In jimbrig/lossrx: Actuarial Loss Development and Reserving with R

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  results = "asis",
  message = FALSE,
  warning = FALSE
)

library(summarytools)
library(kableExtra)
library(actuar)
library(fplot)

st_options(plain.ascii = FALSE, style = "rmarkdown")

devtools::load_all()
# library(lossrx)

data("claims_transactional")
data("losses")
data("exposures")

latest_eval <- losses |> dplyr::filter(eval_date == max(.data$eval_date))
wc_dat <- latest_eval |> dplyr::filter(coverage == "WC")
al_dat <- latest_eval |> dplyr::filter(coverage == "AL")

lossrx Datasets

lossrx comes with some built in data for example usage, including:

• a simulated transactional claims data.frame
• a suite of example WC and AL lossruns combined into a single data.frame
• sample exposure data for WC ($ payroll) and AL (vehicles or miles driven)

Loss Data

plot_distr(
  ~ total_incurred | coverage,
  latest_eval,
  mod.method = "split"
)

Top 10 Rows:

head(losses) |>
  kable(format = "html", digits = 2) |>
  kable_styling()

Summary:

print(dfSummary(losses, 
                varnumbers   = FALSE, 
                valid.col    = FALSE, 
                graph.magnif = 0.76),
      method = 'render')

Worker's Compensation

Distribution of Claims

library(fplot)
fplot::plot_distr(wc_dat$total_incurred)
plot_distr(~ total_incurred | cause, wc_dat, cumul = TRUE)

plot_lines(
  total_incurred ~ program_year,
  losses
)

jimbrig/lossrx documentation built on Dec. 20, 2024, 3:46 a.m.