MCPcopy Index your code

hub / github.com/MengyuanChen21/ICLR2024-REDL / functions

Functions127 in github.com/MengyuanChen21/ICLR2024-REDL

⨍ Functions 127◇ Types & classes 13
↓ 12 callersMethodsvhn
Returns the SVHN dataset, either for training or testing: * Task: classification * Features: [3, 32, 32] * Classes: 1
code_classical/dataset.py:433
↓ 8 callersMethodcifar10
Returns the CIFAR-10 dataset, either for training or testing: * Task: classification * Features: [3, 32, 32] * Classe
code_classical/dataset.py:405
↓ 8 callersMethodcifar100
Returns the CIFAR-100 dataset, either for training or testing: * Task: classification * Features: [3, 32, 32] * Class
code_classical/dataset.py:419
↓ 8 callersFunctionlogger
(*args, **kwargs)
code_fsl/utils/io_utils.py:157
↓ 8 callersFunctionmake_layers
(cfg, batch_norm=False, k_lipschitz=None)
code_classical/architectures/vgg_sequential.py:61
↓ 4 callersMethodGenerateRun
(self, iRun, cfg, regenRState=False, generate=True)
code_fsl/FSLTask.py:83
↓ 4 callersMethod_make_layer
(self, block, planes, num_blocks, stride)
code_classical/architectures/resnet_sequential.py:54
↓ 4 callersFunctionanomaly_detection
(alpha, ood_alpha, score_type='AUROC', uncertainty_type='aleatoric', lamb1=1.0, lamb2=1.0)
code_fsl/metrics.py:81
↓ 4 callersFunctionconfidence
(Y, alpha, score_type='AUROC', uncertainty_type='aleatoric', lamb1=1.0, lamb2=1.0)
code_fsl/metrics.py:21
↓ 4 callersFunctiondiff_entropy
(alpha, ood_alpha, score_type='AUROC', lamb2=1.0)
code_fsl/metrics.py:141
↓ 4 callersFunctiondist_uncertainty
(alpha, ood_alpha, score_type='AUROC')
code_fsl/metrics.py:182
↓ 4 callersMethoddump
(self)
code_fsl/utils/io_utils.py:90
↓ 4 callersMethodfake
Returns the SVHN dataset, either for training or testing: * Task: classification * Features: [3, 32, 32] * Classes: 1
code_classical/dataset.py:461
↓ 4 callersMethodfashion_mnist
Returns the Fashion-MNIST dataset, either for training or testing: * Task: classification * Features: [1, 28, 28] * C
code_classical/dataset.py:391
↓ 4 callersMethodkmnist
Returns the KMNIST dataset, either for training or testing: * Task: classification * Features: [1, 28, 28] * Classes:
code_classical/dataset.py:377
↓ 4 callersMethodmnist
Returns the MNIST dataset, either for training or testing: * Task: classification * Features: [1, 28, 28] * Classes:
code_classical/dataset.py:363
↓ 4 callersMethodsegment_sky_only
Returns the segment dataset with class sky. * Task: classification * Features: [18] * Samples: 330
code_classical/dataset.py:291
↓ 4 callersMethodsegment_window_only
Returns the segment dataset with class window. * Task: classification * Features: [18] * Samples: 330
code_classical/dataset.py:274
↓ 4 callersMethodsegment_window_sky_missing
Returns the segment dataset, either for training or testing. * Task: classification * Features: [18] * Samples: 1650
code_classical/dataset.py:254
↓ 4 callersMethodsensorless_drive_10_11_only
Returns the segment dataset wiht class 10 and 11. * Task: classification * Features: [48] * Samples: 10638
code_classical/dataset.py:345
↓ 4 callersMethodsensorless_drive_9_10_11_missing
Returns the segment dataset, either for training or testing. * Task: classification * Features: [48] * Samples: 42552
code_classical/dataset.py:308
↓ 4 callersMethodsensorless_drive_9_only
Returns the segment dataset with class 9. * Task: classification * Features: [48] * Samples: 5319
code_classical/dataset.py:328
↓ 4 callersMethodsetRandomStates
(self, cfg, cache_dir)
code_fsl/FSLTask.py:105
↓ 4 callersMethodstep
(self)
code_classical/models/ModifiedEvidentialN.py:142
↓ 3 callersMethodGenerateRunSet
(self, start=None, end=None, cfg=None, cache_dir=None)
code_fsl/FSLTask.py:125
↓ 3 callersFunctioncompute_loss_accuracy
(model, loader, epoch, device=torch.device("cpu"), is_fisher=False)
code_classical/train.py:5
↓ 3 callersFunctioncompute_output
