raschtree: Recursive partitioning (also known as trees) based on Rasch...

In mirka-henninger/raschTreeEffectSize: Helper functions to use the Mantel-Haenszel effect size measure as a stopping criterion in Rasch trees

Description

Recursive partitioning (also known as trees) based on Rasch models. for more details see psychotree::raschtree

Usage

raschtree(
  formula,
  data,
  na.action,
  reltol = 1e-10,
  deriv = c("sum", "diff", "numeric"),
  maxit = 100L,
  ...
)

Arguments

formula	A symbolic description of the model to be fit. This should be of type y ~ x1 + x2 where y should be a binary 0/1 item response matrix and x1 and x2 are used as partitioning variables.
data	a data frame containing the variables in the model.
na.action	a function which indicates what should happen when the data contain missing values (NAs).
reltol	arguments passed via raschmodel to optim.
deriv	character. Which type of derivatives should be used for computing gradient and Hessian matrix? Analytical with sum algorithm ("sum"), analytical with difference algorithm ("diff", faster but numerically unstable), or numerical. Passed to raschmodel.
maxit	arguments passed via raschmodel to optim.
...	arguments passed to the underlying functions, i.e., to mob_control for raschtree, and to the underlying predict and plot methods, respectively.

mirka-henninger/raschTreeEffectSize documentation built on Nov. 14, 2021, 11:15 p.m.