Package index • hsquared

Start here

Live verbs for the default job: one Gaussian trait, one animal term, REML. Read current limits before reporting.

hsquared(): Fit a quantitative-genetic model
hs_data() experimental: Create an hsquared data container
hs_control(): Create hsquared control options
animal() experimental: Animal-model formula marker
rr() experimental: Random-regression formula marker
model_spec() experimental: Inspect a parsed hsquared model specification
formula_status() experimental: Inspect formula grammar status
validation_status() experimental: Inspect validation evidence status
data_status() experimental: Inspect hsquared data-container status
backend_info() experimental: Inspect planned compute backends

Core extractors

What you read off a fitted univariate Gaussian animal model — the covered default route. Read fit_diagnostics() (especially at_boundary) before reporting anything else.

variance_components() experimental: Extract variance components
heritability() experimental: Extract heritability estimates
breeding_values() EBV() BLUP() experimental: Extract breeding values
prediction_error_variance() experimental: Extract prediction error variances
reliability() accuracy() experimental: Extract reliability and accuracy estimates
fixef() experimental: Extract fixed effects
ranef() experimental: Extract random effects
fit_diagnostics() experimental: Inspect fitted-model diagnostics
predict(<hsquared_fit>) fitted(<hsquared_fit>) residuals(<hsquared_fit>): Response-scale prediction helpers

Repeatability and independent extra effects

Outputs of the opt-in repeatability, two_effect, and multi_effect targets. The common-environment and arbitrary-N legs are covered; the repeatability and maternal-genetic legs are experimental. Non-animal blocks report variance-explained proportions, not heritabilities.

repeatability() experimental: Extract repeatability estimates
permanent_effects() experimental: Extract permanent-environment effects
common_env_effects() experimental: Extract common-environment effects
maternal_effects() experimental: Extract maternal genetic effects
common_env_proportion() experimental: Extract the common-environment variance ratio (c2)
maternal_proportion() experimental: Extract the maternal variance ratio (m2)

Direct–maternal and random regression

Outputs of the opt-in direct_maternal and random_regression targets, both covered at validation scale (k = 2 only for random regression). Direct h² is not total heritability; heritability() returns the labelled Willham triple for direct–maternal fits.

direct_heritability() direct_variance() partner_variance() direct_maternal_covariance() total_heritability() experimental: Direct-maternal correlated model extractors
rr_covariance() random_coefficients() rr_genetic_variance() rr_heritability() rr_correlation() rr_eigenfunctions() experimental: Random-regression (reaction-norm) extractors

Visualization

Figures carry a machine-readable hsquared_meta honest-status attribute. Uncertainty displays are not calibrated-precision statements.

autoplot(<hsquared_fit>) autoplot(<hs_gwas>) experimental: ggplot2 visualizations for hsquared results
plot(<hsquared_fit>): Diagnostic plots for a fitted animal model
theme_hsquared() experimental: hsquared ggplot2 theme
hs_recovery_forest() experimental: Forest plot of a known-truth recovery study

Uncertainty (experimental)

Asymptotic/delta-method standard errors, intervals, and tests, surfaced when the engine provides them. No route has nominal coverage calibration; named univariate-pedigree methods are directional-conservative only. confint() and vcov() are deliberately absent.

heritability_interval() experimental: Extract an experimental heritability confidence interval
variance_component_standard_errors() heritability_standard_error() experimental: Extract experimental variance-component and heritability standard errors
covariance_standard_errors() experimental: Extract experimental multivariate covariance standard errors
covariance_structure_lrt() experimental: Likelihood-ratio test for genetic covariance structure
repeatability_interval() experimental: Extract an experimental repeatability confidence interval
common_env_proportion_interval() maternal_proportion_interval() experimental: Extract an experimental common-environment / maternal variance-ratio interval
confint(<hsquared_fit>) vcov(<hsquared_fit>) profile(<hsquared_fit>) anova(<hsquared_fit>): Block unsupported likelihood-inference helpers

Multivariate (experimental)

Outputs of the opt-in multivariate and metafounder targets. Experimental: REML-only, animal-model-only, dense/validation-scale.

genetic_covariance() G_matrix() residual_covariance() R_matrix() genetic_correlation() residual_correlation() experimental: Extract multivariate covariance and correlation matrices
gamma_matrix() experimental: Extract a supplied metafounder Gamma matrix
metafounder_groups() experimental: Extract supplied metafounder group assignments
evolvability() respondability() conditional_evolvability() autonomy() mean_evolvability() g_max() variance_along_gradient() eigen_G() experimental: G-matrix geometry and evolvability

Reserved syntax

Formula markers the parser recognises so it can name the planned or opt-in path. They are not a finished genomic, QTL, or inheritance system.

genomic() single_step() markers() marker_scan() qtl_scan() experimental: Genomic and QTL formula markers
permanent() common_env() maternal_genetic() maternal_env() paternal_genetic() paternal_env() cytoplasmic() imprinting() dominance() epistasis() relmat() precision() group() unknown_parent_group() metafounder() inbreeding() experimental: Quantitative-genetic formula markers

Reserved extractors

Names reserved so help pages exist. Most currently error. gwas() is an experimental post-fit scan, not a formula-level QTL workflow.

metafounder_effects() experimental: Reserved metafounder effect extractor
genetic_loadings() specific_variance() latent_breeding_values() experimental: Reserved factor-analytic and G-matrix extractors
marker_effects() marker_variance_explained() qtl_table() gwas_table() eqtl_table() lod_scores() experimental: Extract planned marker, QTL, GWAS, and eQTL results
gwas() experimental: Post-fit relatedness-corrected marker scan (GWAS)

Package

hsquared-package: hsquared: R Interface for Julia-Backed Quantitative-Genetic Models