Distribution Classes — Mastiff-Distributions • mastiff

Distributions implemented as R6 classes in mastiff

Details

mastiff introduces a R6 class structure which is used to combine the density, distribution function, quantile function and generation of random deviates into a single object.

Parameters for a distribution are set at initialisation and can be updated via a named list $params stored in the distribution class.

Each distribution includes methods

`$d(x)`		Evaluates the density at values `x`
`$p(q)`		Evaluates the distribution function at values `q`
`$q(p)`		Evaluates the quantile function at values `p`
`$r(n)`		Generates `n` random values from the distribution
`$mean`		Returns the mean of the distribution
`$sd`		Returns the standard deviation of the distribution
`$var`		Returns the variance of the distribution

Discrete Distributions

Discrete distributions with class distribution.discrete.class

distribution.binomial	Binomial distribution with size `size` and success probability `prob`
distribution.negative_binomial	Negative Binomial distribution with size `size` and success probability `prob`
distribution.point_mass	Point mass at `value`
distribution.poisson	Poisson distribution with mean `lambda`

Continuous Distributions

Discrete distributions with class distribution.continuous.class

distribution.exponential	Exponential distribution with rate `rate`
distribution.gamma	Gamma distribution with shape `shape` and rate `rate`
distribution.normal	Normal distribution with mean `mean` and standard deviation `sd`
distribution.uniform	Uniform distribution on `[min, max]`

Examples

# Construct a Poisson( 1 ) random variable
Pois_RV <- distribution.poisson( lambda = 1 )

# Evaluate the density, equivalent to dpois( 0 : 5, lambda = 1 )
Pois_RV$d( 0 : 5 )
#> [1] 0.367879441 0.367879441 0.183939721 0.061313240 0.015328310 0.003065662

# Evaluate the distribution function, equivalent to ppois( 0 : 5, lambda = 1 )
Pois_RV$p( 0 : 5 )
#> [1] 0.3678794 0.7357589 0.9196986 0.9810118 0.9963402 0.9994058

# Evaluate the quantile function, equivalent to qpois( c( 0.5, 0.8 ), lambda = 1 )
Pois_RV$q( c( 0.5, 0.8 ) )
#> [1] 1 2

# Generate random deviates, equivalent to rpois( 10, lambda = 1 )
Pois_RV$r( 10 )
#>  [1] 0 2 1 0 0 1 1 0 1 2

# Update parameters to a Poisson( 10 ) random variable
Pois_RV$params <- list( lambda = 10 )
mean( Pois_RV$r( 1e5 ) )
#> [1] 9.99972