meanshift: The Meanshift mode estimator

In modeest: Mode Estimation

Description

The Meanshift mode estimator.

Usage

meanshift(
  x,
  bw = NULL,
  kernel = "gaussian",
  par = shorth(x),
  iter = 1000,
  tolerance = sqrt(.Machine$double.eps)
)

Arguments

x	numeric. Vector of observations.
bw	numeric. The smoothing bandwidth to be used.
kernel	character. The kernel to be used. Available kernels are "biweight", "cosine", "eddy", "epanechnikov", "gaussian", "optcosine", "rectangular", "triangular", "uniform". See density for more details on some of these kernels.
par	numeric. The initial value used in the meanshift algorithm.
iter	numeric. Maximal number of iterations.
tolerance	numeric. Stopping criteria.

Value

meanshift returns a numeric value, the mode estimate, with an attribute "iterations". The number of iterations can be less than iter if the stopping criteria specified by eps is reached.

Note

The user should preferentially call meanshift through mlv(x, method = "meanshift", ...).

References

• Fukunaga, K. and Hostetler, L. (1975). The estimation of the gradient of a density function, with applications in pattern recognition. IEEE Transactions on Information Theory, 21(1):32–40.

See Also

mlv, tsybakov.

Examples

# Unimodal distribution
x <- rweibull(100, shape = 12, scale = 0.8)

## True mode
weibullMode(shape = 12, scale = 0.8)

## Estimate of the mode
mlv(x, method = "meanshift", par = mean(x))

modeest documentation built on Nov. 18, 2019, 5:07 p.m.