mirror/eigen
2
0
Fork 0
mirror of https://gitlab.com/libeigen/eigen.git synced 2025-01-06 14:14:46 +08:00
Code Issues Packages Projects Releases Wiki Activity
7,521 Commits 11 Branches 89 Tags 168 MiB
3b614a2358
Commit Graph

4437 Commits

Author SHA1 Message Date
Rasmus Munk Larsen
acce4dd050 Change Eigen's ColPivHouseholderQR to use the numerically stable norm 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.
 2016-01-28 15:07:26 -08:00
Gael Guennebaud
b908e071a8 bug #178: get rid of some const_cast in SparseCore 2016-01-28 22:11:18 +01:00
Gael Guennebaud
c1d900af61 bug #178: remove additional const on nested expression, and remove several const_cast. 2016-01-28 21:43:20 +01:00
Gael Guennebaud
f50bb1e6f3 Fix compilation with gcc 2016-01-28 13:25:26 +01:00
Gael Guennebaud
ddf64babde merge 2016-01-28 13:21:48 +01:00
Gael Guennebaud
df15fbc452 bug #1158: PartialReduxExpr is a vector expression, and it thus must expose the LinearAccessBit flag 2016-01-28 13:16:30 +01:00
Gael Guennebaud
9bcadb7fd1 Disable stupid MSVC warning 2016-01-28 12:14:16 +01:00
Gael Guennebaud
b4d87fff4a Fix MSVC warning. 2016-01-28 12:12:30 +01:00
Gael Guennebaud
2bad3e78d9 bug #96, bug #1006: fix by value argument in result_of. 2016-01-28 12:12:06 +01:00
Benoit Steiner
291069e885 Fixed some compilation problems with nvcc + clang 2016-01-27 15:37:03 -08:00
Gael Guennebaud
4865e1e732 Update link to suitesparse. 2016-01-27 22:48:40 +01:00
Eugene Brevdo
c8d94ae944 digamma special function: merge shared code. 
Moved type-specific code into a helper class digamma_impl_maybe_poly<Scalar>.
 2016-01-27 09:52:29 -08:00
Gael Guennebaud
9c8f7dfe94 bug #1156: fix several function declarations whose arguments were passed by value instead of being passed by reference 2016-01-27 18:34:42 +01:00
Gael Guennebaud
9aa6fae123 bug #1154: move to dynamic scheduling for spmv products. 2016-01-27 18:03:51 +01:00
Gael Guennebaud
9801c959e6 Fix tri = complex * real product, and add respective unit test. 2016-01-27 17:12:25 +01:00
Gael Guennebaud
21b5345782 Add meta_least_common_multiple helper. 2016-01-27 17:11:39 +01:00
Gael Guennebaud
fecea26d93 Extend doc on shifting strategy 2016-01-27 15:55:15 +01:00
Gael Guennebaud
cfa21f8123 Remove dead code. 2016-01-26 23:33:15 +01:00
Gael Guennebaud
6850eab33b Re-enable blocking on rows in non-l3 blocking mode. 2016-01-26 23:32:48 +01:00
Gael Guennebaud
aa8c6a251e Make sure that micro-panel-size is smaller than blocking sizes (otherwise we might get a buffer overflow) 2016-01-26 23:31:48 +01:00
Gael Guennebaud
5b0a9ee003 Make sure that block sizes are smaller than input matrix sizes. 2016-01-26 23:30:24 +01:00
Christoph Hertzberg
44d4674955 bug #1153: Don't rely on __GXX_EXPERIMENTAL_CXX0X__ to detect C++11 support 2016-01-26 16:45:33 +01:00
Gael Guennebaud
8328caa618 bug #51: add block preallocation mechanism to selfadjoit*matrix product. 2016-01-25 22:06:42 +01:00
Gael Guennebaud
