alpine.vis package

Subpackages

• alpine.vis.geometry package
  • Submodules
  • alpine.vis.geometry.local_complexity module
    • flatten_model()
    • get_intersections_for_hulls()
    • get_layer_intersections_batched()
    • get_ortho_hull_around_samples()
    • get_ortho_hull_around_samples_w_orig()
    • sample_ortho_random()
    • split_generator()
  • alpine.vis.geometry.partitions module
    • partitions
      • partitions.device
      • partitions.get_partitions_from_inr()
      • partitions.model_flatten_func()
      • partitions.n
      • partitions.r
      • partitions.show_partitions()
    • plot_partitions()
    • show_partitions()
  • Module contents

Submodules

alpine.vis.clustering module

class alpine.vis.clustering.KMeans

Bases: object

KMeans clustering algorithm. This class wraps the sklearn KMeans implementation and provides a simple interface for fitting and predicting clusters.

__init__(n_clusters, random_state=None)

fit(X)

fit_predict(X, new_shape=None)

predict(X, new_shape=None)

alpine.vis.pca module

alpine.vis.pca.check_tensor_dtype(x)

alpine.vis.pca.compute_pca_features(inr_features, num_components, signal_shape, pca_preprocess_whiten=True, min_max_normalize=True, normalization_axis=(-1,), **kwargs)

_summary_

Parameters:

• inr_features (_type_) – _description_

• num_components (_type_) – _description_

• signal_shape (_type_) – _description_

• pca_preprocess_whiten (bool, optional) – _description_. Defaults to True.

• min_max_normalize (bool, optional) – _description_. Defaults to True.

• kwargs – keyword arguments for sklearn.decomposition.PCA

alpine.vis.pca.compute_pca_features_batch(inr_features, num_components, signal_shape, pca_preprocess_whiten=True, min_max_normalize=True, normalization_axis=(-1,), **kwargs)

_summary_

Parameters:

• inr_features (_type_) – _description_

• num_components (_type_) – _description_

• signal_shape (_type_) – _description_

• pca_preprocess_whiten (bool, optional) – _description_. Defaults to True.

• min_max_normalize (bool, optional) – _description_. Defaults to True.

• kwargs – keyword arguments for sklearn.decomposition.PCA

alpine.vis.pca.compute_pca_layerwise(inr_features, num_components, signal_shape, pca_preprocess_whiten=True, min_max_normalize=True, normalization_axis=(-1,), **kwargs)

alpine.vis.pca.normalize_fn(x, axis=None)

alpine.vis.pca.return_as_torch(x)

Module contents