tangram.mapping_optimizer.Mapper

class tangram.mapping_optimizer.Mapper(S, G, d=None, d_source=None, lambda_g1=1.0, lambda_d=0, lambda_g2=0, lambda_r=0, device='cpu', adata_map=None, random_state=None)

Allows instantiating and running the optimizer for Tangram, without filtering. Once instantiated, the optimizer is run with the ‘train’ method, which also returns the mapping result.

Mapper.train(num_epochs[, learning_rate, …])

Run the optimizer and returns the mapping outcome.