e58827d2ed bug #51: make general_matrix_matrix_triangular_product use L3-blocking helper so that general symmetric rank-updates and general-matrix-to-triangular products do not trigger dynamic memory allocation for fixed size matrices. 2016-01-25 17:16:33 +01:00
Gael Guennebaud
b114e6fd3b Improve documentation. 2016-01-25 11:56:25 +01:00
Gael Guennebaud
869b4443ac Add SparseVector::conservativeResize() method. 2016-01-25 11:55:39 +01:00
Gael Guennebaud
acf6f7af6b Merged in larsmans/eigen (pull request PR-156) 
Documentation fixes
 2016-01-24 22:28:49 +01:00
Lars Buitinck
cc482e32f1 Method is called visit, not visitor 2016-01-24 15:50:59 +01:00
Gael Guennebaud
1cf85bd875 bug #977: add stableNormalize[d] methods: they are analogues to normalize[d] but with carefull handling of under/over-flow 2016-01-23 22:40:11 +01:00
Gael Guennebaud
369d6d1ae3 Add link to reference paper. 2016-01-23 22:16:03 +01:00
Gael Guennebaud
0caa4b1531 bug #1150: make IncompleteCholesky more robust by iteratively increase the shift until the factorization succeed (with at most 10 attempts). 2016-01-23 22:13:54 +01:00
Gael Guennebaud
5358c38589 bug #1095: add Cholmod*::logDeterminant/determinant (from patch of Joshua Pritikin) 2016-01-22 16:05:29 +01:00
Gael Guennebaud
06971223ef Unify std::numeric_limits and device::numeric_limits within numext namespace 2016-01-22 15:02:21 +01:00
Gael Guennebaud
ee37eb4eed bug #977: avoid division by 0 in normalize() and normalized(). 2016-01-21 20:43:42 +01:00
Gael Guennebaud
7cae8918c0 Fix compilation on old gcc+AVX 2016-01-21 20:30:32 +01:00
Gael Guennebaud
8dca9f97e3 Add numext::sqrt function to enable custom optimized implementation. 
This changeset add two specializations for float/double on SSE. Those
are mostly usefull with GCC for which std::sqrt add an extra and costly
check on the result of _mm_sqrt_*. Clang does not add this burden.

In this changeset, only DenseBase::norm() makes use of it.
 2016-01-21 20:18:51 +01:00
Gael Guennebaud
34340458cb bug #1151: remove useless critical section 2016-01-21 14:29:45 +01:00
Gael Guennebaud
ed8ade9c65 bug #1149: fix Pastix*::*parm() 2016-01-20 19:01:24 +01:00
Gael Guennebaud
4c5e96aab6 bug #1148: silent Pastix by default 2016-01-20 18:56:17 +01:00
Gael Guennebaud
db237d0c75 bug #1145: fix PastixSupport LLT/LDLT wrappers (missing resize prior to calls to selfAdjointView) 2016-01-20 18:49:01 +01:00
Gael Guennebaud
0b7169d1f7 bug #1147: fix compilation of PastixSupport 2016-01-20 18:15:59 +01:00
Gael Guennebaud
234a1094b7 Add static assertion to y(), z(), w() accessors 2016-01-20 09:18:44 +01:00
Eugene Brevdo
6a75e7e0d5 Digamma cleanup 
* Added permission from cephes author to use his code
* Cleanup in ArrayCwiseUnaryOps
 2016-01-15 16:32:21 -08:00
Benoit Steiner
bbdabbb379 Made the blas utils usable from within a cuda kernel 2016-01-11 17:26:56 -08:00
Gael Guennebaud
8b9dc9f0df bug #1144: fix regression in x=y+A*x (aliasing), and move evaluator_traits::AssumeAliasing to evaluator_assume_aliasing. 2016-01-09 08:30:38 +01:00
