prediction: Compute prediction

In jpasquier/ad_test: What the Package Does (One Line, Title Case)

Description

Compute a prediction with ...

Usage

prediction(time, comp, id, cor, k, w, V = "naive")

Arguments

time	individual (and time dependent) times
comp	individual and time dependent compliances (0/1)
id	(individual and) time dependent ids
cor	correlation structure of the GEE model
k	knots for the splines method
w	weights of the weighted adherence method
V	variance estimate of model coefficients (default="naive")

Value

a list containing

• predict: dataframe containing prediction and confidence intervals of the model

• q: QIC goodness of fit statistics

Examples

data(onco)
onco$compliance <- as.numeric(onco$observed >= onco$expected)
prediction(onco$drel, onco$compliance, onco$id, "ar1",
           c(120, 240), rep(1, nrow(onco)))

jpasquier/ad_test documentation built on Nov. 14, 2019, 8:09 p.m.