Plotting

class scarches.plotting.SCVI_EVAL(model: SCVI | SCANVI | TOTALVI, adata: AnnData, trainer: Trainer | None = None, cell_type_key: str = None, batch_key: str = None)[source]

Bases: object

Methods

plot_latent([show, save, dir_path, ...])

if save:

get_asw
get_classification_accuracy
get_ebm
get_f1_score
get_knn_purity
get_latent_score
get_model_arch
get_nmi
latent_as_anndata
plot_history

get_asw()[source]

get_classification_accuracy()[source]

get_ebm(n_neighbors=50, n_pools=50, n_samples_per_pool=100, verbose=True)[source]

get_f1_score()[source]

get_knn_purity(n_neighbors=50, verbose=True)[source]

get_latent_score()[source]

get_model_arch()[source]

get_nmi()[source]

latent_as_anndata()[source]

plot_history(show=True, save=False, dir_path=None)[source]

plot_latent(show=True, save=False, dir_path=None, n_neighbors=8, predictions=False, in_one=False, colors=None)[source]

if save:

if dir_path is None:: name = ‘scanvi_latent.png’
else:: name = f’{dir_path}.png’

else:

name = False

class scarches.plotting.TRVAE_EVAL(model: trVAE | TRVAE, adata: AnnData, trainer: trVAETrainer | None = None, condition_key: str | None = None, cell_type_key: str | None = None)[source]

Bases: object

Methods

get_asw
get_ebm
get_knn_purity
get_latent_score
get_model_arch
get_nmi
latent_as_anndata
plot_history
plot_latent

get_asw()[source]

get_ebm(n_neighbors=50, n_pools=50, n_samples_per_pool=100, verbose=True)[source]

get_knn_purity(n_neighbors=50, verbose=True)[source]

get_latent_score()[source]

get_model_arch()[source]

get_nmi()[source]

latent_as_anndata()[source]

plot_history(show=True, save=False, dir_path=None)[source]

plot_latent(show=True, save=False, dir_path=None, n_neighbors=8)[source]

scarches.plotting.plot_abs_bfs(adata, scores_key='bf_scores', terms: str | list = 'terms', keys=None, n_cols=3, **kwargs)[source]: Plot the absolute bayes scores rankings.

scarches.plotting.sankey_diagram(data, save_path=None, show=False, **kwargs)[source]: Draws Sankey diagram for the given data. :param data: array with 2 columns. One for predictions and another for true values. :type data: ndarray :param save_path: Path to save the drawn Sankey diagram. if None, the diagram will not be saved. :type save_path: str :param show: if True will show the diagram. :type show: bool :param kwargs: additional arguments for diagram configuration. See _alluvial.plot function.