(model, inputs, act_type='exp')
code_fsl/evaluation.py:12
↓ 3 callersFunctioncompute_output
(model, inputs, act_type, lamb1, lamb2)
code_fsl/our_evaluation.py:13
↓ 3 callersFunctiondirichlet_kl_divergence
(alphas, target_alphas)
code_fsl/train.py:45
↓ 3 callersMethoddump
(self)
code_classical/utils/io_utils.py:90
↓ 3 callersFunctiongen_table
(summary_df, row_tree, col_tree, y_col, y_col_ci=None, str_maker=None)
code_fsl/utils/summ_utils.py:44
↓ 3 callersFunctionget_dataset
(dataset_name, batch_size, split=[.8, .2], seed=1, test_shuffle_seed=None, batch_size_eval=1024,
code_classical/dataset.py:109
↓ 3 callersMethodloadDataSet
(self, dsname, features_dir)
code_fsl/FSLTask.py:53
↓ 2 callersFunctionaccuracy
(Y, alpha)
code_fsl/metrics.py:6
↓ 2 callersFunctionaccuracy
(Y, alpha)
code_classical/utils/metrics.py:40
↓ 2 callersFunctionclosure
()
code_fsl/train.py:109
↓ 2 callersFunctioncompute_X_Y_alpha
(model, loader, device, noise_epsilon=0.0)
code_classical/utils/metrics.py:20
↓ 2 callersFunctioncompute_fisher_loss
(labels_1hot_, evi_alp_)
code_fsl/train.py:11
↓ 2 callersFunctionconvert_before_after_delta_to_3_rows
There are 3 accuracy columns that we care about: 1. 'base_test_acc' otherwise known as 'Before' 2. 'test_acc' otherwise known as 'Aft
code_fsl/utils/summ2tbls.py:13
↓ 2 callersMethodis_attachment
(self, var)
code_fsl/utils/io_utils.py:84
↓ 2 callersMethodis_attachment
(self, var)
code_classical/utils/io_utils.py:84
↓ 2 callersFunctionlinear_sequential
(input_dims, hidden_dims, output_dim, k_lipschitz=None, p_drop=None)
code_classical/architectures/linear_sequential.py:6
↓ 2 callersMethodrandom_noise_image_dataset
(cls, num_classes, num_images_per_class, mean=0, sigma=1, dims=(1, 28, 28),
code_classical/dataset.py:473
↓ 2 callersMethodreset_global_vars
(self)
code_fsl/FSLTask.py:22
↓ 1 callersMethod__init__
(self, block, num_blocks, output_dim=10)
code_classical/architectures/resnet_sequential.py:41
↓ 1 callersMethod_load_pickle
(self, file)
code_fsl/FSLTask.py:40
↓ 1 callersMethodadd
(self, row_dict, file_path)
code_fsl/utils/io_utils.py:56
↓ 1 callersMethodadd
(self, row_dict, file_path)
code_classical/utils/io_utils.py:56
↓ 1 callersFunctionadd_missing_dflt_vals
(df, def_dict)
code_fsl/utils/summ_utils.py:26
↓ 1 callersFunctionanomaly_detection
(alpha, ood_alpha, uncertainty_type='max_prob', save_path=None, return_scores=False)
code_classical/utils/metrics.py:179
↓ 1 callersFunctioncompute_kl
(alphas, target_concentration=1.0)
code_fsl/train.py:204
↓ 1 callersFunctioncompute_kl_loss
(alphas, labels=None, target_concentration=1.0, concentration=1.0, reverse=True)
code_fsl/train.py:29
↓ 1 callersMethodcompute_kl_loss
(self, alphas, target_concentration, epsilon=1e-8)
code_classical/models/ModifiedEvidentialN.py:122
↓ 1 callersFunctioncompute_mse
(labels_1hot, alpha, lamb1, lamb2)
code_fsl/train.py:180
↓ 1 callersMethodcompute_mse
(self, labels_1hot, evidence)
code_classical/models/ModifiedEvidentialN.py:112
↓ 1 callersFunctionconfidence
(Y, alpha, uncertainty_type='max_prob', save_path=None, return_scores=False)
code_classical/utils/metrics.py:47
↓ 1 callersFunctionconvolution_linear_sequential
(input_dims, linear_hidden_dims, conv_hidden_dims, output_dim, kernel_dim, k_lipschitz=None, p_drop=None)
code_classical/architectures/convolution_linear_sequential.py:30
↓ 1 callersFunctionconvolution_sequential
(input_dims, hidden_dims, output_dim, kernel_dim, k_lipschitz=None, p_drop=None)
code_classical/architectures/convolution_sequential.py:7
↓ 1 callersFunctionget_csvh5files
(fldr_mini, results_dir)
code_fsl/utils/summ_utils.py:79
↓ 1 callersFunctionget_dataset_info
(dataset_name)
code_classical/dataset.py:15
↓ 1 callersFunctionload_model
(directory_model, name_model, model_type, batch_size_eval=1024)
code_classical/models/model_loader.py:10
↓ 1 callersFunctionmain
(config_dict)
code_fsl/main.py:21
↓ 1 callersFunctionmain
