join_node: Node joiner

In cat.dt: Computerized Adaptive Testing and Decision Trees

Description

Given all the nodes from one level, join_node evaluates all possible pairs one by one and decides whether or not to join them based on the similarity between the estimated ability levels and the density functions. If a pair of nodes is joined, the density function of the resulting node is the mean of the density functions of the joined nodes and the confluencies are summed.

Usage

join_node(nodes, level, limit, tol, inters)

Arguments

nodes	list of node lists. Every node list must contain the ID of the node, the vector of density function values of the evaluated abiity levels, the vector of previous items, the estimated ability level and the node confluency
level	level of the CAT decision tree
limit	maximum number of level nodes
tol	minimum distance between estimated ability levels to join two nodes
inters	minimum common area between density functions in the nodes of the evaluated pair in order to join them

Value

A list of node lists. This list is the input list updated with the joined nodes

Author(s)

Javier Rodr?guez-Cuadrado

cat.dt documentation built on March 31, 2021, 5:07 p.m.