model.runtime.loaders.base_data_loader

Shared low-level loading helpers for ADELM data loaders.

Module Contents

Classes

BaseDataLoader

Base class for ADELM data loaders.

Data

__all__

API

class model.runtime.loaders.base_data_loader.BaseDataLoader

Base class for ADELM data loaders.

Subclasses reuse the NetCDF opening, coordinate decoding, tensor conversion, and ADELM variable-mapping helpers defined here.

static open_netcdf(path)

Open a NetCDF4 dataset.

static read_site_ids(ds)

Return the site coordinate as a list of strings.

static read_time_values(ds)

Return the time coordinate as a numpy datetime64 array.

static to_tensor(array)

Convert a numpy / masked array to a float32 torch.Tensor.

static collect_mapped_nc_vars(mapping)

Collect all NetCDF variable names referenced in a mapping dict.

classmethod load_variables(ds, expected_dims_per_var, registry_keys=None, only_vars=None)

Read data variables from an open dataset.

static apply_mapping(data, mapping)

Select and rename variables using an ADELM-name -> NC-name mapping.

classmethod apply_layered_mapping(data, section_mapping, num_soil_layers)

Apply a mapping section that may contain both scalar and layer-wise variables.

model.runtime.loaders.base_data_loader.__all__

[‘BaseDataLoader’]