find_split: Find the Best Split

In puls: Partitioning Using Local Subregions

Find the Best Split

Description

Find the best split in terms of reduction in inertia for the transferred node, indicate by row. Find the terminal node with the greatest change in inertia and bi-partition it.

Usage

find_split(
  toclust.fd,
  frame_row,
  cloc,
  dist,
  dsubs,
  dsubsname,
  weights,
  minbucket,
  minsplit,
  spliton,
  method
)

Arguments

toclust.fd	A functional data object (i.e., having class fd) created from fda package. See fda::fd().
frame_row	One row of the split tree as data frame.
cloc	Vector of current cluster membership.
dist	Distance matrix of all observations in the data.
dsubs	Distance matrix calculated on each subregion. A three-dimensional matrix.
dsubsname	Subregion names.
weights	(Currently unused) Weights on observations.
minbucket	The minimum number of data points in one cluster allowed.
minsplit	The minimum number of observations that must exist in a node in order for a split to be attempted.
spliton	Restrict the partitioning on a specific set of subregions.
method	The clustering method you want to run in each subregion. Can be chosen between pam and ward.

Value

The updated frame_row with the next split updated.

puls documentation built on June 8, 2025, 1:48 p.m.