aboutsummaryrefslogtreecommitdiffhomepage
path: root/Eigen/src/QR
Commit message (Collapse)AuthorAge
* Handle some Index to int conversions in BLAS/LAPACK support.Gravatar Gael Guennebaud2016-05-26
|
* Prevent crash in CompleteOrthogonalDecomposition if object was default ↵Gravatar Rasmus Munk Larsen2016-04-21
| | | | constructed.
* Removed executable bit from header filesGravatar Benoit Steiner2016-03-23
|
* Rename inverse -> pseudoInverse.Gravatar Rasmus Munk Larsen2016-02-10
|
* Enable inverse() method for computing pseudo-inverse.Gravatar Rasmus Munk Larsen2016-02-09
|
* Make applyZAdjointOnTheLeftInPlace protected.Gravatar Rasmus Munk Larsen2016-02-08
|
* Add missing calls to tests of COD.Gravatar Rasmus Munk Larsen2016-02-08
| | | | Fix a few mistakes in 3.2 -> 3.3 port.
* Nevermind.Gravatar Rasmus Munk Larsen2016-02-06
|
* Add my name to copyright notice in ColPivHouseholder.h, mostly for previous ↵Gravatar Rasmus Munk Larsen2016-02-06
| | | | work on stable norm downdate formula.
* Implement complete orthogonal decomposition in Eigen.Gravatar Rasmus Munk Larsen2016-02-06
|
* Fix bad line break. Don't repeat Kahan matrix test since it is deterministic.Gravatar Rasmus Munk Larsen2016-02-03
|
* Make the array of directly compute column norms a member to avoid allocation ↵Gravatar Rasmus Munk Larsen2016-02-03
| | | | in computeInPlace.
* Change Eigen's ColPivHouseholderQR to use the numerically stable norm ↵Gravatar Rasmus Munk Larsen2016-01-28
| | | | | | | | | | | | | | | | | | | | | | | | downdate formula from http://www.netlib.org/lapack/lawnspdf/lawn176.pdf, which has been used in LAPACK's xGEQPF and xGEQP3 since 2006. With the old formula, the code chooses the wrong pivots and fails to correctly determine rank on graded matrices. This change also adds additional checks for non-increasing diagonal in R11 to existing unit tests, and adds a new unit test with the Kahan matrix, which consistently fails for the original code. Benchmark timings on Intel(R) Xeon(R) CPU E5-1650 v3 @ 3.50GHz. Code compiled with AVX & FMA. I just ran on square matrices of 3 difference sizes. Benchmark Time(ns) CPU(ns) Iterations ------------------------------------------------------- Before: BM_EigencolPivQR/64 53677 53627 12890 BM_EigencolPivQR/512 15265408 15250784 46 BM_EigencolPivQR/4k 15403556228 15388788368 2 After (non-vectorized version): Benchmark Time(ns) CPU(ns) Iterations Degradation -------------------------------------------------------------------- BM_EigencolPivQR/64 63736 63669 10844 18.5% BM_EigencolPivQR/512 16052546 16037381 43 5.1% BM_EigencolPivQR/4k 15149263620 15132025316 2 -2.0% Performance-wise there seems to be a ~18.5% degradation for small (64x64) matrices, probably due to the cost of more O(min(m,n)^2) sqrt operations that are not needed for the unstable formula.
* Fix numerous doxygen shortcomings, and workaround some clang -Wdocumentation ↵Gravatar Gael Guennebaud2016-01-01
| | | | warnings
* Fix MKL compilation issueGravatar Gael Guennebaud2015-12-11
|
* Fix compilation of MKL support.Gravatar Gael Guennebaud2015-12-11
|
* Generalize matrix ctor and compute() method of dense decomposition to 1) ↵Gravatar Gael Guennebaud2015-09-07
| | | | limit temporaries, 2) forward expressions to nested decompositions, 3) fix ambiguous ctor instanciation for square decomposition
* bug #949: add static assertion for incompatible scalar types in dense ↵Gravatar Gael Guennebaud2015-03-13
| | | | end-user decompositions.
* New scoring functor to select the pivot.Gravatar Marc Glisse2015-03-03
| | | | This is can be useful for non-floating point scalars, where choosing the biggest element is generally not the best choice.