Gael Guennebaud
ee738321aa rm remaining debug code 2016-01-06 14:49:40 +01:00
Christoph Hertzberg
54bf582303 bug #1143: Work-around gcc bug 2016-01-06 11:59:24 +01:00
Gael Guennebaud
715f6f049f Improve inline documentation of SparseCompressedBase and its derived classes 2016-01-03 21:56:30 +01:00
Gael Guennebaud
8b0d1eb0f7 Fix numerous doxygen shortcomings, and workaround some clang -Wdocumentation warnings 2016-01-01 21:45:06 +01:00
Gael Guennebaud
9900782e88 Mark AlignedBit and EvalBeforeNestingBit with deprecated attribute, and remove the remaining usages of EvalBeforeNestingBit. 2015-12-30 16:47:49 +01:00
Gael Guennebaud
70404e07c2 Workaround clang -Wdocumentation warning about "/*<" 2015-12-30 16:46:45 +01:00
Gael Guennebaud
addb7066e8 Workaround "empty paragraph" warning with clang -Wdocumentation 2015-12-30 16:45:44 +01:00
Gael Guennebaud
eadc377b3f Add missing doc of Derived template parameter 2015-12-30 16:43:19 +01:00
Gael Guennebaud
29bb599e03 Fix numerous doxygen issues in auto-link generation 2015-12-30 16:04:24 +01:00
Gael Guennebaud
25f2b8d824 bug #1141: add missing initialization of CholmodBase::m_*IsOk 2015-12-29 15:50:11 +01:00
Eugene Brevdo
f2471f31e0 Modify constants in SpecialFunctions to lowercase (avoid name conflicts). 2015-12-28 17:48:38 -08:00
Eugene Brevdo
afb35385bf Change PI* to M_PI* in SpecialFunctions to avoid possible breakage 
with external DEFINEs.
 2015-12-28 17:34:06 -08:00
Eugene Brevdo
cef81c9084 Merged eigen/eigen into default 2015-12-24 21:17:33 -08:00
Eugene Brevdo
f7362772e3 Add digamma for CPU + CUDA. Includes tests. 2015-12-24 21:15:38 -08:00
Gael Guennebaud
d2e288ae50 Workaround compilers that do not even define _mm256_set_m128. 2015-12-24 16:53:43 +01:00
Benoit Steiner
3504ae47ca Made it possible to run the lgamma, erf, and erfc functors on a CUDA gpu. 2015-12-21 15:20:06 -08:00
Benoit Steiner
a6c243617b Fixed a typo in previous change. 2015-12-21 09:05:45 -08:00
Benoit Steiner
51be91f15e Added support for CUDA architectures that don's support for 3.5 capabilities 2015-12-21 08:42:58 -08:00
Benoit Steiner
6d777e1bc7 Fixed a typo. 2015-12-18 19:25:50 -08:00
Gael Guennebaud
3abd8470ca bug #1140: remove custom definition and use of _mm256_setr_m128 2015-12-18 14:18:59 +01:00
Gael Guennebaud
9f9de1aaa9 bump to 3.3-beta1 2015-12-16 21:48:48 +01:00
Gael Guennebaud
ae8b217a01 Update doc to make it clear that only SuperLU 4.x is supported 2015-12-16 10:47:03 +01:00
Gael Guennebaud
140f3a02a8 Fix MKL wrapper for ComplexSchur 2015-12-11 23:31:21 +01:00
Gael Guennebaud
4483c0fdf6 Fix unused variable warning. 2015-12-11 23:29:53 +01:00
Gael Guennebaud
774dba87c8 merge 2015-12-11 23:28:44 +01:00
Gael Guennebaud
c884a8e7f4 merge 2015-12-11 23:07:33 +01:00
Gael Guennebaud
b60a8967f5 bug #1134: fix JacobiSVD pre-allocation 
(grafted from f22036f5f8
)
 2015-12-11 11:59:11 +01:00
Gael Guennebaud
ca39b1546e Merged in ebrevdo/eigen (pull request PR-148) 
Add special functions to eigen: lgamma, erf, erfc.
