GFLlogit: Coordinate optimization for GFL logistic regression

In ohishim/GFLlogit: Coordinate descent algorithm of generalized fused Lasso logistic regression for multivariate trend filtering

Coordinate optimization for GFL logistic regression

Description

GFLlogit This function calculates estimates for GFL logistic regression via coordinate descent algorithm.

Usage

GFLlogit(
  m,
  y,
  adjCD,
  lambda = NULL,
  lambda.type = NULL,
  weight = NULL,
  conv = "coef",
  progress = FALSE,
  thres = 1e-05
)

Arguments

m	a vector of the numbers of trials
y	a vector of the numbers of successes
adjCD	a list of adjacent information
lambda	tuning parameter; "NULL" or vector/scalar
lambda.type	option when is.null(lambda)=FALSE; value: lambda is searching points; rate: lam.max*lambda is searching points
weight	penalty weights; "NULL" or list
conv	"coef" or "pll"; A convergence judgement is based on "conv"
progress	If TRUE, progress is displayed
thres	a threshold for convergence judgement

Value

mu.hat: vector of estimates

lam.hat: the optimal tuning parameter

pll: the minimum of the objective function

time: runtime (s)

Examples

#GFLlogit(m, y, adjCD)

ohishim/GFLlogit documentation built on Feb. 6, 2023, 9 p.m.