tomodrgn.convergence.plot_latent_umap

plot_latent_umap(workdir: str, outdir: str, plot_format: Literal['png', 'svg'], epochs: ndarray, nptcls: int, subset: int | None = 50000, random_seed: int | RandomState | None = 42, random_state: int | None = 42) → None

Calculates, plots, and saves UMAP embeddings of subset of particles’ selected epochs’ latent encodings

Parameters:

• workdir – path to directory containing tomodrgn training results

• outdir – path to base directory to save outputs

• plot_format – file format with which to save plots

• epochs – array of epochs for which to calculate UMAPs

• nptcls – number of particles in dataset

• subset – size of subset on which to calculate umap, None means all

• random_seed – seed for random selection of subset particles with numpy

• random_state – random state seed used by UMAP for reproducibility at slight cost of performance (None means faster but non-reproducible)

Returns:

None