Skip to contents

Run DGRDL with the specified parameters

dgrdl_result.Rd

Runs the DGRDL algorithm from Pan et al., with the specified hyperparamters. To determine the hyperparameters, you can use dgrdl_grid_search().

Usage

dgrdl_result(
  object,
  dgrdl_params = params_dgrdl(),
  seed = 42L,
  .verbose = TRUE
)

Arguments

object

The class, see BulkCoExp(). Ideally, you should run preprocess_bulk_coexp() before applying this function.

dgrdl_params

List. Output of params_dgrdl():

  • sparsity - Integer. Sparsity constraint (max non-zero coefficients per signal)

  • dict size - Integer. The dictionary size.

  • alpha - Float. Sample context regularisation weight.

  • beta - Float. Feature effect regularisation weight.

  • max_iter - Integer. Maximum number of iterations for the main algorithm.

  • k_neighbours - Integer. Number of neighbours for the KNN graph for the feature and sample Laplacian.

  • admm_iter - Integer. ADMM iterations for sparse coding.

  • rho - Float. ADMM step size.

seed

Integer. Seed for the initialisation of the dictionary.

.verbose

Boolean. Controls verbosity of the function.

References

Pan et al., Cell Syst, 2022