marp {marp}

A function to apply model-averaged renewal process

Description

A function to apply model-averaged renewal process

Usage

marp(data, t, m, y, which.model = 1)

Arguments

data

input inter-event times

t

user-specified time intervals (used to compute hazard rate)

m

the number of iterations in nlm

y

user-specified time point (used to compute time-to-event probability)

which.model

user-specified generating (or true underlying if known) model

Value

returns list of estimates obtained from different renewal processes and after applying model-averaging

par1

Estimated scale parameters (if applicable) of all six renewal models

par2

Estimated shape parameters (if applicable) of all six renewal models

logL

Negative log-likelihood

AIC

Akaike information criterion (AIC)

BIC

Bayesian information criterion (BIC)

mu_hat

Estimated mean

pr_hat

Estimated (logit) probabilities

haz_hat

Estimated (log) hazard rates

weights_AIC

Model weights calculated based on AIC

weights_BIC

Model weights calculated based on BIC

model_best

Model selected based on the lowest AIC

mu_best

Estimated mean obtained from the model with the lowest AIC

pr_best

Estimated probability obtained from the model with the lowest AIC

haz_best

Estimated hazard rates obtained from the model with the lowest AIC

mu_gen

Estimated mean obtained from the (true or hypothetical) generating model

pr_gen

Estimated probability obtained from the (true or hypothetical) generating model

haz_gen

Estimated hazard rates obtained from the (true or hypothetical) generating model

mu_aic

Estimated mean obtained from model-averaging (using AIC weights)

pr_aic

Estimated probability obtained from model-averaging (using AIC weights)

haz_aic

Estimated hazard rates obtained from model-averaging (using AIC weights)

Examples

set.seed(42)
data <-  rgamma(100,3,0.01)

# set some parameters
m = 10  # number of iterations for MLE optimization
t = seq(100, 200, by=10)  # time intervals
y = 304  # cut-off year for estimating probability
which.model <- 2 # specify the generating model

# model selection and averaging
result <- marp::marp(data, t, m, y, which.model)