part_name, part_type, voxel_dim, voxel_channels, point_dim) #Import model architecture output_dimension = assembly_kccs dl_model = Bayes_DLModel(model_type, output_dimension, optimizer, loss_func, regularizer_coeff, output_type) model = dl_model.bayes_cnn_model_3d(voxel_dim, voxel_channels) #Inference from simulated data inference_model = deploy_model.get_model(model, model_path, voxel_dim, voxel_channels) kcc_dataset = get_data.data_import(kcc_files, kcc_folder) input_conv_data, kcc_subset_dump, kpi_subset_dump = get_data.data_convert_voxel_mc( vrm_system, dataset, point_index, kcc_dataset) y_pred = np.zeros_like(kcc_dataset) y_pred, y_std, y_aleatoric_std = deploy_model.model_inference( input_conv_data, inference_model, y_pred, kcc_dataset.values, plots_path) avg_std = np.array(y_std).mean(axis=0) avg_aleatoric_std = np.array(y_std).mean(axis=0) print("Average Epistemic Uncertainty of each KCC: ", avg_std) print("Average Aleatoric Uncertainty of each KCC: ", avg_aleatoric_std) evalerror = 1 if (evalerror == 1): metrics_eval = MetricsEval() eval_metrics, accuracy_metrics_df = metrics_eval.metrics_eval_base(
file_names_y=[file_names_y] file_names_z=[file_names_z] np.savetxt(file_path, initial_samples, delimiter=",") print('Sampling Completed...') test_samples=initial_samples cae_status=cae_simulations.run_simulations(run_id=0,type_flag='test') print("Pre-processing simulated test data") dataset_test=[] dataset_test.append(get_data.data_import(file_names_x,data_folder)) dataset_test.append(get_data.data_import(file_names_y,data_folder)) dataset_test.append(get_data.data_import(file_names_z,data_folder)) input_conv_data_test, kcc_subset_dump_test,kpi_subset_dump_test=get_data.data_convert_voxel_mc(vrm_system,dataset_test,point_index,test_samples) for i in tqdm(range(max_run_length)): run_id=i print('Training Run ID: ',i) file_name=sampling_config['output_file_name_train']+'_'+str(i)+'.csv' file_names_x=sampling_config['datagen_filename_x']+'train'+'_'+str(i)+'.csv' file_names_y=sampling_config['datagen_filename_y']+'train'+'_'+str(i)+'.csv' file_names_z=sampling_config['datagen_filename_z']+'train'+'_'+str(i)+'.csv' file_names_x=[file_names_x] file_names_y=[file_names_y] file_names_z=[file_names_z]
get_data.data_import(test_output_file_names_y, data_folder)) test_output_dataset.append( get_data.data_import(test_output_file_names_z, data_folder)) kcc_dataset = get_data.data_import(kcc_files, kcc_folder) test_kcc_dataset = get_data.data_import(test_kcc_files, kcc_folder) if (kcc_sublist != 0): print("Sub-setting Process Parameters: ", kcc_sublist) kcc_dataset = kcc_dataset.iloc[:, kcc_sublist] test_kcc_dataset = test_kcc_dataset[:, kcc_sublist] else: print("Using all Process Parameters") #Pre-processing to point cloud data input_conv_data, kcc_subset_dump, kpi_subset_dump = get_data.data_convert_voxel_mc( vrm_system, input_dataset, point_index, kcc_dataset) test_input_conv_data, test_kcc_subset_dump, test_kpi_subset_dump = get_data.data_convert_voxel_mc( vrm_system, test_input_dataset, point_index, test_kcc_dataset) output_conv_data, kcc_subset_dump, kpi_subset_dump = get_data.data_convert_voxel_mc( vrm_system, output_dataset, point_index, kcc_dataset) test_output_conv_data, test_kcc_subset_dump, test_kpi_subset_dump = get_data.data_convert_voxel_mc( vrm_system, test_output_dataset, point_index, test_kcc_dataset) unet_train_model = Unet_TrainModel(batch_size, epocs, split_ratio) trained_model, eval_metrics, accuracy_metrics_df = unet_train_model.unet_run_train_model( model, input_conv_data, kcc_subset_dump, output_conv_data, test_input_conv_data, test_kcc_subset_dump, test_output_conv_data, model_path, logs_path, plots_path, activate_tensorboard)
file_names_z = sampling_config[ 'datagen_filename_z'] + 'test' + '_' + str(0) + '.csv' np.savetxt(file_path, test_samples, delimiter=",") print('Sampling Completed...') cae_status = cae_simulations.run_simulations(run_id=0, type_flag='test') print("Pre-processing simulated test data") dataset_test = [] dataset_test.append(get_data.data_import([file_names_x], data_folder)) dataset_test.append(get_data.data_import([file_names_y], data_folder)) dataset_test.append(get_data.data_import([file_names_z], data_folder)) input_conv_data_test, kcc_subset_dump_test, kpi_subset_dump_test = get_data.data_convert_voxel_mc( vrm_system, dataset_test, point_index, test_samples) if (sampling_validation_flag == 1): print('Generating Adaptive Sampling Data...') print('LHS Sampling for validation samples') #get prediction errors #get uncertainty estimates from cae_simulations import CAESimulations cae_simulations = CAESimulations(simulation_platform, simulation_engine, max_run_length, case_study) validate_samples = adaptive_sampling.inital_sampling_uniform_random( kcc_struct, sampling_config['sample_validation_dim']) file_name = sampling_config['output_file_name_validate'] + ".csv"