Big Data

How to study

• Chapters are in dependency order: read them from beginning to end. Each chapter consolidates, in one place, everything the lectures said about a topic.
• The course has four parts. Foundations (ch. 1–2) frame the field; Storage (ch. 3) gives you a place to put bytes; Batch processing (ch. 4–9) is the heart of the course, from the MapReduce era (ch. 4–5) through the Spark era (ch. 6–8) to designing parallel algorithms (ch. 9); Stream processing (ch. 10) closes with the near-real-time lane and the examinable theory of sketches and windows.
• Chapter 9 is the bridge chapter: once the framework is fast, the hard part is rethinking algorithms for it — Amdahl’s law, prefix sum, and the long line of parallel frequent-itemset-mining formulations. Read it as engineering trade-offs, not code.
• The widgets (the Amdahl calculator, the window simulators, the sketch explorers, the quizzes) are reconstructed from the original lessons: use them actively, not just as reading material.
• Every chapter ends with a Test your knowledge section and lists at the end the deck it was reconstructed from; the page-per-date lessons no longer exist but remain in the git history of the repo.