.plot(values_K, line_search.all_values(metric), label=metric) plt.legend(loc=3) # Also print out all values in a dictionary all_values = {} for metric in metrics: all_values[metric] = line_search.all_values(metric) print "all_values = %s" % all_values
tau = alpha / beta lambdaU = numpy.ones((I,true_K)) lambdaV = numpy.ones((J,true_K)) classifier = bnmf_gibbs_optimised initUV = 'random' # Generate data (_,_,_,_,R) = generate_dataset(I,J,true_K,lambdaU,lambdaV,tau) M = numpy.ones((I,J)) #M = try_generate_M(I,J,fraction_unknown,attempts_M) # Run the line search. The priors lambdaU and lambdaV need to be a single value (recall K is unknown) priors = { 'alpha':alpha, 'beta':beta, 'lambdaU':lambdaU[0,0]/10, 'lambdaV':lambdaV[0,0]/10 } line_search = LineSearch(classifier,values_K,R,M,priors,initUV,iterations,restarts) line_search.search(burn_in,thinning) # Plot the performances of all three metrics - but MSE separately metrics = ['loglikelihood', 'BIC', 'AIC', 'MSE'] for metric in metrics: plt.figure() plt.plot(values_K, line_search.all_values(metric), label=metric) plt.legend(loc=3) # Also print out all values in a dictionary all_values = {} for metric in metrics: all_values[metric] = line_search.all_values(metric) print "all_values = %s" % all_values