(config_dict)
code_classical/main.py:29
↓ 1 callersFunctionour_anomaly_detection
(alpha, ood_alpha, uncertainty_type='max_prob', save_path=None, return_scores=False, lamb1=1.0, lamb2=1.0)
code_classical/utils/metrics.py:263
↓ 1 callersFunctionour_confidence
(Y, alpha, uncertainty_type='max_prob', save_path=None, return_scores=False, lamb1=1.0, lamb2=1.0)
code_classical/utils/metrics.py:105
↓ 1 callersFunctionour_test_misclassication
(model, act_type, id_x, id_y, lamb1, lamb2)
code_fsl/our_evaluation.py:24
↓ 1 callersFunctionour_test_ood_uncertainty
(model, act_type, id_x, ood_x, ood_y, lamb1, lamb2)
code_fsl/our_evaluation.py:43
↓ 1 callersMethodpredict
(self, p)
code_classical/models/ModifiedEvidentialN.py:147
↓ 1 callersFunctionread_csvh5
(filename)
code_fsl/utils/summ_utils.py:88
↓ 1 callersMethodset_path
(self, file_path)
code_fsl/utils/io_utils.py:40
↓ 1 callersMethodset_path
(self, file_path)
code_classical/utils/io_utils.py:40
↓ 1 callersFunctionstr_maker
(flt_mean, flt_ci=None, is_bold=False)
code_fsl/utils/summ2tbls.py:58
↓ 1 callersFunctionsummarizer
(df, prop_cols=None, y_col=None, reg_column=None, cond=None)
code_fsl/utils/summ_utils.py:125
↓ 1 callersFunctiontest_misclassication
(model, act_type, id_x, id_y)
code_fsl/evaluation.py:30
↓ 1 callersFunctiontest_ood_uncertainty
(model, act_type, id_x, ood_x, ood_y)
code_fsl/evaluation.py:48
↓ 1 callersFunctiontrain
(model, train_loader, val_loader, max_epochs=200, frequency=2, patience=5, model_path='saved_model',
code_classical/train.py:55
↓ 1 callersFunctiontrain_iedl
(X, Y, loss_type='EDL', act_type='softplus', fisher_c=0.0, kl_c=-1.0, target_c=1.0, max_iter=10
code_fsl/train.py:66
↓ 1 callersFunctiontrain_medl
(X, Y, loss_type='EDL', act_type='softplus', lamb1=1.0, lamb2=1.0, fisher_c=0.0, kl_c=-1.0, ta
code_fsl/train.py:210
↓ 1 callersFunctionvgg16_bn
VGG 16-layer model (configuration "D") with batch normalization
code_classical/architectures/vgg_sequential.py:139
MethodClassesInRun
(self, iRun, cfg)
code_fsl/FSLTask.py:100
Method__init__
(self)
code_fsl/FSLTask.py:9
Method__init__
(self, exp_bs, in_dim, out_dim, device, tch_dtype, init=False)
code_fsl/classifier.py:7
Method__init__
(self, dump_period=10)
code_fsl/utils/io_utils.py:28
Method__init__
(self, dump_period=10)
code_classical/utils/io_utils.py:28
Method__init__
(self, in_planes, planes, stride=1)
code_classical/architectures/resnet_sequential.py:15
Method__init__
(self, features, output_dim, k_lipschitz=None, p_drop=None)
code_classical/architectures/vgg_sequential.py:23
Method__init__
(self, output_dim=10)
code_classical/architectures/alexnet_sequential.py:7
Method__init__
(self, input_dim, output_dim, kernel_dim, padding, k_lipschitz=1.0)
code_classical/architectures/SpectralConv.py:7
Method__init__
(self, input_dim, output_dim, k_lipschitz=1.0)
code_classical/architectures/SpectralLinear.py:7
Method__init__
(self, input_dims, linear_hidden_dims, conv_hidden_dims, output_dim, kernel_dim, k_lipschitz, p_drop)
code_classical/architectures/convolution_linear_sequential.py:6
Method__init__
(self, input_dims, # Input dimension. list of ints output_dim, # Output di
code_classical/models/ModifiedEvidentialN.py:12
Functionalexnet
(output_dim, **kwargs)
code_classical/architectures/alexnet_sequential.py:41
Functionappend_to_tar
(tar_path, file_name, file_like_obj)
code_fsl/utils/io_utils.py:145
Functionappend_to_tar
(tar_path, file_name, file_like_obj)
code_classical/utils/io_utils.py:145
Functionbrier_score
(Y, alpha)
code_fsl/metrics.py:11
Functionbrier_score
(Y, alpha)
code_classical/utils/metrics.py:168
Methoddata_len
(self)
code_fsl/utils/io_utils.py:34
Methoddata_len
(self)
code_classical/utils/io_utils.py:34
Functiondefault_str_maker
(flt_mean, flt_ci=None)
code_fsl/utils/summ_utils.py:35
Functiondict_hash
MD5 hash of a dictionary.
code_fsl/utils/io_utils.py:17
next →1–100 of 127, ranked by callers