Convert data frame or matrix object to kde class

A density specified as a data frame or matrix can be converted to a kde object. This is useful for plotting the density using autoplot.kde. As kde objects are defined on a grid, the density values are interpolated based on the points in the data frame or matrix.

Usage

as_kde(object, density_column, ngrid, ...)

Arguments

object: Data frame or matrix with numerical columns, where one column (specified by density_column) contains the density values, and the remaining columns define the points at which the density is evaluated.
density_column: Name of the column containing the density values, specified as a bare expression. If missing, the last column is used.
ngrid: Number of points to use for the grid in each dimension. Default is 10001 for univariate densities and 101 for multivariate densities.
...: Additional arguments are ignored.

Value

An object of class "kde"

Author

Rob J Hyndman

Examples

tibble(y = seq(-4, 4, by = 0.01), density = dnorm(y)) |>
  as_kde()
#> Density of: [y]
#> Computed on a grid of size 10001