Spark GraphX in Action

This is the Chinese translation of Manning's Spark GraphX in Action, translated by Shi Jinkui and Huang Guangyuan and published by Publishing House of Electronics Industry in 2017. Based on Spark GraphX 1.6, it is an introductory-to-intermediate book on graph computing that requires neither mastery of Scala nor prior GraphX experience—basic programming ability is enough to follow along. The book has 10 chapters in three parts. Part 1 (Chapters 1–3) covers prerequisites: one chapter explains the relationship among big data, Hadoop, Spark, and graph computing, clarifying property graphs, graph partitioning, graph parallelism, and data parallelism; then Spark Shell is used to demonstrate interactive, quick-start GraphX, with basics of Scala functional programming, RDDs, and graph terminology filled in. Part 2 (Chapters 4–7) gets into GraphX itself: core classes such as Vertex and Edge and core operations such as map, Map/Reduce, and Pregel; graph serialization and generation; built-in algorithms including PageRank, triangle counting, shortest paths, connected components, strongly connected components, and label propagation; then classic algorithms such as weighted shortest paths, the greedy traveling salesman, and minimum spanning trees show how to implement your own algorithms on top of GraphX; the machine learning chapter demonstrates GraphX with MLlib through SVDPlusPlus movie recommendation, LDA topic modeling, logistic regression spam filtering, and power-iteration clustering for image segmentation. Part 3 (Chapters 8–10) discusses operations GraphX lacks (subgraphs, graph union, reading RDF, global clustering coefficient), performance monitoring and tuning (caching, checkpointing, serialization, graph partitioning), and peripheral tools such as calling GraphX from Java, Zeppelin + d3.js visualization, Spark JobServer, and GraphFrames. Four appendices cover installing Spark, Gephi visualization, further resources, and Scala tips, making the book suitable for hands-on reading.