mediation_analysis: Apply mediation analysis for non-rare binary outcome with two...

In roqe/BY2M: Mediation analysis for a binary non-rare outcome with Sequentially Ordered Multiple Mediators

Apply mediation analysis for non-rare binary outcome with two continuous mediators

Description

Apply mediation analysis for non-rare binary outcome with two continuous mediators

Usage

mediation_analysis(
  dt,
  cnfd = c(),
  dt2 = NULL,
  cnfd2 = c(),
  dt3 = NULL,
  cnfd3 = c(),
  nb = 0,
  intv = 4,
  unit = 1,
  reNAME = NULL,
  grpID = NULL,
  mc = 5,
  autoR = T,
  stra = 1,
  seed = 217
)

Arguments

dt	Input data.
cnfd	Confounder values.
nb	Number of bootstrapping. Default is 0 (no bootstrapping applied).
intv	Number of intervention, 3 or 4. Default is 3.

Examples

para=c(rep(-0.5,9),1,1)
dat1=sim_mediation_data(0.5,1000,para) #binary exposure
apply(dat1,2,mean) #the proportion of Y should not be too skew (nearly 0 or 1)
res1_3=mediation_analysis(dat1)
tru1_3=calc_true_value(para)
dat2=sim_mediation_data(c(0,1),1000,para) #continous exposure
res2_4=mediation_analysis(dat2, intv=4, nb=500)
tru2_4=calc_true_value(para, intv=4)
summary(eIVF) #subset eIVF data for demo
res3_3=mediation_analysis(eIVF, cnfd=c(age=log(36),bmi=log(26)))
res3_4=mediation_analysis(eIVF, cnfd=c(age=log(36),bmi=log(26)), intv=4)
para_sg=c(rep(-0.5,3),rep(0,4),rep(-0.5,2),1,1) # for single mediator, set exposure-related parameters into 0
dat_sg=sim_mediation_data(0.5,1000,para_sg)
dat_sg$S=0 # set the second mediator into 0
res_sg_3=mediation_analysis(dat_sg)

roqe/BY2M documentation built on March 24, 2024, 2:48 a.m.