Time Series Analysis with Python Cookbook, Second Edition

A Practical Cookbook for Modern Time Series Work with Python 📊

The second edition of Time Series Analysis with Python Cookbook offers a recipe-based approach to the entire time series workflow. Tarek A. Atwan’s updated guide walks through data ingestion, cleaning, forecasting, anomaly detection, and deep learning with realistic examples.

What This Book Covers

Each recipe focuses on a concrete task you can apply immediately. The book opens with reading time series data from CSV, Excel, Parquet, SQL databases, Snowflake, and cloud storage. From there, it moves into date/time handling, missing data imputation, and outlier detection using statistical methods.

Later chapters explore exploratory analysis, stationarity, autocorrelation, exponential smoothing, ARIMA, Prophet, VAR, GARCH, and automated forecasting libraries. The final third of the book shifts to supervised machine learning and deep learning with scikit-learn, Keras, Darts, NeuralForecast, and Transformers.

Updated for the Second Edition

• New and expanded recipes for cloud data engineering, Snowflake, and modern Python libraries
• End-to-end coverage from data preparation to advanced forecasting
• Practical code examples you can adapt to real projects

Who Will Find This Book Useful

Data scientists, analysts, and engineers working with time-dependent data will appreciate the hands-on recipes. The book balances conceptual explanations with code, making it suitable for practitioners who need reliable solutions for forecasting and anomaly detection.

If you are building your first forecasting model or refining a production pipeline, this cookbook gives you actionable techniques to move from raw time series data to reliable insights.