scvi_criticism.PPCPlot#

class scvi_criticism.PPCPlot(ppc)#

Plotting utilities for posterior predictive checks

Parameters:

ppc (PPC) – An instance of the PPC class containing the computed metrics

Methods table#

plot_cv(model_name[, cell_wise, plt_type])

Plot coefficient of variation metrics results.

plot_diff_exp(model_name, plot_kind[, ...])

Plot differential expression results.

Methods#

PPCPlot.plot_cv(model_name, cell_wise=True, plt_type='hist2d')#

Plot coefficient of variation metrics results.

See our tutorials for a demonstration of the generated plot along with detailed explanations.

Parameters:

  • model_name (str) – Name of the model

  • cell_wise (bool (default: True)) – Whether to plot the cell-wise or gene-wise metric

  • plt_type (Literal['scatter', 'hist2d'] (default: 'hist2d')) – The type of plot to generate.

PPCPlot.plot_diff_exp(model_name, plot_kind, figure_size=None)#

Plot differential expression results.

Parameters:

  • model_name (str) – Name of the model

  • var_gene_names_col – Column name in the adata.var attribute containing the gene names, if different from adata.var_names

  • figure_size (default: None) – Size of the figure to plot. If None, we will use a heuristic to determine the figure size.