estimateHazards: Conditional Hazard estimation in survtmle package From...

In wilsoncai1992/onestep.survival: One-step TMLE for survival analysis

Description

Conditional Hazard estimation in survtmle package From survtmle (David Benkeser, 2016)

Usage

estimateHazards(dataList, J, verbose, strata = NULL, adjustVars,
  SLlibrary.event, glmFormula.event, superLearnerSummary, bounds)

Arguments

dataList	output of makeDataList
J	same as above
verbose	print messages as SL runs?
strata	would leave as NULL, dont remember exactly how it is coded
adjustVars	data.frame of predictors for SL, dont know why I have it coded this way... just put in e.g. dat[,c("W1","W2")]. by default it will add the Z column and t column in dataList[[1]] when it fits the SL
SLlibrary.event	Super Learner library
glmFormula.event	a glm formula if you don't want to use SL
superLearnerSummary	doesn't do anything, just leave blank
bounds	NULL

Value

added column for the initial data.frame QjHaz <- result, for each subject, h_t from time 0 to T-1; organized in T_i rows QjPseudoHaz <- returns list of length = length(dataList) with added columns name QjPseudoHaz and QjHaz for all j in J. Ignore the PseudoHaz variables – they're used when there are multiple event types; if there's only one type of event they'll == Haz values anyway

Examples

# TO DO

wilsoncai1992/onestep.survival documentation built on May 29, 2019, 11:58 a.m.