remove commented codes

tests/unit/machine/gp/SingleFITCLaplacianInferenceMethodWithLBFGS_unittest.cc

@@ -91,7 +91,6 @@ TEST(SingleFITCLaplacianInferenceMethodWithLBFGS,get_cholesky)
 	CDenseFeatures<float64_t>* latent_features_train=new CDenseFeatures<float64_t>(lat_feat_train);
 	CBinaryLabels* labels_train=new CBinaryLabels(lab_train);
 
-	// choose Gaussian kernel with sigma = 2 
 	float64_t ell=1.0;
 
 	CLinearARDKernel* kernel=new CGaussianARDFITCKernel(10, 2*ell*ell);
@@ -151,50 +150,6 @@ TEST(SingleFITCLaplacianInferenceMethodWithLBFGS,get_cholesky)
 	abs_tolorance = CMath::get_abs_tolerance(-0.006876868793106, rel_tolorance);
 	EXPECT_NEAR(L(2,2),  -0.006876868793106,  abs_tolorance);
 
-	/*
-	// build parameter dictionary
-	CMap<TParameter*, CSGObject*>* parameter_dictionary=new CMap<TParameter*, CSGObject*>();
-	inf->build_gradient_parameter_dictionary(parameter_dictionary);
-
-	// compute derivatives wrt parameters
-	CMap<TParameter*, SGVector<float64_t> >* gradient=
-		inf->get_negative_log_marginal_likelihood_derivatives(parameter_dictionary);
-
-	// get parameters to compute derivatives
-	//TParameter* width_param=kernel->m_gradient_parameters->get_parameter("width");
-	TParameter* scale_param=inf->m_gradient_parameters->get_parameter("scale");
-	TParameter* mean_param=mean->m_gradient_parameters->get_parameter("mean");
-
-	TParameter* weights_param=kernel->m_gradient_parameters->get_parameter("weights");
-	SGVector<float64_t> dnlz_weights_vec=gradient->get_element(weights_param);
-	SGMatrix<float64_t> dnlz_weights(dnlz_weights_vec.vector,t_dim, dim, false);
-	dnlz_weights(0,0)*=weight1;
-	dnlz_weights(1,1)*=weight4;
-	Map<MatrixXd> out(dnlz_weights.matrix, t_dim, dim);
-	std::cout<<"\nweights="<<out<<std::endl;
-
-
-	float64_t dnlZ_sf2=scale*(gradient->get_element(scale_param))[0];
-	std::cout<<"\nscale="<<dnlZ_sf2<<std::endl;
-
-	TParameter* noise_param=inf->m_gradient_parameters->get_parameter("inducing_noise");
-	float64_t dnlZ_noise=(gradient->get_element(noise_param))[0];
-
-	std::cout<<dnlZ_noise<<std::endl;
-
-
-	TParameter* lat_param=inf->m_gradient_parameters->get_parameter("inducing_features");
-	SGVector<float64_t> tmp=gradient->get_element(lat_param);
-	SGMatrix<float64_t> deriv_lat(tmp.vector, dim, m, false);
-	Map<MatrixXd> eigen_out(deriv_lat.matrix, dim, m);
-	std::cout<<"lat_mat=\n"<<eigen_out<<std::endl;
-
-	// clean up
-	SG_UNREF(gradient);
-	SG_UNREF(parameter_dictionary);
-
-	SG_UNREF(inf);
-	*/
 	SG_UNREF(inf);
 }
 
@@ -244,7 +199,6 @@ TEST(SingleFITCLaplacianInferenceMethodWithLBFGS,get_alpha)
 	CDenseFeatures<float64_t>* latent_features_train=new CDenseFeatures<float64_t>(lat_feat_train);
 	CBinaryLabels* labels_train=new CBinaryLabels(lab_train);
 
-	// choose Gaussian kernel with sigma = 2
 	float64_t ell=1.0;
 
 	CLinearARDKernel* kernel=new CGaussianARDFITCKernel(10, 2*ell*ell);
@@ -340,7 +294,6 @@ TEST(SingleFITCLaplacianInferenceMethodWithLBFGS,get_negative_log_marginal_likel
 	CDenseFeatures<float64_t>* latent_features_train=new CDenseFeatures<float64_t>(lat_feat_train);
 	CBinaryLabels* labels_train=new CBinaryLabels(lab_train);
 
-	// choose Gaussian kernel with sigma = 2 and zero mean function
 	float64_t ell=1.0;
 	float64_t weight1=2.0;
 	float64_t weight2=3.0;
@@ -422,7 +375,6 @@ TEST(SingleFITCLaplacianInferenceMethodWithLBFGS,get_marginal_likelihood_derivat
 	CDenseFeatures<float64_t>* latent_features_train=new CDenseFeatures<float64_t>(lat_feat_train);
 	CBinaryLabels* labels_train=new CBinaryLabels(lab_train);
 
-	// choose Gaussian kernel with sigma = 2 and zero mean function
 	float64_t ell=1.0;
 	float64_t weight1=2.0;
 	float64_t weight2=3.0;