nbfar_control: Control parameters for NBFAR and NBRRR

In nbfar: Negative Binomial Factor Regression Models ('nbfar')

Control parameters for NBFAR and NBRRR

Description

Default value for a list of control parameters that are used to estimate the parameters of negative binomial co-sparse factor regression (NBFAR) and negative binomial reduced rank regression (NBRRR).

Usage

nbfar_control(
  maxit = 5000,
  epsilon = 1e-07,
  elnetAlpha = 0.95,
  gamma0 = 1,
  spU = 0.5,
  spV = 0.5,
  lamMaxFac = 1,
  lamMinFac = 1e-06,
  initmaxit = 10000,
  initepsilon = 1e-08,
  objI = 0
)

Arguments

maxit	maximum iteration for each sequential steps
epsilon	tolerance value required for convergence of inner loop in GCURE
elnetAlpha	elastic net penalty parameter
gamma0	power parameter for generating the adaptive weights
spU	maximum proportion of nonzero elements in each column of U
spV	maximum proportion of nonzero elements in each column of V
lamMaxFac	a multiplier of the computed maximum value (lambda_max) of the tuning parameter
lamMinFac	a multiplier to determine lambda_min as a fraction of lambda_max
initmaxit	maximum iteration for minimizing the objective function while computing the initial estimates of the model parameter
initepsilon	tolerance value required for the convergence of the objective function while computing the initial estimates of the model parameter
objI	1 or 0 to indicate that the convergence will be on the basis of objective function or not

Value

a list of controlling parameter.

References

Mishra, A., Müller, C. (2022) Negative binomial factor regression models with application to microbiome data analysis. https://doi.org/10.1101/2021.11.29.470304

Examples

control <- nbfar_control()

nbfar documentation built on March 18, 2022, 7:31 p.m.