library("FlashAustria")
## Visualization of fitted model
if(require("bamlss") && require("gamlss.dist") && require("gamlss.tr")) {
data("FlashAustria", package = "FlashAustria")
data("FlashAustriaModel", package = "FlashAustria")
ztSICHEL <- trun(0, family = "SICHEL", local = FALSE)
plot(flash_model_ztSICHEL)
}
## Replication code:
## Required packages.
library("bamlss")
library("gamlss.dist")
library("gamlss.tr")
## Load the data.
data("FlashAustria", package = "FlashAustria")
## Generate zero truncated Sichel distribution.
ztSICHEL <- trun(0, family = "SICHEL", local = FALSE)
## Model formula, up to four parameters.
f <- list(
counts ~ s(d2m, bs = "ps") + s(q_prof_PC1, bs = "ps") +
s(cswc_prof_PC4, bs = "ps") + s(t_prof_PC1, bs = "ps") +
s(v_prof_PC2, bs = "ps") + s(sqrt_cape, bs = "ps"),
~ s(sqrt_lsp, bs = "ps")
)
## Estimate models.
set.seed(123)
flash_model_ztnbinom <- bamlss(f,
family = "ztnbinom", data = FlashAustriaTrain,
optimizer = opt_boost, cores = 3, light = TRUE, binning = TRUE,
maxit = 1000, thin = 3, burnin = 1000, n.iter = 2000)
set.seed(123)
flash_model_ztSICHEL <- bamlss(f,
family = ztSICHEL, data = FlashAustriaTrain,
optimizer = opt_boost, cores = 3, light = TRUE, binning = TRUE,
maxit = 1000, thin = 3, burnin = 1000, n.iter = 2000)Fitted Distributional Regression Model Object (bamlss)
Description
A zero-truncated negative binomial bamlss model trained on the FlashAustriaTrain data using gradient boosting with subsequent MCMC sampling. The response variable are the lightning counts and the regression terms are P-splines based on ERA5 covariates.
Usage
data("FlashAustriaModel", package = "FlashAustria")Format
An object of class bamlss.
References
Umlauf N, Klein N, Zeileis A (2018). BAMLSS: Bayesian Additive Models for Location, Scale and Shape (and Beyond). Journal of Computational and Graphical Statistics, 27(3), 612–627. doi:10.1080/10618600.2017.1407325