Local Inferential Feature Significance for Multivariate Kernel Density Estimation.
Introduction
The feature package contains functions to display and compute kernel density estimates, significant gradient and significant curvature regions. Significant gradient and/or curvature regions often correspond to significant features (e.g. local modes).
There are two main functions in this package. featureSignifGUI() is the interactive function where the user can select bandwidths from a pre-defined range. This mode is useful for initial exploratory data analysis. featureSignif() is the non-interactive function. This is useful when the user has a more definite idea of suitable values for the bandwidths. The latter is closely related to the ks::kfs() function.
For a more detailed example for 1-, 2- and 3-d data, see vignette("feature").
Installation
Install from CRAN:
install.packages("feature")
Further reading
Duong, T., Cowling, A., Koch, I., and Wand, M. P. (2008) Feature significance for multivariate kernel density estimationComputational Statistics and Data Analysis52, 4225--4242.