addKnot.binary: addKnot.binary

In kreidles/informativeDropout: informativeDropout: varying coefficient models to account for informative dropout

Description

For binary outcomes, add a new knot and use a Metropolis-Hastings step to determine if we keep the changes to the model.

Usage

addKnot.binary(model.options, knots.previous, knots.positions.candidate,
  outcomes, times.dropout, times.observation, covariates, X.previous,
  Theta.previous, Z, alpha, betaCovariates, eta.null)

Arguments

model.options	model options
knots.previous	previous set of knots
outcomes	vector of outcomes
times.dropout	vector of dropout times
times.observation	vector of observation times
covariates	data frame of covariates
X.previous	previous X matrix
Theta.previous	previous Theta values
Z	random effects design matrix
alpha	random effects
betaCovariates	regression coefficients for covariates

Value

list containing updated X, knots, Theta, and boolean indicating if change accepted

kreidles/informativeDropout documentation built on Sept. 13, 2020, 12:15 a.m.