aiqclib.classify.step5_extract_features package

Submodules

aiqclib.classify.step5_extract_features.dataset_all module

This module provides the ExtractDataSetAll class, which is designed for extracting features from Copernicus CTD (Conductivity, Temperature, and Depth) datasets. It inherits from ExtractFeatureBase and utilizes a configuration-driven approach to define data targets and output paths.

class aiqclib.classify.step5_extract_features.dataset_all.ExtractDataSetAll(config, input_data=None, selected_profiles=None, selected_rows=None, summary_stats=None)[source]

Bases: ExtractFeatureBase

Feature extraction implementation specifically for Copernicus CTD data.

This class serves as a concrete implementation of the ExtractFeatureBase interface, specializing in the configuration and file naming conventions required for full CTD dataset processing.

Variables:

expected_class_name (str) – The identifier used to match this class with configuration settings.

Parameters:

  • config (ConfigBase)

  • input_data (DataFrame | None)

  • selected_profiles (DataFrame | None)

  • selected_rows (Dict[str, DataFrame] | None)

  • summary_stats (DataFrame | None)

default_file_name: str

Default string template for naming exported feature files.

drop_col_names: list[str]

List of columns to be excluded or dropped during the extraction process.

expected_class_name: str = 'ExtractDataSetAll'
normalization_role: str = 'apply'

At classification time, normalization values are loaded from the file produced during preparation rather than re-derived from data.

output_file_names: Dict[str, str]

Resolved mapping of target names to their specific output file paths.