## gllvmTMB licensing
##
## gllvmTMB is a standalone Template Model Builder (TMB) package for
## stacked-trait, long-format multivariate generalised linear latent
## variable models (GLLVMs). Except for the components and source lineages
## recorded below, gllvmTMB-specific code (the multi-trait engine, formula
## parser, covariance keyword grid in this release snapshot, response-family
## extensions, diagnostics, and extractors) is licensed GPL-3 by Shinichi
## Nakagawa.
##
## Family constructors (R/families.R)
##
## The 0.7 family-constructor layer is an independently implemented,
## data-driven gllvmTMB design by Shinichi Nakagawa. It is built from the public
## R family-object contract and `stats::make.link()` semantics. The earlier
## development implementation derived from sdmTMB/glmmTMB family code was
## replaced in this 0.7.1 source. The earlier implementation is not present
## in this source tree. Compatibility is guarded by formals, object-field,
## inverse-link, validation, delta, mixed-family, and compiled-engine tests.
##
## Spatial helper lineage (R/mesh.R, R/crs.R, and R/plot.R)
##
## The original gllvmTMB spatial helper layer was derived from sdmTMB's GPL-3
## implementation. The current helpers were substantially rewritten in place
## against the public fmesher/sf APIs and the published SPDE/GMRF construction.
## gllvmTMB therefore retains conservative sdmTMB attribution rather than
## claiming clean-room authorship. The source baseline is sdmTMB commit
## 8bc452cb743a9efa7b3212653a0f5f2bda3c5ffc (version 0.7.4.9024):
## https://github.com/sdmTMB/sdmTMB/tree/8bc452cb743a9efa7b3212653a0f5f2bda3c5ffc
##
## sdmTMB is GPL-3. Its named authors at that revision include Sean C.
## Anderson, Eric J. Ward, Philina A. English, Lewis A. K. Barnett, and James
## T. Thorson. The DESCRIPTION at the pinned source revision is available at
## https://github.com/sdmTMB/sdmTMB/blob/8bc452cb743a9efa7b3212653a0f5f2bda3c5ffc/DESCRIPTION.
## No sdmTMB runtime or build dependency is required by gllvmTMB.
##
## TMB engine (src/gllvmTMB.cpp)
##
## Except for the same-author drmTMB helpers recorded below, the C++ template
## under src/gllvmTMB.cpp is original gllvmTMB-native code by Shinichi
## Nakagawa, written against the TMB API (Kristensen et al. 2016,
## https://github.com/kaskr/adcomp). TMB is a runtime dependency (Imports +
## LinkingTo in DESCRIPTION) rather than included source.
##
## The reduced-rank covariance interface and its standard identified
## factor-loading parameterisation were informed by glmmTMB's `rr()` model.
## The current cursor-based lower-triangular unpacker is independently written
## for gllvmTMB; no glmmTMB C++ source is included in this engine.
##
## The stable cumulative-logit helpers `gll_log_inv_logit()`,
## `gll_log1m_inv_logit()`, `gll_log1mexp()`, and
## `gll_log_inv_logit_diff()` are copied verbatim from the sister package
## drmTMB (`src/drm_numeric.h`). Both packages are GPL-3 and both are authored
## by Shinichi Nakagawa; this is same-author GPL-3 reuse.
##
## The broader missing-predictor implementation is also same-author GPL-3
## reuse from drmTMB. It includes the R-side grammar and validators in
## R/missing-predictor.R, the Gaussian/binary/ordered/categorical and grouped
## missing-predictor blocks in src/gllvmTMB.cpp, and their focused tests under
## tests/testthat/test-missing-predictor-*.R. The gllvmTMB code adapts those
## pieces to stacked traits and the package's multi-tier parameter map.
##
## Standalone VA research template (inst/tmb/gllvmTMB_va_r3.cpp)
##
## The standalone variational research template is original gllvmTMB GPL-3
## code by Shinichi Nakagawa. Its dense-tier factor loadings use an
## independently written cursor-based implementation of the same standard
## identified lower-triangular parameterisation; no glmmTMB C++ source is
## included in this template.
##
## Standalone EVA research template (inst/tmb/gllvmTMB_eva.cpp)
##
## The separate extended-variational research template is also original
## gllvmTMB GPL-3 code by Shinichi Nakagawa. Its identified factor loadings are
## unpacked by an independently written cursor traversal; no glmmTMB C++
## source is included in this template.
##
## Upstream copyright holder of TMB:
##
##   * Kasper Kristensen (TMB author, https://github.com/kaskr/adcomp)
##
## TMB itself is GPL-2/GPL-3 licensed. See the upstream TMB
## DESCRIPTION at https://github.com/kaskr/adcomp for the full TMB
## author list.
