Description
Model-Averaged Renewal Process.
Description
To implement a model-averaging approach with different renewal models, with a primary focus on forecasting large earthquakes. Based on six renewal models (i.e., Poisson, Gamma, Log-Logistics, Weibull, Log-Normal and BPT), model-averaged point estimates are calculated using AIC (or BIC) weights. Additionally, both percentile and studentized bootstrapped model-averaged confidence intervals are constructed. In comparison, point and interval estimation from the individual or "best" model (determined via model selection) can be retrieved.
README.md
marp
An R package to apply model-averaging on renewal process.
Install
You can install the released version of marp from GitHub with:
if(!require(devtools)){
install.packages("devtools")
library(devtools)
}
devtools::install_github("kanji709/marp")
Example
Here is a basic example which shows you how to use marp:
# load R package - marp
library(marp)
# generate a small dataset
data <- rgamma(100,3,0.01)
# set parameters
m <- 10 # number of iterations for MLE optimization
t <- seq(100,200,by=10) # time intervals
B <- 99 # number of bootstraps
BB <- 99 # number of double-bootstrapps
alpha <- 0.05 # confidence level
y <- 304 # cut-off time point for probablity estimation
model_gen <- 2 # specifying the data generating model (if known)
# step one: fitting differnt renewal models
res1 <- marp::poisson_rp(dat,t,y)
res2 <- marp::gamma_rp(dat,t,m,y)
res3 <- marp::loglogis_rp(dat,t,m,y)
res4 <- marp::weibull_rp(dat,t,m,y)
res5 <- marp::lognorm_rp(dat,t,y)
res6 <- marp::bpt_rp(dat,t,m,y)
# step two: model selection and obtain model-averaged estimates
res <- marp::marp(dat,t,m,y,which.model = 2)
# step three: construct different confidence intervals (including model-averaged CIs)
ci <- marp::marp_confint(dat,m,t,B,BB,alpha,y,model_gen)