Analysis Tools

common.utils.analysis.collect_feature(data_loader, feature_extractor, device, max_num_features=None)

Fetch data from data_loader, and then use feature_extractor to collect features

Parameters

• data_loader (torch.utils.data.DataLoader) – Data loader.

• feature_extractor (torch.nn.Module) – A feature extractor.

• device (torch.device) –

• max_num_features (int) – The max number of features to return

Returns

Features in shape (min(len(data_loader), max_num_features * mini-batch size), \(|\mathcal{F}|\)).

common.utils.analysis.a_distance.calculate(source_feature, target_feature, device, progress=True, training_epochs=10)

Calculate the \(\mathcal{A}\)-distance, which is a measure for distribution discrepancy.

The definition is \(dist_\mathcal{A} = 2 (1-2\epsilon)\), where \(\epsilon\) is the test error of a classifier trained to discriminate the source from the target.

Parameters

• source_feature (tensor) – features from source domain in shape \((minibatch, F)\)

• target_feature (tensor) – features from target domain in shape \((minibatch, F)\)

• device (torch.device) –

• progress (bool) – if True, displays a the progress of training A-Net

• training_epochs (int) – the number of epochs when training the classifier

Returns

\(\mathcal{A}\)-distance

common.utils.analysis.tsne.visualize(source_feature, target_feature, filename, source_color='r', target_color='b')

Visualize features from different domains using t-SNE.

Parameters

• source_feature (tensor) – features from source domain in shape \((minibatch, F)\)

• target_feature (tensor) – features from target domain in shape \((minibatch, F)\)

• filename (str) – the file name to save t-SNE

• source_color (str) – the color of the source features. Default: ‘r’

• target_color (str) – the color of the target features. Default: ‘b’