Causal Data Science with Python

In the age of big data, correlation is everywhere – but causation is what truly drives understanding and decision-making. Causal Data Science with Python: From Correlation to Decision bridges the gap between predictive modeling and causal reasoning, offering a practical, hands-on guide to uncovering cause-and-effect relationships in data.
This book introduces the principles of causal inference and their implementation in Python, combining the rigor of statistics with the flexibility of modern machine learning. Through real-world examples and step-by-step coding exercises, readers learn to move beyond simple associations and make robust causal claims that support confident decisions in business, healthcare, economics, and the social sciences.
Key topics include counterfactual reasoning, randomized experiments, propensity score methods, instrumental variables, causal graphs (DAGs), mediation analysis, and machine learning for causal effect estimation. The text balances theory and practice, providing clear explanations of concepts such as the Rubin Causal Model, do-calculus, and Structural Causal Models (SCMs) – alongside Python implementations using libraries such as DoWhy, EconML, CausalML, and PyMC.
Whether you are a data scientist seeking to build fairer AI systems, a social scientist analyzing interventions, or a policymaker looking for evidence-based insights, this book offers the tools and reasoning framework to transform data into meaningful, actionable understanding.