 2015-12-11 11:52:09 +01:00
Gael Guennebaud
82152f2ae6 bug #1132: add EIGEN_MAPBASE_PLUGIN 2015-12-11 11:43:49 +01:00
Gael Guennebaud
4519fd5d40 Fix MKL compilation issue 2015-12-11 11:11:38 +01:00
Gael Guennebaud
7385e6e2ef Remove useless explicit 2015-12-11 11:11:19 +01:00
Gael Guennebaud
bcb4f126a7 Fix compilation of PardisoSupport 2015-12-11 11:11:00 +01:00
Gael Guennebaud
30b5c4cd14 Remove useless "explicit", and fix inline/static order. 2015-12-11 10:59:39 +01:00
Gael Guennebaud
79c1e6d0a6 Fix compilation of MKL support. 2015-12-11 10:55:07 +01:00
Gael Guennebaud
c684a07eba merge 2015-12-11 10:06:38 +01:00
Benoit Steiner
b820b097b8 Created EIGEN_HAS_C99_MATH define as Gael suggested. 2015-12-10 13:52:05 -08:00
Gael Guennebaud
df6f54ff63 Fix storage order of PartialRedux 2015-12-10 22:24:58 +01:00
Mark Borgerding
22dd368ea0 sign(complex) compiles for GPU 2015-12-10 16:14:29 -05:00
Benoit Steiner
58e06447de Silence a compilation warning 2015-12-10 13:11:36 -08:00
Benoit Steiner
48877a6933 Only implement the lgamma, erf, and erfc functions when using a compiler compliant with the C99 specification. 2015-12-10 13:09:49 -08:00
Gael Guennebaud
7ad1aaec1d bug #1103: fix neon vectorization of pmul(Packet1cd,Packet1cd) 2015-12-10 16:06:33 +01:00
Benoit Steiner
53b196aa5f Simplified the implementation of lgamma, erf, and erfc 2015-12-08 14:17:34 -08:00
Benoit Steiner
e535450573 Cleanup 2015-12-08 14:06:39 -08:00
Benoit Steiner
b1ae39794c Simplified the code a bit 2015-12-07 16:46:35 -08:00
Benoit Steiner
73b68d4370 Fixed a couple of typos 
Cleaned up the code a bit.
 2015-12-07 16:38:48 -08:00
Eugene Brevdo
fa4f933c0f Add special functions to Eigen: lgamma, erf, erfc. 
Includes CUDA support and unit tests.
 2015-12-07 15:24:49 -08:00
Gael Guennebaud
ad3d68400e Add matrix-free solver example 2015-12-07 12:33:38 +01:00
Gael Guennebaud
b37036afce Implement wrapper for matrix-free iterative solvers 2015-12-07 12:23:22 +01:00
Benoit Steiner
e25e3a041b Added rsqrt() method to the Array class: this method computes the coefficient-wise inverse square root much more efficiently than calling sqrt().inverse(). 2015-12-03 18:16:35 -08:00
Benoit Steiner
c41e9e4bd0 Merged in Unril/eigen-1/Unril/fixes-internal-compiler-error-while-comp-1449156092576 (pull request PR-147) 
Fixes internal compiler error while compiling with VC2015 Update1 x64.
 2015-12-03 14:26:14 -08:00
Gael Guennebaud
1562e13aba Add missing Rotation2D::operator=(Matrix2x2) 2015-12-03 22:25:26 +01:00
Nikolay Fedorov
944647c0aa Fixes internal compiler error while compiling with VC2015 Update1 x64. 2015-12-03 15:21:43 +00:00
Benoit Steiner
d2d4c45d55 Made it possible to leverage several binary functor in a CUDA kernel 
Explicitely specified the return type of the various scalar_cmp_op functors.
