mklewarp {MKLE}R Documentation

Maximum kernel likelihood estimation

Description

Computes the maximum kernel likelihood estimator for a given dataset and bandwidth using fast fourier transforms.

Usage

mklewarp(data, bw = 1, gs = 2^11, K="gaussian")

Arguments

data the data for which the log likelihood will be computed.
bw the smoothing bandwidth to be used.
gs the number of gridpoints to be used for the fourier transform
K a character string giving the kernel function to be used. This must be one of '"gaussian"', '"rectangular"', '"triangular"', '"epanechnikov"', '"biweight"', '"cosine"' or '"optcosine"', with default '"gaussian"'.

Details

The underlying shifted kernel density estimator is defined as

frac{1}{nh}sum_{i=1}^nK(frac{y-X_i-bar X+theta}{h})

. The default for the bandwidth is 2*sigma, which is the optimal value if a Gaussian kernel is used.

Value

the MKLE.

Author(s)

Thomas Jaki

References

~coming soon ~

See Also

mkle and klik

Examples

## compares the MKLE and the warped MKLE.
data(state)
attach(state)

mkle(crime)
mklewarp(crime)


[Package MKLE version 0.02 Index]