| Library: | smoother |
| See also: | denbwsel denbwcrit |
| Macro: | denrotp | |
| Description: | determines a rule-of-thumb bandwidth for multivariate density estimation according to Scott. |
| Usage: | hrot = denrotp(x {,K {,opt}}) | |
| Input: | ||
| x | n x 1 vector, the data. | |
| K | string, univariate kernel function on [-1,1] or Gaussian kernel "gau". The rule-of-thumb is determined for the product kernel obtained from K. If K is not given, "qua" is used. | |
| opt | string, if opt equals "robust", a robust estimate for the standard deviations of the data is used. | |
| Output: | ||
| hrot | scalar, rule-of-thumb bandwidth. | |
library("smoother")
x=read("nicfoo")
hrot=denrotp(x)
hrot
Contents of hrot [1,] 0.60341 [2,] 0.41691
library("smoother")
x=read("nicfoo")
hrot=denrotp(x,"qua","robust")
hrot
Contents of hrot [1,] 0.6046 [2,] 0.44921
| Library: | smoother |
| See also: | denbwsel denbwcrit |