Some more fixes.

section_4.1_compute_visualize.py

@@ -29,7 +29,7 @@ hyperparameter = dic_hyperparams[name_dataset]
 
 ## Run everything
 A_start, A_optimized, model, model_vkoga1, model_vkoga2, data, \
-    array_concatenate, _, dic_timings_2L = run_everything(
+    array_concatenate, _, list_timings_2L = run_everything(
     name_dataset,
     hyperparameter.maxIter_vkoga, hyperparameter.N_points,
     hyperparameter.noise_level, hyperparameter.reg_para_optim, hyperparameter.reg_para_vkoga,
@@ -62,7 +62,7 @@ io.savemat(path_for_results + name_dataset + '.mat',
                 array_eps=array_eps,
                 array_cv_f=array_cv_f,
                 array_cv_f_val=array_cv_f_val,
-                dic_timings_2L = dic_timings_2L,
+                list_timings_2L = list_timings_2L,
                 list_timings_1L = list_timings_1L))
 
 # in Matlab:

section_4.2_compute.py

@@ -18,7 +18,6 @@ np.random.seed(1)
 list_datasets = ['fried', 'sarcos', 'ct', 'diamonds', 'stock', 'kegg_undir_uci', 'online_video',
                  'wecs', 'mlr_knn_rng', 'query_agg_count', 'sgemm', 'road_network']
 
-
 ## Loop over reruns and datasets
 for idx_indices in [0, 1, 2, 3, 4]:
     for idx_dataset, name_dataset in enumerate(list_datasets):
@@ -29,7 +28,7 @@ for idx_indices in [0, 1, 2, 3, 4]:
 
         ## Run everything
         A_start, A_optimized, model, model_vkoga1, model_vkoga2, data, \
-            array_concatenate, array_test_rmse_deep, dic_timings_2L = run_everything(
+            array_concatenate, array_test_rmse_deep, list_timings_2L = run_everything(
             name_dataset,
             hyperparameter.maxIter_vkoga, hyperparameter.N_points,
             hyperparameter.noise_level, hyperparameter.reg_para_optim, hyperparameter.reg_para_vkoga,
@@ -68,7 +67,7 @@ for idx_indices in [0, 1, 2, 3, 4]:
                         array_cv_f_val=array_cv_f_val,
                         array_test_rmse_deep=array_test_rmse_deep,
                         array_test_rmse_cv=array_test_rmse_cv,
-                        dic_timings_2L=dic_timings_2L,
+                        list_timings_2L=list_timings_2L,
                         list_timings_1L=list_timings_1L))
 
         # in Matlab:

utils_code/main_function.py

@@ -29,7 +29,7 @@ def run_everything(name_dataset, maxIter_vkoga, N_points, noise_level, reg_para_
                    flag_vkoga_verbose=False, flag_plots=False, flag_optim_verbose=True,
                    flag_std_vkoga=True, idx_rerun=None):
 
-    dic_timings = {}
+    list_timings = []
 
     ## Load dataset
     dataset = Dataset(N_points=N_points)
@@ -82,7 +82,7 @@ def run_everything(name_dataset, maxIter_vkoga, N_points, noise_level, reg_para_
     print(datetime.now().strftime("%H:%M:%S"), name_dataset, '2layered kernel optimization finished.')
     t_optim_stop = time.time()
 
-    dic_timings['2L_optim'] = t_optim_stop - t_optim_start
+    list_timings.append(t_optim_stop - t_optim_start)
 
     ## Application of VKOGA
     # VKOGA with modified modified kernel, simply pre-apply the linear transformation
@@ -96,7 +96,7 @@ def run_everything(name_dataset, maxIter_vkoga, N_points, noise_level, reg_para_
     print(datetime.now().strftime("%H:%M:%S"), name_dataset, '2layered VKOGA finished.')
     t_vkoga1_stop = time.time()
 
-    dic_timings['2L_vkoga'] = t_vkoga1_stop - t_vkoga1_start
+    list_timings.append(t_vkoga1_stop - t_vkoga1_start)
 
     # VKOGA with standard kernel
     model_vkoga2 = VKOGA(kernel=kernel, greedy_type='f_greedy',
@@ -219,7 +219,7 @@ def run_everything(name_dataset, maxIter_vkoga, N_points, noise_level, reg_para_
     data = [X_train, X_test, y_train, y_test]
 
     return A_start, model.A.detach().numpy(), model, model_vkoga1, model_vkoga2, \
-           data, array_concatenate, array_test_rmse, dic_timings
+           data, array_concatenate, array_test_rmse, list_timings