distribution: Create a 'Distribution' object.
In trim: Tree methods for classification of imbalanced data

Description Usage Arguments Value Author(s) See Also Examples

The distribution function is a generic S4 method for creating a Distribution object.

1	distribution(x, precision = .Machine$double.eps, p.estimator = NA)

`x`	A R object, see `distribution-methods`.
`precision`	A single numeric value. To be valid a `Distribution` object has to satisfy sum(object) == 1 +- precision.
`p.estimator`	Used only when `x` is a `character` or a `factor`. In this case `p.estimator` has to be function able to estimate probabilities, see for example `p.estimator.laplace`. A such function takes two arguments: `y` and `weights`, with `y` a `character` or a `factor`, and return a `numeric`.

A Distribution object.

Emmanuel Rousseaux

Distribution, distribution.comparable, distribution.reorder.

distribution(c(0.2,0.8))
data(iris)
distribution(iris$Species)

distribution(
  wvc(iris$Species, weights=c(rep(1,50), rep(0,100))),
  p.estimator = NA
)
distribution(
  wvc(iris$Species, weights=c(rep(1,50), rep(0,100))),
  p.estimator = p.estimator.laplace
)