hydromodpy.results#
Result storage for HydroModPy simulations (DuckDB + Zarr).
Hydrological metrics live in hydromodpy.core.metrics (canonical
location for nse, rmse, mae, kge, log_nse, bias,
pbias, correlation).
Modules
Input provenance fingerprinting for simulation reproducibility. |
|
Catalog facade for the results layer. |
|
Strong-typed return contracts for |
|
Pure-function derivations of secondary fields from primary outputs. |
|
Typed exceptions for the results / reader layer. |
|
FAIR export formats for HydroModPy simulations. |
|
Canonical field registry for HydroModPy simulation outputs. |
|
Workspace-level content-addressable geographic cache. |
|
Compatibility re-export for shared GeoParquet I/O helpers. |
|
Scalar spatial metadata for a simulation: cell size, shape, extent. |
|
Small HTML helpers shared by calibration and analysis report builders. |
|
Importers symmetric to |
|
Lazy loaders for the per-simulation Parquet and Zarr stores. |
|
Atomic pyarrow Parquet writer with canonical V1 options forced. |
|
PyArrow schemas for every per-simulation Parquet view. |
|
Variable read dispatch shared by |
|
Single-simulation view on the results catalog. |
|
Array / xarray provider bound to a single |
|
Capture host environment metadata for the |
|
Geographic and mesh accessors for |
|
Hydrographic-network accessors for |
|
Tabular and time-aligned accessors for |
|
Multi-simulation view on the results catalog. |
|
Point-in-cell lookup on unstructured meshes via Shapely STRtree. |
|
Shared result-storage layout contract. |
|
Diagnostics for the workspace result-storage layout. |
|
Time alignment helpers for observed and simulated series. |
|
LTTB downsampling for dense time series stored in the catalog. |
|
Lazy catchment-scale views computed on demand from a Run. |
|
On-the-fly derived fields computed from stored primary variables. |
|
Zarr v2 per-simulation store (CF-1.11 + ACDD-1.3 + UGRID-1.0). |