README.md

In sebpardo/rtails: Simulate heavy-tailed deviations

rtails: Simulate heavy-tailed deviations

Installation

install.packages("remotes")
remotes::install_github("sebpardo/rtails")

Introduction to rtails

The rtails package provides functions to easily obtain heavy-tailed multiplicative random deviates (i.e., positive only, centered around 1). While its primary aim is to simulate heavy-tailed recruitment deviations, these functions could also be used for other applications as well.

The heavy-tailed random deviates are generated using four distributions:

• Normal mixture (i.e. normal with sporadic wider normal draws): rnorm_tails()
• Pareto: rpareto_tails()
• Student-t: rst_tails()
• Generalized Extreme Variance: rgev_tails()

Where possible, functions are analytically bias-corrected around a mean of 1 by default. However, this correction is not always applicable (e.g. Student-t and GEV). We therefore also allow all functions to be bias corrected using the drawn sample mean (sample_bias_correct). All functions also take a seed argument so that different parameterizations of each distribution using a given seed can be more directly comparable.

We also provide the mutate_tails() function that converts an existing set of normal random deviates into deviates with a Student-t or Pareto distribution, and allows for bootstrapping.

Lastly, the plot_tails() function allows for easy visualization of the generated random deviates.

See the vignette for more details.

sebpardo/rtails documentation built on Dec. 22, 2021, 11:17 p.m.