DoWhy

소개

DoWhy MCP v2.0 is a complete rewrite designed to equip LLM agents with sophisticated causal inference tools via the Model Context Protocol (MCP). Built on foundational theories like Structural Causal Models, Graphical Causal Models, Potential Outcomes, and Do-Calculus, it ensures scientific rigor and theoretical depth in its analyses. This version focuses on providing robust statistical inference, comprehensive sensitivity analysis, and a complete suite of 42 specialized causal tools for a wide range of applications, ensuring that rigorous science meets practical application.

주요 기능

• Theoretical rigor, with every method backed by solid causal inference theory
• Advanced visualization for causal graphs, attribution plots, and diagnostic charts
• 0 GitHub stars
• Rigorous statistical inference with true Bootstrap confidence intervals
• Comprehensive sensitivity analysis with a full suite of refutation tests and E-value analysis
• Complete causal toolkit offering 42 specialized functionalities for various analyses

사용 사례

• Estimating causal effects of interventions, treatments, or policies on specific outcomes
• Performing root cause analysis to identify underlying drivers of anomalies or changes in data
• Conducting counterfactual 'what-if' scenario analysis to simulate outcomes under different conditions