GraphFrames Project Blog https://graphframes.io/05-blog/ Mon, 30 Mar 2026 00:00:00 +0000 ssinchenko@apache.org (Sem Sinchenko) https://graphframes.io/05-blog/1000-graphframes-is-back.html fbf59b6a-83f0-3f8f-8ce3-7d4fe5468e81 Fri, 01 Aug 2025 00:00:00 +0000 GraphFrames 0.9.2 is out on PyPi as graphframes-py and as io.graphframes on Maven Sonatype Central! Documentation is now available on graphframes.io… and we even have a new logo! https://graphframes.io/05-blog/997-graphframes-011-release.html 87f2ba15-cc54-3819-97af-80a1135a78f1 Mon, 30 Mar 2026 00:00:00 +0000 This release brings a new Connected Components algorithm based on Randomized Contraction, automatic Pregel optimization that skips unnecessary joins, graph embeddings via random walks with Word2Vec and Hash2Vec, a PySpark Property Graph API, approximate triangle counting with DataSketches, and various improvements. https://graphframes.io/05-blog/998-graphframes-010-release.html 529f18e2-aa30-31ff-b98b-0a890df90767 Sat, 11 Oct 2025 00:00:00 +0000 This release comes with significant performance improvements to most algorithms, as well as fixed memory leaks. The PySpark APIs for Spark Connect and Spark Classic are now synchronized with the Scala core, allowing PySpark users to benefit from the latest improvements in the GraphFrame APIs and configurations. This is the first release in which GraphFrames relies on its own internal fork of GraphX instead of Spark's built-in version. There are also improvements in motif finding. Undirected, bidirectional, and an arbitrary amount of edges can now be included in the pattern string. New algorithms for cycle detection, maximal independent set, k-Core were added. Shortest Path algorithm can now consider graph as undirected and return any path instead of directed only. A new set of methods to compute degree by edge type was added. The documentation has also been significantly improved. https://graphframes.io/05-blog/999-graphframes-093-release.html 7c694d1a-22dd-3fcf-8bf0-774fb37b4e45 Wed, 10 Sep 2025 00:00:00 +0000 This release comes with redesigned documentation and improvements. A new PropertyGraph model was introduced. An option has been added to use local checkpoints in DataFrame-based Pregel and Connected Components. A bug in DataFrame-based LabelPropagation was fixed. Benchmarks for GraphFrames have been introduced. There is better compatibility with Scala 3. There are also minor updates and improvements.