curryer.correction.results_io
=============================

.. py:module:: curryer.correction.results_io

.. autoapi-nested-parse::

   NetCDF I/O helpers for the correction pipeline.

   This module owns all read/write operations on the NetCDF result file:

   - :func:`_build_netcdf_structure` -- initialise the in-memory data dict.
   - :func:`_save_netcdf_results` -- write the final NetCDF output file.
   - :func:`_save_netcdf_checkpoint` -- write a partial checkpoint after each GCP pair.
   - :func:`_load_checkpoint` -- reload a checkpoint to resume an interrupted run.
   - :func:`_cleanup_checkpoint` -- delete the checkpoint file after a successful run.



Attributes
----------

.. autoapisummary::

   curryer.correction.results_io.logger


Functions
---------

.. autoapisummary::

   curryer.correction.results_io._build_netcdf_structure
   curryer.correction.results_io._save_netcdf_checkpoint
   curryer.correction.results_io._load_checkpoint
   curryer.correction.results_io._cleanup_checkpoint
   curryer.correction.results_io._save_netcdf_results


Module Contents
---------------

.. py:data:: logger

.. py:function:: _build_netcdf_structure(setup: curryer.correction.config.GeolocationSetup, sweep: curryer.correction.config.Sweep, netcdf_config: curryer.correction.config.NetCDFConfig, n_param_sets: int, n_gcp_pairs: int) -> dict

   Build NetCDF data structure dynamically from the sweep + output config.

   This creates the netcdf_data dictionary with proper variable names based on
   the parameters defined in the sweep, avoiding hardcoded mission-specific names.

   :param setup: GeolocationSetup (performance threshold for metric naming)
   :param sweep: Sweep with the parameters to vary
   :param netcdf_config: Resolved NetCDFConfig with metadata
   :param n_param_sets: Number of parameter sets (iterations)
   :param n_gcp_pairs: Number of GCP pairs

   :returns: Dictionary with initialized arrays for all NetCDF variables


.. py:function:: _save_netcdf_checkpoint(netcdf_data, output_file, setup, sweep, netcdf_config, pair_idx_completed)

   Save NetCDF checkpoint with partial results after each GCP pair completes.

   This enables resuming correction runs if they are interrupted.
   Adapted for pair-outer loop order where each pair processes all parameters.

   :param netcdf_data: Dictionary with current NetCDF data
   :param output_file: Path to final output file (checkpoint uses .checkpoint.nc suffix)
   :param setup: GeolocationSetup (performance threshold for metric naming)
   :param sweep: Sweep with parameters/iterations/seed
   :param netcdf_config: Resolved NetCDFConfig with metadata
   :param pair_idx_completed: Index of the last completed GCP pair (for pair-outer loop)


.. py:function:: _load_checkpoint(output_file)

   Load checkpoint if it exists and convert back to netcdf_data dict.

   :param output_file: Path to final output file (will check for .checkpoint.nc)

   :returns: Tuple of (netcdf_data dict, start_idx) or (None, 0) if no checkpoint


.. py:function:: _cleanup_checkpoint(output_file)

   Remove checkpoint file after successful completion.

   :param output_file: Path to final output file (will remove .checkpoint.nc)


.. py:function:: _save_netcdf_results(netcdf_data, output_file, setup, sweep, netcdf_config)

   Save results to NetCDF file using config-driven metadata.

   This function dynamically builds the NetCDF file structure from the
   netcdf_data dictionary, using configuration for all metadata rather
   than hardcoding mission-specific values.

   :param netcdf_data: Dictionary with all NetCDF variables and data
   :param output_file: Path to output NetCDF file
   :param setup: GeolocationSetup (performance threshold for metric naming)
   :param sweep: Sweep with parameters/iterations/seed
   :param netcdf_config: Resolved NetCDFConfig with metadata