LinearGaussianObservation#
- class hybrid_aif.LinearGaussianObservation[source]#
Bases:
Moduley = H x + c + eps: a linear-Gaussian modality over one Gaussian state factor.Construct with
H=/c=to haveheadassembled.H,cand the noise are each either shared ((d_obs, d)/(d_obs,)/(d_obs, d_obs)) or per-mode, with leading parent axesparent_shape(the categorical parents named after the observed state factor inA_deps), broadcast like the per-modeF_jofLinearGaussianTransition. Every inference consumer reads theR_cholproperty, which lowers a distributional noise by its expected precision.- head: Array | MatrixNormal#
The augmented
[H | c]:(d_obs, d + 1), or(*parent_shape, d_obs, d + 1)per mode; a point value or aMatrixNormalposterior over it (never separate posteriors overHandc, which are coupled).
- noise: Array | InverseGamma | InverseWishart | Parameter#
A point
R_chol(lower triangular(d_obs, d_obs), or(*parent_shape, d_obs, d_obs)per mode), or the conjugate posterior overR(InverseGamma,InverseWishart), which makes it learnable and samplable; orParameter(name), reading a noise held inSystemParams.parameters(one noise several modalities or transitions read; design: process-noise-learning.md D4).
- property R_chol: Array#
The noise as an
R_cholarray: the point value itself, or the expected-precision lowering of a distributional noise. A noise map has no one value: read it at a mean withR_chol_at().
- property noise_map: bool#
Capability, per-field: the noise is a
NoiseMapover the observed state, not a value or a posterior.
- property noise_levels: int#
How many leading level axes the noise map carries, one per categorical parent (zero for a plain map, a point or a posterior).
- R_chol_at(loc)[source]#
The noise as an
R_cholarray at a belief’s meanloc: a point or conjugate noise is the same everywhere; a noise map is evaluated there, the extended filter’s convention (design: planning-with-observations.md D3). A levelled map returns every level’s factor, leading(*levels, ...)like a per-modeR_chol.