{"categories":[{"label":"Mathematics","url":"https://skillfed.io/packages/category/scientific-engineering-mathematics"}],"enrichment":{"capability":"Provides a Python interface to QDLDL, a free LDL factorization routine for solving linear systems with quasi-definite matrices in sparse format.","skillfed_tags":["sparse-linear-algebra","matrix-factorization","numerical-computing"],"use_cases":["Solve multiple linear systems with the same quasi-definite matrix by factorizing once and reusing the factorization.","Implement interior-point optimization algorithms that require repeated solves with slowly-changing sparse matrices.","Perform LDL decomposition of indefinite symmetric matrices in control theory or signal processing applications.","Update factorizations in iterative solvers without recomputing from scratch when only matrix values change.","Integrate specialized sparse linear algebra into larger numerical pipelines that use scipy."],"what_it_does":"QDLDL is a Python binding to a specialized linear algebra routine that factors quasi-definite matrices using LDL decomposition. It takes sparse matrices in scipy's CSC format and computes a factorization that can be reused to solve multiple linear systems efficiently. The package is designed for problems where the matrix structure is sparse and quasi-definite (a generalization of symmetric indefinite matrices), which arise in optimization, control theory, and other numerical applications.\n\nThe solver provides two main operations: initial factorization via the Solver constructor, and subsequent solves via the solve() method. It also supports updating the factorization when the matrix values change but the sparsity pattern remains fixed, which is useful for iterative algorithms. The implementation handles format conversions internally (e.g., converting to upper triangular form) and depends on numpy and scipy for matrix representation and numerical operations.","worth_installing":"Yes, if you need to solve linear systems with quasi-definite sparse matrices and want a specialized, actively-maintained solver. The medium install friction is typical for compiled numerical packages. Apache 2.0 licensing poses no restrictions. No known security vulnerabilities. Best suited for optimization and numerical computing workflows; not needed for general-purpose linear algebra."},"id":"qdldl","links":{"html":"https://skillfed.io/packages/qdldl","md":"https://skillfed.io/packages/qdldl.md","pypi":"https://pypi.org/project/qdldl/"},"maintenance":{"status":"active"},"meta":{"latest_release":"2026-02-19","license_spdx":null,"license_treatment":"permissive","name":"qdldl","python_support":"unspecified","summary":"QDLDL, a free LDL factorization routine."},"popularity":{"monthly_downloads":1962872,"position":3407,"tier":"top_5000"},"security":{"n_vulnerabilities":0},"version":"0.1.9.post1"}