sensitivity-analyzer

# Sensitivity Analyzer ## Overview The Sensitivity Analyzer skill provides comprehensive capabilities for identifying critical inputs and understanding how model outputs respond to parameter changes. It supports both local (one-at-a-time) and global sensitivity analysis methods, enabling robust decision-making under uncertainty. ## Capabilities - One-at-a-time (OAT) sensitivity - Global sensitivity analysis (Sobol indices, Morris screening) - Tornado diagram generation - Spider plot creation - Parameter importance ranking - Threshold identification - Breakeven analysis - Scenario comparison ## Used By Processes - Monte Carlo Simulation for Decision Support - Multi-Criteria Decision Analysis (MCDA) - Prescriptive Analytics and Optimization - What-If Analysis Framework ## Usage ### One-at-a-Time (OAT) Analysis ```python # Define OAT analysis oat_config = { "base_case": { "price": 100, "volume": 10000, "cost": 60, "fixed_costs": 200000 }, "variations": { "price": {"range": [-20, 20], "step": 5, "unit": "%"}, "volume": {"range": [-30, 30], "step": 10, "unit": "%"}, "cost": {"range": [-15, 15], "step": 5, "unit": "%"}, "fixed_costs": {"range": [-10, 10], "step": 5, "unit": "%"} }, "output_variable": "profit" } ``` ### Global Sensitivity (Sobol Indices) ```python # Define Sobol analysis sobol_config = { "parameters": { "price": {"bounds": [80, 120], "distribution": "uniform"},...