| Library: | rpclib |
| See also: | xgobiisotonic xgobilinreg xgobisknn |
| Quantlet: | xgobiknn | |
| Description: | conmputes an k-nn fit for XGobi |
| Usage: | yh = xgobiknn(x,y,g,h) | |
| Input: | ||
| x | n x 1 matrix | |
| y | n x 1 matrix | |
| g | n x 1 matrix with integers starting at zero | |
| kp | scalar number of neighbours | |
| Output: | ||
| yh | n x 1 matrix smoothed fit | |
; load the library for XGobi
library ("rpclib")
; create a display
d=createdisplay(1,1)
; read a data set
x=read("motcyc")
; generate randomly three subgroups
g=floor(3.*uniform(rows(x)))
; color the data after the subgroups
setmaskp (x, g, 3, 4)
; show the data and the smooth
show (d, 1, 1, x, x[,1]~xgobiknn(x[,1],x[,2],g,7))
shows a k-nn fit for the three different subgroups in the data
| Library: | rpclib |
| See also: | xgobiisotonic xgobilinreg xgobisknn |