f_fit_linERR_wf: fit Excess Relative Risk Model

In fbr600/rERR: Excess Relative Risk Models

Description

function that calls the optimization (mle from stats4 package, so use optim) from an event format data set, and return a rERR object with the estimation and summary

Usage

f_fit_linERR_wf(formula, data, id_name, doses, times, covars, lag,
  exclusion_done = F)

Arguments

formula	Surv(entry_time,exit_time,outcome)~loglin(loglin_var1,..,loglin_varn)+ lin(lin_var1,..,lin_varm)+strata(strat_var1,...strat_varp)
data	data set returned from f_to_model_data
id_name	name of variable containing the names of subjects
doses	sub data set of grouped doses
times	sub data set of times relatives to doses
covars	sub data set of the covars that will be involved in the model (adjustments and stratification)
lag	latency period
exclusion_done	a logical indicating wheather the exclusion is already done or not

Value

rERR object with the estimation

Examples

# set the formulas for the models
formula1 <- Surv(AgeAtEntry,age_at_event,outcome) ~ lin(dose_cum) + strata(sex)
formula2 <- Surv(AgeAtEntry,age_at_event,outcome) ~ loglin(factor(country)) + lin(dose_cum) + 
                                                    strata(sex)

  
# fit the models
fit1 <- f_fit_linERR_wf(formula1,data=cohort_wf,id_name="id",doses=cohort_wf[,45:79],
                        times=cohort_wf[,10:44],covars=cohort_wf[,c("sex","country")],
                        lag=2,exclusion_done = FALSE)

fit2 <- f_fit_linERR_wf(formula2,data=cohort_wf,id_name="id",doses=cohort_wf[,45:79],
                        times=cohort_wf[,10:44],covars=cohort_wf[,c("sex","country")],
                        lag=2,exclusion_done = FALSE)

# display a summary
summary(fit1)
summary(fit2)

# confidence intervals
confint(fit1)
confint(fit2) 

# likelihood ratio test between nested and nesting models#' 
f_lrt(fit1,fit2)

fbr600/rERR documentation built on May 7, 2019, 9:36 a.m.