src.llm_interpreter.results_aggregator

Results Aggregator for AMMM Model Outputs.

This module consolidates all model outputs (CSV files, JSON config, diagnostics) into a detailed JSON structure for LLM interpretation.

Module Contents

class src.llm_interpreter.results_aggregator.ResultsAggregator(results_dir: str = 'results')

Aggregates all model outputs into a detailed JSON structure.

aggregate_convergence_diagnostics() → Dict[str, Any]

Aggregate convergence diagnostics from model summary.

Returns:

Dictionary of convergence metrics.

aggregate_model_fit_metrics() → Dict[str, Any]

Aggregate model fit metrics including ELPD and R-squared.

Returns:

Dictionary of model fit metrics.

aggregate_channel_performance() → Dict[str, Any]

Aggregate marketing channel performance metrics.

Returns:

Dictionary of channel performance metrics.

aggregate_optimization_results() → Dict[str, Any]

Aggregate optimization results and recommendations.

Returns:

Dictionary of optimization results.

aggregate_model_configuration() → Dict[str, Any]

Aggregate model configuration and priors.

Returns:

Dictionary of model configuration.

aggregate_all_results() → Dict[str, Any]

Aggregate all results into a detailed JSON structure.

Returns:

Complete aggregated results dictionary.

save_aggregated_results(output_path: str | None = None) → str

Save aggregated results to JSON file.

Parameters:

output_path – Optional custom output path.

Returns:

Path to saved file.