Fit kernel density estimator (KDE) object — fit_kde

This function fits a kernel density estimator to a set of data points. Another wrapper of ks::kde.

fit_kde_object(df, h_band = NULL, position = 1:2, grid_size = rep(1000, 2))

Arguments

df	data.frame with all the points
h_band	optional argument for the bandwidth of the kde object. If NULL, the optimal band would be selected through the `ks::kde` function. Default is NULL.
position	Columns position of x/y pair. Default is 1:2
grid_size	size of the grid which is going to be used for the evaluation of kde object. Can be reduced to speed-up computation.

KDE object fitted to the df data.frame.

This function is shared with TCpredictionbands on github: TCpredictionbands.

if (FALSE) {
set.seed(8192)
x <- 2^rnorm(100)
y <- rnorm(100)
dfmat <- cbind(x,y)

kde_object <- fit_kde_object(dfmat)
}