lambdastar: Measurement and Linear Hypothesis Models for Lambda Star
Estimates intrinsic and captured noncentrality from parallel
measurements and evaluates the numerical parsimony functional for an
explicitly encoded hypothesis matrix, formula, or compatible linear model.
Includes common-case model comparisons, quantized entropy capacity,
controlled temperature integration, and explicit singularity diagnostics.
Uses manuscript projection estimates by default, with an explicit alternative
population estimator. Supports common nuisance adjustment and ordinary case
or cluster percentile bootstrap intervals, with diagnostics for undefined
estimates and preserved model coding.
Provides reusable row-bound design specifications and explicit conditional
term blocks, named linear restrictions, explicit predictor-grid contrasts,
and fixed or reevaluated basis recipes under a homogeneous isotropic
measurement-fluctuation assumption. Supports explicit known-reference
mean hypotheses and paired differences from parallel measurement pairs.
Encodes fixed person-by-occasion models with parallel indicators, implicit
person adjustment and whole-person bootstrap with distinct sampled copies.
The underlying method is described in Hammes (2026)
<doi:10.5281/zenodo.22962377>.
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=lambdastar
to link to this page.