Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Library: xclust
See also: cartsplit cartsplitopt cartcv leafnum maketr pred prune prunecv pruneseq prunetot ssr kuva

Macro: prederr
Description: Given a regression tree and observations gives the mean predicton error of the regression tree for that data.

Usage: mssr = prederr (tr, x, y, type)
Input:
tr list of vectors: data structure which represents a binary tree and is produced by cartsplit procedure, contains vectors tr.val, tr.vec, tr.mean, tr.ssr, tr.nelem. See cartsplit for the description of tr.
x n x p vector: represents n points in the sample space at which the prediction of the regression tree will be calculated.
y n x 1 vector: contains the values of the response variable.
type p x 1 vector: contains the types of the original regression variables, 1 means that the corresponding variable is continuous and 0 that it is categorical. This vector should be similar to the vector which was given originally as an input to the cartsplit which produced the regression tree tr.
Output:
mssr real number >0: mean of the squared residuals, that is, sum of squared differences between predictions and observations, divided by the number of observations n.

Example:
; loads the library xclust
library ("xclust")
;let us generate a tree by cartsplit procedure
x1=#(0,0,0,0,1,1,1,1,1,2)
x2=#(0,0,0,0,0,0,0,1,1,1)
x=x1~x2
y=#(0,0,0,0,100,100,100,120,120,120)
tree=cartsplit(x,y,#(0,1))
tree
xmat=uniform(5,2)
xmat
ymat=uniform(5,1)
ymat
mssr=prederr(tree,xmat,ymat,#(0,1))
mssr
Result:

Content of object tr.val.split0
[1,] 0 
[2,] 1,2
Content of object tr.val.split1
[1,] NaN 
Content of object tr.val.split2
[1,] 0 
Content of object tr.val.split3
[1,] NaN 
Content of object tr.val.split4
[1,] NaN 
Content of object tr.vec
[1,] 1.000000 
[2,] NaN 
[3,] 2.000000 
[4,] NaN 
[5,] NaN 
Content of object tr.mean
[1,] 66.000000 
[2,] 0.000000 
[3,] 110.000000 
[4,] 100.000000 
[5,] 120.000000 
Content of object tr.ssr
[1,] 29640.000000 
[2,] 0.000000 
[3,] 600.000000 
[4,] 0.000000 
[5,] 0.000000 
Content of object tr.nelem
[1,] 10.000000 
[2,] 4.000000 
[3,] 6.000000 
[4,] 3.000000 
[5,] 3.000000 
Content of object xmat
[1,] 0.000000 0.231587 
[2,] 0.003027 0.429810 
[3,] 0.124776 0.604880 
[4,] 0.420721 0.615851 
[5,] 0.003547 0.262006 
Content of object ymat
[1,] 0.749410 
[2,] 0.997687 
[3,] 0.248579 
[4,] 0.437815 
[5,] 0.791837 
Content of object mssr
[1,] 0.487495 

Library: xclust
See also: cartsplit cartsplitopt cartcv leafnum maketr pred prune prunecv pruneseq prunetot ssr kuva

Keywords - Function groups - @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Author: Jussi Klemelae, 980323
(C) MD*TECH Method and Data Technologies, 28.6.1999