 2015-12-02 17:21:33 -08:00
Gael Guennebaud
c5b86893e7 bug #1123: add missing documentation of angle() and axis() 2015-12-01 14:45:08 +01:00
Gael Guennebaud
0bb12fa614 Add LU::transpose().solve() and LU::adjoint().solve() API. 2015-12-01 14:38:47 +01:00
Rasmus Munk Larsen
1663d15da7 Add internal method _solve_impl_transposed() to LU decomposition classes that solves A^T x = b or A^* x = b. 2015-11-30 13:39:24 -08:00
Gael Guennebaud
6c02cbbb0f Fix matrix to quaternion (and angleaxis) conversion for matrix expression. 2015-12-01 09:45:56 +01:00
Gael Guennebaud
1d906d883d Fix degenerate cases in syrk and trsm 2015-11-30 22:20:31 +01:00
Gael Guennebaud
afa11d646d Fix UmfPackLU ctor for exppressions 2015-11-27 22:04:22 +01:00
Gael Guennebaud
6bdeb8cfbe bug #918, umfpack: add access to umfpack return code and parameters 2015-11-27 21:58:36 +01:00
Gael Guennebaud
3f32f5ec22 ArrayBase::sign: add unit test and fix doc 2015-11-27 16:27:53 +01:00
Gael Guennebaud
1261d020c3 bug #1120, superlu: mem_usage_t is now uniquely defined, so let's use it. 2015-11-27 10:39:09 +01:00
Gael Guennebaud
ca001d7c2a Big 1009, part 2/2: add static assertion on LinearAccessBit in coeff(index)-like methods. 2015-11-27 10:06:47 +01:00
Gael Guennebaud
91a7059459 bug #1009, part 1/2: make sure vector expressions expose LinearAccessBit flag. 2015-11-27 10:06:07 +01:00
Mark Borgerding
7ddcf97da7 added scalar_sign_op (both real,complex) 2015-11-24 17:15:07 -05:00
Gael Guennebaud
f9fff67a56 Disable "decorated name length exceeded, name was truncated" MSVC warning. 2015-11-23 15:03:24 +01:00
Gael Guennebaud
f3dca16a1d bug #1117: workaround unused-local-typedefs warning when EIGEN_NO_STATIC_ASSERT and NDEBUG are both defined. 2015-11-23 14:07:52 +01:00
Gael Guennebaud
82bd4e546a Merged in dr15jones/eigen (pull request PR-146) 
Use a class constructor to initialize CPU cache sizes
 2015-11-22 22:50:31 +01:00
Gael Guennebaud
35c17a3fc8 Use overload instead of template full specialization to please old MSVC 2015-11-22 22:09:57 +01:00
Gael Guennebaud
b265979a70 Make FullPivLU::solve use rank() instead of nonzeroPivots(). 2015-11-21 15:03:04 +01:00
Chris Jones
4946d758c9 Use a class constructor to initialize CPU cache sizes 
Using a static instance of a class to initialize the values for
the CPU cache sizes guarantees thread-safe initialization of the
values when using C++11. Therefore under C++11 it is no longer
necessary to call Eigen::initParallel() before calling any eigen
functions on different threads.
