cultivars.engine._core

Primitive layer: constants, value objects, contracts, and pure functions.

Nothing in this subpackage knows what a model is. Every symbol here has more than one consumer – that is the test for belonging in _core rather than in _internals, which houses mono-consumer mechanism plumbing.

Constants and defaults

_containers	Reusable value objects shared by internal mechanisms and public results.
_converters	Converters between the numpy core and the optional dataframe ecosystems.
_defaults	Package-wide numeric constants and estimator defaults.
_estimators
_loaders
_mappings
_matrices
_notes
_polynomials
_protocols	Structural contracts satisfied by fitted results and model specifications.
_recursions
_reparams	Unconstrained reparameterization of stationary autoregressive coefficients.
_rotations	Rotation sampling for set-identified structural models.
_samplers	Sampling primitives: single conditional draws with no recursion over data.
_scales
_scores	Scoring primitives: exact functionals of a predictive sample.
_spectra	Frequency-domain primitives: grids, transfer functions, spectral matrices.
_transforms	Stationarity-inducing data transforms and their inverses.
_types	Public type aliases for categorical specification arguments.
_validators	Input validation primitives.
_wavelets	Wavelet primitives: MODWT pyramid, multiresolution synthesis, Morlet transform, smoothing.