* Merged in chtz/eigen-indexconversion (pull request PR-92)Gravatar Gael Guennebaud2015-02-16
|\ | | | | | | | | | | | | | | | | | | | | | | bug #877, bug #572: Get rid of Index conversion warnings, summary of changes: - Introduce a global typedef Eigen::Index making Eigen::DenseIndex and AnyExpr<>::Index deprecated (default is std::ptrdiff_t). - Eigen::Index is used throughout the API to represent indices, offsets, and sizes. - Classes storing an array of indices uses the type StorageIndex to store them. This is a template parameter of the class. Default is int. - Methods that *explicitly* set or return an element of such an array take or return a StorageIndex type. In all other cases, the Index type is used.
| * Remove deprecated usage of expr::Index.Gravatar Gael Guennebaud2015-02-16
| |
* | bug #952: Missing \endcode made doxygen fail to build ColPivHouseholderQRGravatar Christoph Hertzberg2015-02-15
| |
| * Index refactoring: StorageIndex must be used for storage only (and locally ↵Gravatar Gael Guennebaud2015-02-13
| | | | | | | | when it make sense). In all other cases use the global Index type.
| * Merge Index-refactoring branch with default, fix PastixSupport, remove some ↵Gravatar Gael Guennebaud2015-02-13
| |\ | |/ |/| | | useless typedefs
* | bug #941: fix accuracy issue in ColPivHouseholderQR, do not stop ↵Gravatar Gael Guennebaud2015-01-30
| | | | | | | | decomposition on a small pivot
| * bug #877, bug #572: Introduce a global Index typedef. Rename Sparse*::Index ↵Gravatar Christoph Hertzberg2014-12-04
|/ | | | to StorageIndex, make Dense*::StorageIndex an alias to DenseIndex. Overall this commit gets rid of all Index conversion warnings.
* quieted more g++ warnings of the form: warning: typedef XXX locally defined ↵Gravatar Mark Borgerding2014-10-16
| | | | but not used [-Wunused-local-typedefs]
* Make constructors explicit if they could lead to unintended implicit conversionGravatar Christoph Hertzberg2014-09-23
|
* Remove deprecated code not used by evaluatorsGravatar Gael Guennebaud2014-09-18
|
* Optimization: enable cache-efficient application of HouseholderSequence.Gravatar Gael Guennebaud2014-09-04
|
* Refactoring of sparse solvers through a SparseSolverBase class and usage of ↵Gravatar Gael Guennebaud2014-09-01
| | | | the Solve<> expression. Introduce a SolveWithGuess expression on top of Solve.
* merge default and evaluator branchesGravatar Gael Guennebaud2014-03-12
|\
* | Port QR module to Solve/InverseGravatar Gael Guennebaud2014-03-11
| |
| * Make pivoting HouseholderQR compatible with custom scalar typesGravatar Christoph Hertzberg2014-02-25
|/
* Minor corrections in QR docs.Gravatar Jitse Niesen2013-12-31
|
* Applied patch from Richard JW Roberts, resolving bug #704Gravatar Christoph Hertzberg2013-12-21
|
* Fix FullPivHouseholderQR ctors for non squared fixed size matrix typesGravatar Gael Guennebaud2013-11-19
|
* merge with default branchGravatar Gael Guennebaud2013-11-05
|\
* | Add a rank method with threshold control to JacobiSVD, and make solve uses ↵Gravatar Gael Guennebaud2013-11-01
| | | | | | | | it to return the minimal norm solution for rank-deficient problems
* | Fix changeset 2702788da71303afa71e53dfa6542a560115425aGravatar Gael Guennebaud2013-11-01
| | | | | | | | for fixed size matrices
* | Fix bug #678: vectors of row and columns transpositions were not properly ↵Gravatar Gael Guennebaud2013-10-29
| | | | | | | | resized in FullPivQR
* | Make FullPivHouseholderQR::solve returns the least-square solution instead ↵Gravatar Gael Guennebaud2013-08-20
| | | | | | | | of aborting if no exact solution exist
| * Disable some shortcuts with nvccGravatar Gael Guennebaud2013-07-31
|/
* Fix various scalar type conversion warnings.Gravatar Gael Guennebaud2013-07-12
|
* Fix HouseholderSequence::conjugate() and ::adjoint() and add respective unit ↵Gravatar Gael Guennebaud2013-06-17
| | | | tests.
* Fix bug #314: move remaining math functions from internal to numext namespaceGravatar Gael Guennebaud2013-06-10
|
* Fix bug #599: add missing documentation of some members in QR module.Gravatar Gael Guennebaud2013-06-10
|
* Clean source code and unit tests with respect to -Wunused-local-typedefsGravatar Gael Guennebaud2013-04-10
|
* Add assertion on the input matrix size in factorizations relying on ↵Gravatar Gael Guennebaud2013-02-26
| | | | permutations of 32bits int
* Remove superfluous cast.Gravatar Gael Guennebaud2013-02-24
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-10-18 18:36:27 +0000