uniport.get_prior

uniport.get_prior(celltype1, celltype2, alpha=2)[source]

Create a prior correspondence matrix according to cell labels

Parameters:

  • celltype1 – cell labels of dataset X

  • celltype2 – cell labels of dataset Y

  • alpha – the confidence of label, ranges from (1, inf). Higher alpha means better confidence. Default: 2.0

Returns:

a prior correspondence matrix between cells

Return type:

torch.tensor