xchrom.tr.XChrom_model

xchrom.tr.XChrom_model(n_cells: int, cell_vec: int = 32, seq_len: int = 1344, show_summary: bool = True, noise_factor: float = 0.01)[source]

XChrom model construction

Parameters:
  • n_cells (int) – number of cells

  • cell_vec (int) – dimension of cell embeddings, default is 32 with rna.obsm[‘X_pca’], it is consistent with the final dimension of the peak embeddings. if you want to use other cell embeddings, you can set it to the dimension of the cell/peak embeddings.

  • seq_len (int) – length of the sequence, default is 1344

  • show_summary (bool) – whether to show the summary of the model, default is True

  • noise_factor (float) – noise factor for the cell embeddings, default is 0.01