ISadditionalblocksearch: Internal function to carry out additional Block Search in...

In gets: General-to-Specific (GETS) Modelling and Indicator Saturation Methods

Internal function to carry out additional Block Search in isat: Leave-One-Out Block Search for Excessive Indicators

Description

Internal function that handles excessive indicator retention by re-running GETS on smaller blocks of indicators using a leave-one-out method. This can be necessary when following a block search, e.g. for IIS, more indicators than observations have been identified. Then another block search is carried out. This can be avoided by choosing tighter significance levels in t.pval or improving the specification.

Usage

ISadditionalblocksearch(mXis, isNames, y, y.n, mX, mXnames, mxkeep, 
  ISmatrixname, indicator.set, print.searchinfo, tol, LAPACK, userEstArg, 
  t.pval, gof.method, gofFunArg, max.regs, max.paths, arLjungB, archLjungB, 
  normality.JarqueB, user.diagnostics, include.gum, include.1cut, 
  include.empty, turbo, do.pet, ratio.threshold, max.block.size, 
  wald.pval, additional.block.search)

Arguments

mXis	Initial regressor matrix including previously selected indicators.
isNames	Character vector with names of the candidate indicator variables.
y	Response variable, typically a numeric vector or time-series.
y.n	Integer indicating the number of observations in the sample.
mX	Matrix of original regressors, excluding indicators.
mXnames	Character vector of column names corresponding to mX.
mxkeep	Indices of regressors to always retain during selection.
ISmatrixname	Character string naming the current IS matrix (e.g., "SIS", "TIS").
indicator.set	Matrix of candidate indicators from which subsets are selected.
print.searchinfo	Logical; if TRUE, outputs progress messages during execution.
tol	Numeric tolerance used in collinearity checks (see qr).
LAPACK	Logical; if TRUE, LAPACK is used for QR decomposition.
userEstArg	List specifying the user-defined estimation function and its arguments.
t.pval	Significance level for coefficient t-tests in the GETS algorithm.
gof.method	Character indicating whether to minimize or maximize the goodness-of-fit criterion.
gofFunArg	List containing the name and arguments of the GOF function used.
max.regs	Maximum number of regressions allowed per deletion path.
max.paths	Maximum number of GETS paths to be explored.
arLjungB	List with order and significance level for the Ljung-Box test on residual autocorrelation.
archLjungB	List with order and significance level for the Ljung-Box test on squared residuals (ARCH).
normality.JarqueB	Significance level for the Jarque-Bera test for normality; if NULL, test is skipped.
user.diagnostics	List containing user-specified diagnostic function and p-value thresholds.
include.gum	Logical; if TRUE, the General Unrestricted Model is included in terminal models.
include.1cut	Logical; if TRUE, the 1-cut model is included in terminal models.
include.empty	Logical; if TRUE, the empty model is included in terminal models.
turbo	Logical; avoids redundant GETS path evaluations if TRUE.
do.pet	Logical; if TRUE, performs Parsimonious Encompassing Tests (PETs).
ratio.threshold	Numeric ratio used to determine block sizes in relation to sample size.
max.block.size	Maximum number of indicators allowed per block during the leave-one-out search.
wald.pval	Significance threshold for Wald tests in the GETS algorithm.
additional.block.search	Default is FALSE. If isat() identifies (retains) too many indicators in the block search, a sensible model might not be estimable (e.g. due to a lack of available degrees of freedom). In this case, setting additional.block.search = TRUE carries out an additional block search using a Leave-one-Out method (e.g. for three blocks, indicators are assigned to block 1,2,3,1,2,3,1,2,3). This is an experimental feature.

Value

A list containing:

ISfinalmodel	The final selected model.
estimations.counter	Updated estimation counter.
getsFun.counter	Updated function call counter.
ISblocks	Final block structure.

Author(s)

Moritz Schwarz, https://www.moritzschwarz.org/

See Also

ISMatricesLoop, isat

gets documentation built on July 29, 2026, 9:07 a.m.