The goal of distributions is to have a standard collection of objects that define random variables in R. The philosophy of this package follows the same principles found in Numpy or Julia. That is, the emphasis is having an object representing a random variable for which we can query according to our interests. For example, we could evaluate the density function or generate random numbers.
This is mainly done to avoid the typical R syntax of the family of functions that surround a random variable. See for example the functions needed for a normal random variable.
The main objective of this package is to be used in my teaching but I am happy to extend it further to a more general audience if needed.
Installation
You can install the development version of distributions from GitHub with:
# install.packages("devtools")
devtools::install_github("agarbuno/distributions")Example
This is a basic example which shows how to define a standard normal random variable.
library(distributions)
x <- GaussianRandomVariable$new()
x
#> <Gaussian>
#> Inherits from: <ContinuousRandomVariable>
#> Public:
#> density: function (x, log = TRUE)
#> initialize: function (mean = 0, sd = 1)
#> mean: 0
#> sample: function (nsamples = 1, nreps = 1)
#> sd: 1
class(x)
#> [1] "Gaussian" "ContinuousRandomVariable"
#> [3] "R6"You can use the sample or the density methods to either generate random numbers from this random variable or evaluate its density function. For example, we can generate a collection of 100 random points or evaluate the density function in 1.
x$sample(100)
x$density(1)You’ll still need to render README.Rmd regularly, to keep README.md up-to-date. devtools::build_readme() is handy for this. You could also use GitHub Actions to re-render README.Rmd every time you push. An example workflow can be found here: https://github.com/r-lib/actions/tree/v1/examples.