logKDE (version 0.3.2)

bw.logG: Bandwidth estimation for strictly positive distributions.

Description

Computes bandwidth for log domain KDE using the Silverman rule.

Usage

bw.logG(x)

Arguments

x

numeric vector of the data. Must be strictly positive, will be log transformed during estimation.

Value

bw the optimal bandwidth.

References

Silverman, B. W. (1986). Density estimation for statistics and data analysis. Monographs on Statistics and Applied Probability. 26.

Wand, M. P., Marron, J. S., & Ruppert, D. (1991). Transformations in density estimation. Journal of the American Statistical Association, 86(414), 343-353.

Examples

# NOT RUN {
bw.logG(rchisq(100,10))

# }