Find_ridge_best_split_slow: Find_ridge_best_split_slow

In theo-s/Rforestry_R: Random Forests

Description

This is the fast impelementation of the best split function when the leaf prediciton is coming from ridge regression.

Usage

Find_ridge_best_split_slow(feat, y, linear.idx, current.splitting.idx, lambda)

Arguments

feat	all features
y	outcome vector
linear.idx	feature indexes where a Ridge Regression should be fitted
current.splitting.idx	the index of the feature we are currently testing for splits.

Value

It returns a vector of two doubles which determines what the best splitting point is, and what it's MSE is.

split_val_best	the best splitting value
MSE_best	the best MSE

Examples

set.seed(309814)
feat <- matrix(rnorm(80), ncol = 4)
y <- rnorm(20)
linear.idx <- 1:4
current.splitting.idx <- 2
lambda <- .2
Find_ridge_best_split_slow_linear_MASS(
  feat = feat,
  y = y,
  current.splitting.idx = current.splitting.idx,
  linear.idx = linear.idx,
  lambda = lambda
)

theo-s/Rforestry_R documentation built on Dec. 23, 2021, 9:55 a.m.