Monthly summary for sympy/sympy (March 2026): Delivered core enhancements in group-theory utilities with a focus on correctness, performance, and test coverage. Key features delivered: kernel computation for homomorphisms with permutation group codomains (including infinite domains); enhanced PermutationGroup.presentation to respect eliminate_gens for precise presentations. Major bug fixes: corrected is_normal logic for abelian subgroups and added tests to ensure correct normality behavior. Stability and performance: internal improvements across permutation-related utilities, including cleanup of cyclic_form, removal of unused imports, deep copies to prevent aliasing, and strengthened data integrity. Business value: expanded computational capabilities for algebraic research, improved reliability of SymPy's group-theory tooling, and reduced risk of incorrect results in related workflows.

February 2026 performance summary: Delivered a set of engineering enhancements across algebraic tooling and improved data ingestion reliability in pandas. Key kernel computation enhancements for GroupHomomorphism were implemented and optimized, enabling reliable kernel extraction across surjective maps and FpGroups with caching, improved error messages, and integration of the new prod function in _compute_kernel. Added a factorization method for GroupHomomorphism and corresponding tests to expand algebraic capabilities. FreeGroup utilities were improved with a pow optimization and a new prod method for product of words, supported by tests. For permutation-related computations and Molien series, the codebase now uses prod where appropriate, includes a short path to __pow__, and returns a Molien rational function with tests; a TODO was added to extend _compute_kernel to PermutationGroup. Ongoing refactor and quality improvements tightened code paths, added typing improvements, and strengthened test coverage. In pandas, read_excel infinite loop risk was eliminated with a guard when skiprows returns all rows, and JsonReader nrows handling for chunked JSON was hardened, improving reliability in data ingestion pipelines.

January 2026 focused on strengthening correctness and stability of core algebraic primitives in sympy/sympy, with targeted feature work across FreeGroup, permutation groups, and symbolic linear algebra, underpinned by expanded test coverage. The month produced a robust foundation for downstream users building advanced mathematics and algorithms on top of SymPy, while reducing edge-case failures and improving reliability under symbolic conditions.