BivariateProjection: ** RESERVED FOR INTERNAL USE **

In benja0x40/Barbouille: Visualisation of empirical distributions

** RESERVED FOR INTERNAL USE **

Description

Bivariate projection

Usage

BivariateProjection(
  V,
  stencil = c("linear", "cosine", "sigmoid"),
  spray = "uniform",
  fwhm = 1/2
)

Arguments

V	numeric matrix with two columns representing bivariate observations.
stencil	either "linear", "cosine" or "sigmoid". It is possible to use non-ambiguous abbreviated forms of these keywords.
spray	defines the distribution used to generate noise values. This can be either "uniform" (default), "triangle", "cosine", or "normal". It is possible to use non-ambiguous abbreviated forms of these keywords.
fwhm	positive numeric value defining the Full Width at Half Maximum of the normal distribution used when spray = "normal". The default value is 0.5 and produces normally distributed noise values with a similar span as when using the triangle and cosine distributions which both have the same FWHM of 0.5 by design.

Value

BivariateProjection returns a numeric matrix representing 2D coordinates.

See Also

UnivariateProjection, Atomize

Examples

## Not run: 

n <- 50000
V <- cbind(
  c(rnorm(n, 10, 5), rnorm(n, -10, 1)),
  c(rnorm(n, 10, 1), rnorm(n, -10, 5))
)

P <- BivariateProjection(V)
r <- BivariateDensity(P)

## End(Not run)

benja0x40/Barbouille documentation built on March 26, 2023, 11:38 p.m.