Dataset¶

Load completed netCDF measurements and recover explicitly staged Zarr stores.

qanary.dataset.load(path)[source]¶

Load a qanary measurement dataset.

Wraps xarray.load_dataset() with engine selection appropriate for Qanary-generated files (automatic selection for .nc and the Zarr engine for .zarr) and clear error messages.

Parameters:

path (Union[str, Path]) – Path to a .nc file or .zarr directory created by a Qanary Measurement run.

Returns:

The fully-loaded (in-memory) xarray dataset.

Return type:

xr.Dataset

Raises:

DatasetNotFoundError – If path does not exist on disk.
DatasetLoadError – If xarray raises any error while opening the file.

qanary.dataset.convert(nc_path, *, overwrite=False)[source]¶

Convert the paired .zarr store for a measurement to netCDF.

A normal run exports its current in-memory dataset directly to netCDF and removes its temporary live checkpoint. If recovery instead leaves a .zarr directory beside the intended .nc file, this function can complete that conversion manually.

Expected layout on disk:

<stem>.nc      <- target output (may or may not exist yet)
<stem>.zarr/   <- live zarr store (must exist)

Parameters:

nc_path (Union[str, Path]) – Path to the (possibly missing) .nc file. The function derives the .zarr path by replacing the suffix.
overwrite (bool) – If True, overwrite an existing .nc file silently. Defaults to False.

Returns:

Resolved absolute path to the written .nc file.

Return type:

Path

Raises:

ZarrStoreNotFoundError – If the .zarr directory does not exist next to nc_path.
DatasetConvertError – If nc_path already exists and overwrite is False, or if conversion fails for another reason.