mathnet-numerics

Commit Graph

Author	SHA1	Message	Date
Christoph Ruegg	7c6d83464f	Build: .Net 3.5 variant of Numerics project in portable solution	13 years ago
Christoph Ruegg	7d11e73aaa	Inspections/CodeAnalysis Cleanup	13 years ago
Christoph Ruegg	01b54a5168	Drop obsolete code	13 years ago
Christoph Ruegg	5189e49fbd	Control: update obsolete attributes	14 years ago
Christoph Ruegg	d3dc54b4c1	Reworking common parallelization routines #92	14 years ago
Christoph Ruegg	386fb36bfc	Solution: Replace Silverlight with Portable library Instead of a specific Silverlight version, we now provide a separate Portable Library assebly that supports .Net 4.0.1 and higher, Silverlight 5 ad well as .Net for Metro style apps (WinRT). New NuGet Package MathNet.Numerics.Portable with the portable library. Since MathNet.Numerics.Portable now supports silverlight, the normal package MathNet.Numerics no longer supports it (.Net 4 only). Reworked the common parallel execution infrastructure for more reasonable abstractions and better compatibility with portable libraries. Portable libraries do support parts of the TPL, but neither the Parallel class nor straight threading.	15 years ago
Christoph Ruegg	7cced98dd2	Threading: CommonParallel now supports empty loops. Closes gh-19. CommonParallel did break on empty loops with an argument exception instead of just doing nothing. It now exits early if there's nothing to do in For and Invoke, no longer throwing an exception. Also, it now special cases single-element iterations and inlines their execution (trivial loop unrolling), skipping all the threading/task overhead.	15 years ago
Marcus Cuda	57da4a187e	vector parallelization tweaks	15 years ago
Marcus Cuda	89aa96eec8	started optimizing the parallel code for small matrices and vectors	16 years ago
Abratiychuk	3f470bc0bc	Fixed unit tests for Silverlight Signed-off-by: abratiychuk <18.05.2011>	16 years ago
Marcus Cuda	518cdf2fbd	fixed silverlight compile bug from last commit	16 years ago
Marcus Cuda	39ea72370a	fixed a problem were Parallel Invoke and Select where ignoring threading options from Control, there is still an issue with recursively calling invoke.	16 years ago
Marcus Cuda	2cf99d8184	commented out CommonParallel.Aggregate modified the affected code	16 years ago
Marcus Cuda	407ba19669	sparse matrix is no longer thread safe to improve performance	16 years ago
Marcus Cuda	d9bbaa4a90	refactored the Vector classes to use an intermediate, type specific class	16 years ago
Marcus Cuda	8585733257	la: added Andriy's complex32 LA code	16 years ago
Marcus Cuda	17d5941ea2	added andriy's single LA	16 years ago
Marcus Cuda	1a8cce73c6	Merged Andriy's additions as follows: Added optimized versions of L1, Frobenius, and infinity norms to the dense and sparse matrices Added a TransposeAndMultiply method to Matrix DenseMatrix, and SparseMatrix Ported the Gram-Schmidt QR factorization from dnAnalytics Ported the diagonal matrix from dnAnalytics Ported the iterative solvers from dnAnalytics	16 years ago
Marcus Cuda	44bded2c7b	added LA to siliverlight switched LA parallel to CommonParallel class	17 years ago
Christoph Ruegg	722b9736c4	Numerics: udated headers (sorry for the code churn..)	17 years ago
Marcus Cuda	fe5147fcda	using common parallel class for both silverlight and .NET 4.0	17 years ago

21 Commits (optimization-3)