feature {feature}R Documentation

feature

Description

Package for feature significance for multivariate kernel density estimation (for 1- to 4-dimensional data).

Details

There is one main function in this package, featureSignif. It has a range of options which allow the user to display compute and kernel density estimates, significant gradient and significant curvature regions. Significant gradient and/or curvature regions often correspond to significant features (e.g. local modes).

It's available in an interactive mode where the user selects bandwidths in the graphics window and the significant regions are computed in real-time. This interactive mode is useful for initial exploratory data analysis.

Otherwise there is the non-interactive mode where the user can specify a single set of bandwidths. This is useful when the user has a more definite idea of suitable bandwidths.

For a more detailed example for 1-d and 2-d data, see vignette("feature").

Author(s)

Tarn Duong <tarn.duong@gmail.com> & Matt Wand <wand@uow.edu.au>

References

Chaudhuri, P. and Marron, J.S. (1999) SiZer for exploration of structures in curves. Journal of the American Statistical Association, 94, 807-823.

Duong, T., Cowling, A., Koch, I., Wand, M.P. (2008) Feature significance for multivariate kernel density estimation. Computational Statistics and Data Analysis, 52, 4225-4242.

Godtliebsen, F., Marron, J.S. and Chaudhuri, P. (2002) Significance in scale space for bivariate density estimation. Journal of Computational and Graphical Statistics, 11, 1-22.

Scott, D.W. (1992) Multivariate Density Estimation: Theory, Practice, and Visualization. John Wiley & Sons. New York.

Wand, M.P. & Jones, M.C. (1995) Kernel Smoothing. Chapman & Hall/CRC. London.

See Also

sm, KernSmooth


[Package feature version 1.1-15 Index]