VectorDistribution: Vectorise Distributions

Description

A wrapper for creating a vector of distributions.

Value

Returns an R6 object of class VectorDistribution.

Constructor

VectorDistribution$new(distlist = NULL, distribution = NULL, params = NULL, name = NULL, short_name = NULL, description = NULL, decorators = NULL)

Constructor Arguments

Argument	Type	Details
`distlist`	list	List of distributions.
`distribution`	character	Distribution to wrap.
`params`	a R object	Either list of parameters or matrix-type frame, see examples.
`shared_params`	a R object	Either list of shared parameters or matrix-type frame, see examples.
`name`	list	Optional new name for distribution.
`short_name`	list	Optional new short_name for distribution.
`description`	list	Optional new description for distribution.
`decorators`	list	Decorators to pass to wrapped distributions on construction.

Constructor Details

A vector distribution can either be constructed by a list of distributions passed to distlist or by passing the name of one or more distributions implemented in distr6 to distribution, as well as a list or table of parameters to params. The former case provides more flexibility in the ability to use wrapped distributions but the latter is vastly faster for distributions of class SDistribution or custom distributions. The shared_params parameter decreases memory usage and improves speed by storing any parameters shared between distributions only once (instead of repeated in a list).

Public Variables

Variable	Return
`name`	Name of distribution.
`short_name`	Id of distribution.
`description`	Brief description of distribution.

Public Methods

Accessor Methods	Link
`wrappedModels(model = NULL)`	`wrappedModels`
`decorators()`	`decorators`
`traits()`	`traits`
`valueSupport()`	`valueSupport`
`variateForm()`	`variateForm`
`type()`	`type`
`properties()`	`properties`
`support()`	`support`
`symmetry()`	`symmetry`
`sup()`	`sup`
`inf()`	`inf`
`dmax()`	`dmax`
`dmin()`	`dmin`
`skewnessType()`	`skewnessType`
`kurtosisType()`	`kurtosisType`



d/p/q/r Methods	Link
`pdf(x1, ..., log = FALSE, simplify = TRUE)`	`pdf`
`cdf(x1, ..., lower.tail = TRUE, log.p = FALSE, simplify = TRUE)`	`cdf`
`quantile(p, ..., lower.tail = TRUE, log.p = FALSE, simplify = TRUE)`	`quantile.Distribution`
`rand(n, simplify = TRUE)`	`rand`



Statistical Methods	Link
`prec()`	`prec`
`stdev()`	`stdev`
`median()`	`median.Distribution`
`iqr()`	`iqr`
`cor()`	`cor`



Parameter Methods	Link
`parameters(id)`	`parameters`
`getParameterValue(id, error = "warn")`	`getParameterValue`
`setParameterValue(..., lst = NULL, error = "warn")`	`setParameterValue`



Validation Methods	Link
`liesInSupport(x, all = TRUE, bound = FALSE)`	`liesInSupport`
`liesInType(x, all = TRUE, bound = FALSE)`	`liesInType`



Representation Methods	Link
`strprint(n = 2)`	`strprint`
`print(n = 2)`	`print`
`summary(full = T)`	`summary.Distribution`

Details

A vector distribution is intented to vectorize distributions more efficiently than storing a list of distributions. To improve speed and reduce memory usage, distributions are only constructed when methods (e.g. d/p/q/r) are called. Whilst it is recommended to first extract distributions using `[` before querying them for results, all common methods are available in `VectorDistribution` as they are wrapped in `apply`.

Examples

Run this code

# NOT RUN {
vecDist <- VectorDistribution$new(list(Binomial$new(prob = 0.5,
                           size = 10), Normal$new(mean = 15)))
vecDist$pdf(x1 = 2, x2 =3)
# Equivalently
vecDist[1]$pdf(2); vecDist[2]$pdf(3)
# Or to evaluate every distribution at the same point
vecDist$pdf(1)

# Same wrapping for statistical functions
vecDist$mean()
c(vecDist[1]$mean(), vecDist[2]$mean())
vecDist$entropy()
c(vecDist[1]$entropy(), vecDist[2]$entropy())

vecDist$cdf(1:5, 12:16)
vecDist$rand(10)

vecBin = VectorDistribution$new(distribution = "Binomial",
       params = list(list(prob = 0.1, size = 2),
                   list(prob = 0.6, size = 4),
                   list(prob = 0.2, size = 6)))
vecBin$pdf(x1=1,x2=2,x3=3)
vecBin$cdf(x1=1,x2=2,x3=3)
vecBin$rand(10)

#Equivalently
vecBin = VectorDistribution$new(distribution = "Binomial",
       params = data.table::data.table(prob = c(0.1,0.6,0.2), size = c(2,4,6)))
vecBin$pdf(x1=1,x2=2,x3=3)
vecBin$cdf(x1=1,x2=2,x3=3)
vecBin$rand(10)

# sharedparams is very useful for vectorized custom distributions
shared_params = list(name = "A Distribution", short_name = "Dist", type = Reals$new())
params = list(list(pdf = function(x) return(1)), list(pdf = function(x) return(2)))
vecdist = VectorDistribution$new(distribution = "Distribution", params = params,
                                   shared_params = shared_params)
vecdist$pdf(1)

# }

Run the code above in your browser using DataLab