.. _programmatic_usage:

Running from Python
====================

The entry script (``__main__.py``) does this for you, but you can drive the pipeline directly if
you need finer control (e.g. from a notebook or a custom orchestration script):

.. code-block:: python

   import logging
   from pathlib import Path

   from kaxanuk.attribution_analysis.config_handlers.excel_configurator import ExcelConfigurator
   from kaxanuk.attribution_analysis.performance_attribution import main
   from kaxanuk.attribution_analysis.services.configuration_logger import configure_logger
   from kaxanuk.attribution_analysis.services.env_loader import load_config_env

   # Load Config/.env (e.g. FRED_API_KEY)
   load_config_env()

   configure_logger(
       logger_name="kaxanuk.attribution_analysis",
       logger_level=logging.INFO,
       logger_format="[%(levelname)s] %(message)s",
       logger_file=None,
   )

   # Load and validate the Excel configuration
   configurator = ExcelConfigurator(Path("Config") / "attribution_analysis_parameters.xlsx")
   configuration = configurator.get_configuration()

   # Run the full attribution pipeline
   main(
       configuration,
       launch_dashboard=True,
       dashboard_port=configurator.get_dashboard_port(),
   )

See :ref:`api_reference` for the full reference on ``main()``, ``ExcelConfigurator``, and the
other public entry points used above.