Fit the MFAI model using greedy algorithm.

Usage

fitGreedy(
  object,
  K_max = NULL,
  learning_rate = 0.1,
  minsplit = 10,
  minbucket = round(minsplit/3),
  maxdepth = 2,
  other_tree_para = list(),
  null_check = TRUE,
  tol_snr = 0.002,
  verbose_greedy = TRUE,
  save_init = FALSE,
  sf_para = list()
)

Arguments

object: MFAIR object.
K_max: Integer. The maximum rank allowed in the MFAI model.
learning_rate: Numeric. Parameter for the gradient boosting part.
minsplit: Integer. Parameter for the gradient boosting part.
minbucket: Integer. Parameter for the gradient boosting part.
maxdepth: Integer. Parameter for the gradient boosting part.
other_tree_para: A list containing other parameters for the gradient boosting part. See rpart::rpart.control() for details.
null_check: Logical. If TRUE, then mfair will check whether the currently inferred factor is close to zero compared to the estimated noise strength. If the check is performed and successful, the factor will be deleted in the returned fit, and the greedy algorithm will be terminated.
tol_snr: Numeric. The convergence criterion which determine the inferred rank of data.
verbose_greedy: Logical. Whether to display the detailed information when fitting the model.
save_init: Logical. Whether to save the initialization of the model.
sf_para: A list containing parameters for fitting the single factor MFAI model. See fitSFFully(), fitSFMissing(), or fitSFSparse() for details.

Value

An MFAIR object containing the information about the fitted MFAI model using greedy algorithm.