Time-varying mediation analysis using g-computation
Usage
tvmedg(
data,
id,
basec,
expo,
med,
tvar,
outc,
time,
lag = 2,
norev = NULL,
cont_exp = FALSE,
cont_exp_std = FALSE,
tvar_to_med = FALSE,
mreg = NULL,
lreg = NULL,
yreg = NULL,
sp_list = NULL,
sp_type = NULL,
sp_df = NULL,
followup = NULL,
seed = 0,
montecarlo = 1000,
boot = FALSE,
nboot = 1,
ci = 0.95,
parallel = FALSE
)Arguments
- data
input data
- id
patient id
- basec
time-fixed variables
- expo
exposure variable
- med
mediator variable
- tvar
time-varying variable
- outc
outcome variable
- time
time variable
- lag
number of lag
- norev
non-reversible variable (among expo,med,tvar)
- cont_exp
continuous exposure
- cont_exp_std
standardize the continuous exposure
- tvar_to_med
time-varying varible to mediator
- mreg
regression model for mediator
- lreg
regression model for time-varying variable
- yreg
regression model for outcome variable
- sp_list
splines list
- sp_type
splines type
- sp_df
spines degree of freedom
- followup
length of follow up
- seed
set seed
- montecarlo
number of repeated samples for accept-reject algorithm
- boot
doing boostrap
- nboot
bootstraping times
- ci
confident interval
- parallel
run parallel