##
## Sparse phylogenetic precision builder (R/phylo-tree-precision.R)
##
## The tree -> sparse A^{-1} construction in R/phylo-tree-precision.R
## (`.gllvm_phylo_tree_precision()` and its validators
## `.gllvm_validate_phylo_tree()` / `.gllvm_validate_phylo_species()` /
## `.gllvm_phylo_node_depths()` / `.gllvm_phylo_augmented_node_labels()`)
## is ported from the univariate sister package drmTMB
## (R/phylo-utils.R: `drm_phylo_augmented_precision()` and helpers).
## It uses only `ape` + `Matrix` and replaces the earlier
## MCMCglmm::inverseA() call on the `phylo_latent(tree = ...)` path, so
## MCMCglmm is no longer required to turn a tree into A^{-1}. Both
## gllvmTMB and drmTMB are GPL-3 by Shinichi Nakagawa (same-author
## reuse); the algorithm is the standard Hadfield & Nakagawa (2010)
## sparse phylogenetic inverse.
##
## Sparse pedigree precision builder (R/pedigree-precision.R)
##
## The pedigree -> sparse A^{-1} construction in R/pedigree-precision.R
## (`.gllvm_pedigree_precision()` and its helpers
## `.gllvm_standardize_pedigree()` / `.gllvm_normalize_pedigree_parent()` /
## `.gllvm_pedigree_topological_order()` /
## `.gllvm_pedigree_additive_relationship()`) reuses the pedigree
## standardisation, missing-parent normalisation, topological ordering,
## and dense tabular relatedness method ported from the univariate sister
## package drmTMB (R/phylo-utils.R: `drm_standardize_pedigree()`,
## `drm_normalize_pedigree_parent()`, `drm_pedigree_topological_order()`,
## `drm_pedigree_additive_relationship()`), which never depended on
## MCMCglmm. The sparse A^{-1} is then assembled on top via the standard
## Henderson (1976) / Quaas (1976) inverse rules. It uses only `Matrix`
## and replaces the earlier MCMCglmm::inverseA() call in
## `pedigree_to_Ainv_sparse()`, so MCMCglmm is no longer required to turn
## a pedigree into A^{-1}. Both gllvmTMB and drmTMB are GPL-3 by Shinichi
## Nakagawa (same-author reuse). The result is validated to reproduce
## `MCMCglmm::inverseA()` (Hadfield 2010) exactly, including inbreeding.
##
## Other same-author and conceptual lineage
##
## The loading-profile plotting design in R/loading-profile.R follows the
## sister drmTMB profile-plot pattern; both packages are GPL-3 by Shinichi
## Nakagawa. R/plot-gllvmTMB.R acknowledges the gllvm2lev plotting interface as
## conceptual inspiration and is independently implemented. The production
## logo at man/figures/logo.png was restored from Shinichi Nakagawa's legacy
## gllvmTMB repository. On 2026-08-08 the maintainer explicitly confirmed that
## gllvmTMB has the right to redistribute this logo. This records the package's
## redistribution basis and does not assert a creator identity that is absent
## from the repository record.
##
## lifecycle documentation badges (MIT)
##
## The files man/figures/lifecycle-deprecated.svg,
## man/figures/lifecycle-superseded.svg, and
## man/figures/lifecycle-experimental.svg are exact documentation assets from
## lifecycle 1.0.5. Copyright (c) 2023 lifecycle authors; Posit Software, PBC
## is the upstream copyright holder. They are distributed under the MIT
## licence:
##
## Permission is hereby granted, free of charge, to any person obtaining a copy
## of this software and associated documentation files (the "Software"), to
## deal in the Software without restriction, including without limitation the
## rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
## sell copies of the Software, and to permit persons to whom the Software is
## furnished to do so, subject to the following conditions:
##
## The above copyright notice and this permission notice shall be included in
## all copies or substantial portions of the Software.
##
## THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
## IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
## FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
## AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
## LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
## OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
## THE SOFTWARE.
##
## Generated package fixtures and visual snapshots
##
## The RDS/JSON files under inst/extdata are synthetic or package-generated
## gllvmTMB fixtures produced by the corresponding data-raw scripts and package
## validation records. The SVG files under tests/testthat/_snaps are generated
## visual regression snapshots of gllvmTMB-owned plots. They are distributed
## with the package under GPL-3 by Shinichi Nakagawa.