 2015-11-20 19:58:08 +01:00
Gael Guennebaud
027a846b34 Use .data() instead of &coeffRef(0). 2015-11-20 15:30:10 +01:00
Gael Guennebaud
4fc36079e7 Fix overload instantiation for clang 2015-11-20 15:29:03 +01:00
Gael Guennebaud
5c9c0dca4d Add missing using statement to enable fast Array<complex> / real operations. (was ok for Matrix only) 2015-11-20 14:51:36 +01:00
Gael Guennebaud
e1b27bcb0b Workaround MSVC missing overloads of std::fpclassify for integral types 2015-11-20 13:55:34 +01:00
Gael Guennebaud
e52d4f8d8d Add is_integral<> type traits 2015-11-20 13:54:28 +01:00
Benoit Steiner
7d1cedd0fe Added numeric limits for unsigned integers 2015-11-18 17:17:44 -08:00
Benoit Jacob
4926251f13 bug #1115: enable static alignment on ARM outside of old-GCC 2015-11-18 10:55:23 -05:00
Benoit Steiner
bf792f59e3 Only enable the use of constexpr with nvcc if we're using version 7.5 or above 2015-11-13 12:24:22 -08:00
Benoit Steiner
1e1755352d Made it possible to compute atan, tanh, sinh and cosh on GPU 2015-11-12 20:19:38 -08:00
Benoit Steiner
e4d45f3440 Only enable the use of const expression when nvcc is called with the -std=c++11 option 2015-11-12 18:18:35 -08:00
Benoit Steiner
8037826367 Simplified more of the IndexList code. 2015-11-12 17:19:45 -08:00
Gael Guennebaud
dfbb889fe9 Fix missing Dynamic versus HugeCost changes 2015-11-12 12:09:48 +01:00
Gael Guennebaud
e701cb2c7c Update EIGEN_FAST_MATH doc 2015-11-12 12:09:19 +01:00
Benoit Steiner
4f471146fb Allow the vectorized version of the Binary and the Nullary functors to run on GPU 2015-11-11 15:19:00 -08:00
Gael Guennebaud
e73ef4f25e bug #1109: use noexcept instead of throw for C++11 compilers 2015-12-10 14:21:23 +01:00
Gael Guennebaud
145ad5d800 Use more explicit names. 2015-12-10 12:03:38 +01:00
Gael Guennebaud
75f0fe3795 Fix usage of "Index" as a compile time integral. 2015-12-10 12:01:06 +01:00
Gael Guennebaud
f248249c1f bug #1113: fix name conflict with C99's "I". 2015-12-10 11:57:57 +01:00
Gael Guennebaud
fbe18d5507 Forbid the creation of SparseCompressedBase object 2015-12-09 15:47:32 +01:00
Gael Guennebaud
dc73430d4b bug #1074: forbid the creation of PlainObjectBase object by making its ctor protected 2015-12-09 15:47:08 +01:00
Gael Guennebaud
1257fbd2f9 Fix sign-unsigned issue in enum 2015-12-09 10:06:42 +01:00
Gael Guennebaud
4549549992 Fix and clarify documentation of Transform wrt operator*(MatrixBase) 2015-12-08 16:21:49 +01:00
Gael Guennebaud
543bd28a24 Fix Alignment in coeff-based product, and enable unaligned vectorization 2015-12-08 11:28:05 +01:00
Benoit Steiner
d27e4f1cba Added missing EIGEN_DEVICE_FUNC statements 2015-11-06 09:23:58 -08:00
Benoit Steiner
ed1962b464 Reimplement the tensor comparison operators by using the scalar_cmp_op functors. This makes them more cuda friendly. 2015-11-06 09:18:43 -08:00
Gael Guennebaud
bfd6ee64f3 bug #1105: fix default preallocation when moving from compressed to uncompressed mode 2015-11-06 15:05:37 +01:00
Gael Guennebaud
ae87f094eb Fix "," in non SSE4 mode 2015-11-05 12:08:36 +01:00
Gael Guennebaud
90323f1751 Fix AVX round/ceil/floor, and fix respective unit test 2015-11-04 22:15:57 +01:00
Gael Guennebaud
3dd24bdf99 Merged in aavenel/eigen (pull request PR-142) 
Add round, ceil and floor for SSE4.1/AVX (Bug #70)
 2015-11-04 18:26:38 +01:00
Gael Guennebaud
902750826b Add support for dense.cwiseProduct(sparse) 
This also fixes a regression regarding (dense*sparse).diagonal()
 2015-11-04 17:42:07 +01:00
Gael Guennebaud
f6b1deebab Fix compilation of sparse-triangular to dense assignment 2015-11-04 17:02:32 +01:00
Benoit Steiner
36cd6daaae Made the CUDA implementation of ploadt_ro compatible with cuda implementations older than 3.5 2015-11-03 16:36:30 -08:00
Gael Guennebaud
29a94c8055 compilation issue 2015-11-02 16:11:59 +01:00
Alexandre Avenel
38832e0791 Merge 2015-11-01 10:55:42 +01:00