API Reference¶

This reference is generated from the docstrings in the source code via sphinx.ext.autosummary / sphinx.ext.autodoc. Each entry below links to a page with the full signature, parameter descriptions, and (for classes) a summary of public methods and attributes. Private helpers (names prefixed with _) are intentionally not part of this reference.

Main package¶

Importing geossm re-exports the main classes and functions from its submodules (geossm.covmodel, geossm.datasets, geossm.ssm, geossm.stmodel, plus geossm.data_preparation and geossm.utils) at the top level, e.g. geossm.LRStateSpaceModel is the same object as geossm.stmodel.LRStateSpaceModel. The sections below document each name once, grouped by the submodule it is defined in; geossm.__version__ holds the installed package version.

Models¶

State-space model¶

The full-rank linear-Gaussian state-space model: Kalman filtering, smoothing, simulation, and prediction.

`StateSpaceModel`	A linear-Gaussian state-space model with Kalman filtering, smoothing and simulation capabilities.
`StateSpaceResults`	Container for state-space model estimation results (filter and/or smoother outputs), with convenience methods for residuals, error metrics, confidence intervals, diagnostics and a printable summary.

Low-rank state-space model¶

geossm’s flagship model for large spatial domains: a state-space model whose latent spatial field is represented through a low-rank, SPDE-based covariance basis (see Covariance below).

`LRStateSpaceModel`	Low-Rank State-Space Model (LRSSM) for large-scale spatio-temporal data.
`LRStateSpaceResults`	Results of fitting a `geossm.stmodel.LRStateSpaceModel`.
`Param`	A single named model parameter, with its value, fixed/free status, and (once available) standard error.
`ModelParams`	The full set of parameters estimated by `fit()`.
`FitOptions`	Stopping-rule and logging options for `fit()`.

Covariance¶

Finite-element (SPDE) approximation of the Matérn covariance, and the mesh builders used to discretize the spatial domain it is defined on.

`spdeAppoxCov`	The SPDE approximation of the Matérn covariance model.
`FEMSolver`	Low-level finite-element engine behind spdeAppoxCov.
`buildMesh2d`	Build a 2D triangular (gmsh/pygmsh) mesh around a set of observed locations, in the spirit of R-INLA's `inla.mesh.2d()`.
`buildMesh2d_density`	Build a 2D triangular mesh whose interior vertices are a density- matched subset of points, rather than free vertices placed by a smooth size field (as buildMesh2d does).
`buildMesh2d_new`	Simplified front door to buildMesh2d_density's density-matched mesh: picks a landmark-placement method instead of juggling snap_to_points/relax_iters directly, and leaves every other knob (min_edge, offset, cutoff, tol, max_iter) at buildMesh2d_density's own defaults, which are self-tuning (derived from lowrank, the interest domain, and max_edge) and rarely need hand-adjustment.
`buildMeshGrid2d`	Build a regular (structured) 2D triangular mesh: a lattice of nx by ny vertices spanning the bounding box of domain (or the convex hull of points if domain is not given), extended by offset.

Spatial data preparation¶

Turning a geopandas.GeoDataFrame plus a model formula into the design matrices consumed by the models above, and the datasets bundled with the package.

`DesignMatrices`	Container for the response and covariate design matrices of a spatio-temporal dataset, organized on an [N x T] (site x time) grid, together with the spatial/temporal/formula metadata needed to interpret and reproduce them.
`DesignMatricesBuilder`	Turn a spatio-temporal `geopandas.GeoDataFrame` plus a Patsy/R-style model formula into a `DesignMatrices` object (response + covariates organized on an [N x T] site x time grid), ready to be consumed by geossm's state-space models.
`check_regular_timestamps`	Check whether a collection of timestamps is regularly spaced.

`load_dataset`	Load one of the example datasets bundled with geossm.
`list_datasets`	List available example datasets bundled with the package.

Utilities¶

General-purpose helpers shared across the package.

`block_diag_3D`	Stack 3D arrays into a single 3D block-diagonal array.
`write`	Append a run log (observation grid, model, and hardware info) to a text file.
`getHardware`	Build a human-readable summary of the current machine's hardware.
`KeyStream`	Stateful helper that hands out fresh JAX PRNG keys on demand.