{quantrra}
is a package for reproducible risk assessment developed for the statistical software R.
You can install quantrra
development version from GitHub using the following code in your R console (make sure you have R > 3.5):
# make sure you have the package devtools installed
devtools::install_github('spablotemporal/quantrra')
quantrra
was developed for the implementation of stochastic risk assessment models as an open access alternative to software such as @Risk. quantrra
provides a set of functions to be used in R, and also offers a more user friendly interface trough shiny that can be accessed using:
quantrra::run_quantrra()
The main usage for quantrra is to simulate stochastic events using the function ra_run()
, which requires two arguments:
- m, the model file
- n number of simulations.
A model file is just a data.frame with rows that represent inputs and outputs of the model. The columns include ID, label, whether the event is input or output, a distribution (for inputs), and a formula (for outputs).
The model file then can created in using the shiny interface, R or other software editors. The idea behind model files is that can be easy to share and modify.
# Load a model file from the package
M <- quantrra::OIRSA_M
# Run the model 5000 times
Mo <- ra_run(m = M, nsim = 5000)
# Visualize the results:
ra_plot_dist(Mo$P)
quantrra also provides functions to evaluate the model and creating stratified models to estimate regional or strata-specific risk estimation of events. see the package example for more information.