DSJM: A Software Toolkit for Direct Determination of Sparse Jacobian Matrices. We describe the architecture and implementation of DSJM, a software toolkit written in portable C++ that enables direct determination of sparse Jacobian matrices. Our design exploits the recently proposed unifying framework ”pattern graph” and employs cache-friendly array-based sparse data structures. The pattern graph remains invariant for one-sided, two-sided, full column, and column-segments compression algorithms.The DSJM implements a greedy partitioning algorithm after the sparse matrix has been preprocessed with ordering heuristics for efficiency. In our numerical testing on 20 large-scale test instances (see Graph Models and their efficient implementation for sparse Jacobian matrix determination, Disc. Appl. Math. 161(2013) 1747-1754) we have found that DSJM consistently produced better timing and partitions compared with similar software.
Keywords for this software
References in zbMATH (referenced in 6 articles , 1 standard article )
Showing results 1 to 6 of 6.
- Hossain, Shahadat; Hakim Mithila, Nasrin: Pattern graph for sparse Hessian matrix (2018)
- Carter, Richard G.; Hossain, Shahadat; Sultana, Marzia: Efficient detection of Hessian matrix sparsity pattern (2016)
- Greuel, Gert-Martin (ed.); Koch, Thorsten (ed.); Paule, Peter (ed.); Sommese, Andrew (ed.): Mathematical software -- ICMS 2016. 5th international conference, Berlin, Germany, July 11--14, 2016. Proceedings (2016)
- Hasan, Mahmudul; Hossain, Shahadat; Khan, Ahamad Imtiaz; Mithila, Nasrin Hakim; Suny, Ashraful Huq: DSJM: a software toolkit for direct determination of sparse Jacobian matrices (2016)
- Gebremedhin, Assefaw H.; Nguyen, Duc; Patwary, Md. Mostofa Ali; Pothen, Alex: ColPack, software for graph coloring and related problems in scientific computing (2013)
- Hossain, Shahadat; Steihaug, Trond: Graph models and their efficient implementation for sparse Jacobian matrix determination (2013)