API Reference

Classes

vae.CISSVAE(*args, **kwargs)	Clustering-Informed Shared-Structure Variational Autoencoder (CISSVAE).
cluster_dataset.ClusterDataset(*args, **kwargs)	Dataset that handles cluster-wise masking and normalization for VAE training.

Training & Tuning

autotune.SearchSpace([num_hidden_layers, ...])	Defines tunable and fixed hyperparameter ranges for the Optuna search.
autotune.autotune(search_space, train_dataset)	Optuna-based hyperparameter search for the CISSVAE model.
run_cissvae.run_cissvae(data[, ...])	End-to-end pipeline for Clustering-Informed Shared-Structure Variational Autoencoder (CISS-VAE).

Utilities & Helpers

clustering.cluster_on_missing(data[, ...])	Cluster samples based on their missingness patterns using KMeans or Leiden.
clustering.cluster_on_missing_prop(...[, ...])	Cluster samples based on their per-feature missingness proportions using KMeans or Leiden.
matrix.create_missingness_prop_matrix(data)	Create a missingness proportion matrix summarizing feature-level missingness per sample.
helpers.plot_vae_architecture(model[, ...])	Plots a horizontal schematic of the VAE architecture, showing shared and cluster-specific layers.
helpers.get_imputed(model, data_loader[, device])	Returns a ClusterDataset where originally missing values have been replaced with model reconstructions.
helpers.get_imputed_df(model, data_loader[, ...])	Given trained model and cluster dataset object, get imputed dataset as pandas DataFrame.