- added a MapBase base xpr on top of which Map and the specialization
of Block are implemented
- MapBase forces both aligned loads (and aligned stores, see below) in expressions
such as "x.block(...) += other_expr"
* Significant vectorization improvement:
- added a AlignedBit flag meaning the first coeff/packet is aligned,
this allows to not generate extra code to deal with the first unaligned part
- removed all unaligned stores when no unrolling
- removed unaligned loads in Sum when the input as the DirectAccessBit flag
* Some code simplification in CacheFriendly product
* Some minor documentation improvements
- removes much code
- 2.5x faster (even though LU uses complete pivoting contrary to what
inverse used to do!)
- there _were_ numeric stability problems with the partial pivoting
approach of inverse(), with 200x200 matrices inversion failed almost
half of the times (overflow). Now these problems are solved thanks to
complete pivoting.
* remove some useless stuff in LU
*in test/CMakeLists : modify EI_ADD_TEST so that 2nd argument is
additional compiler flags. used to add -O2 to test_product_large so it
doesn't take forever.
not allow to easily get the rank), fix a bug (which could have been
triggered by matrices having coefficients of very different
magnitudes).
Part: add an assert to prevent hard to find bugs
Swap: update comments
Note: in fact, inverse() always uses partial pivoting because the algo
currently used doesn't make sense with complete pivoting. No num
stability issue so far even with size 200x200. If there is any problem
we can of course reimplement inverse on top of LU.
pivoting for better numerical stability. For now the only application is
determinant.
* New determinant unit-test.
* Disable most of Swap.h for now as it makes LU fail (mysterious).
Anyway Swap needs a big overhaul as proposed on IRC.
* Remnants of old class Inverse removed.
* Some warnings fixed.
* faster matrix-matrix and matrix-vector products (especially for not aligned cases)
* faster tridiagonalization (make it using our matrix-vector impl.)
Others:
* fix Flags of Map
* split the test_product to two smaller ones
- added explicit enum to int conversion where needed
- if a function is not defined as declared and the return type is "tricky"
then the type must be typedefined somewhere. A "tricky return type" can be:
* a template class with a default parameter which depends on another template parameter
* a nested template class, or type of a nested template class
=> up to 6 times faster !
* Added DirectAccessBit to Part
* Added an exemple of a cwise operator
* Renamed perpendicular() => someOrthogonal() (geometry module)
* Fix a weired bug in ei_constant_functor: the default copy constructor did not copy
the imaginary part when the single member of the class is a complex...
- conflicts with operator * overloads
- discard the use of ei_pdiv for interger
(g++ handles operators on __m128* types, this is why it worked)
- weird behavior of icc in fixed size Block() constructor complaining
the initializer of m_blockRows and m_blockCols were missing while
we are in fixed size (maybe this hide deeper problem since this is a
recent one, but icc gives only little feedback)
Renamed "MatrixBase::extract() const" to "MatrixBase::part() const"
* Renamed static functions identity, zero, ones, random with an upper case
first letter: Identity, Zero, Ones and Random.
Removed EulerAngles, addes typdefs for Quaternion and AngleAxis,
and added automatic conversions from Quaternion/AngleAxis to Matrix3 such that:
Matrix3f m = AngleAxisf(0.2,Vector3f::UnitX) * AngleAxisf(0.2,Vector3f::UnitY);
just works.
* Improve the efficiency of matrix*vector in unaligned cases
* Trivial fixes in the destructors of MatrixStorage
* Removed the matrixNorm in test/product.cpp (twice faster and
that assumed the matrix product was ok while checking that !!)
- remove all invertibility checking, will be redundant with LU
- general case: adapt to matrix storage order for better perf
- size 4 case: handle corner cases without falling back to gen case.
- rationalize with selectors instead of compile time if
- add C-style computeInverse()
* update inverse test.
* in snippets, default cout precision to 3 decimal places
* add some cmake module from kdelibs to support btl with cmake 2.4
It basically performs 4 dot products at once reducing loads of the vector and improving
instructions scheduling. With 3 cache friendly algorithms, we now handle all product
configurations with outstanding perf for large matrices.
and vector * row-major products. Currently, it is enabled only is the matrix
has DirectAccessBit flag and the product is "large enough".
Added the respective unit tests in test/product/cpp.
might be twice faster fot small fixed size matrix
* added a sparse triangular solver (sparse version
of inverseProduct)
* various other improvements in the Sparse module
* added complete implementation of sparse matrix product
(with a little glue in Eigen/Core)
* added an exhaustive bench of sparse products including GMM++ and MTL4
=> Eigen outperforms in all transposed/density configurations !
* rework PacketMath and DummyPacketMath, make these actual template
specializations instead of just overriding by non-template inline
functions
* introduce ei_ploadt and ei_pstoret, make use of them in Map and Matrix
* remove Matrix::map() methods, use Map constructors instead.
* added some glue to Eigen/Core (SparseBit, ei_eval, Matrix)
* add two new sparse matrix types:
HashMatrix: based on std::map (for random writes)
LinkedVectorMatrix: array of linked vectors
(for outer coherent writes, e.g. to transpose a matrix)
* add a SparseSetter class to easily set/update any kind of matrices, e.g.:
{ SparseSetter<MatrixType,RandomAccessPattern> wrapper(mymatrix);
for (...) wrapper->coeffRef(rand(),rand()) = rand(); }
* automatic shallow copy for RValue
* and a lot of mess !
plus:
* remove the remaining ArrayBit related stuff
* don't use alloca in product for very large memory allocation
to "public:method()" i.e. reimplementing the generic method()
from MatrixBase.
improves compilation speed by 7%, reduces almost by half the call depth
of trivial functions, making gcc errors and application backtraces
nicer...
* introduce packet(int), make use of it in linear vectorized paths
--> completely fixes the slowdown noticed in benchVecAdd.
* generalize coeff(int) to linear-access xprs
* clarify the access flag bits
* rework api dox in Coeffs.h and util/Constants.h
* improve certain expressions's flags, allowing more vectorization
* fix bug in Block: start(int) and end(int) returned dyn*dyn size
* fix bug in Block: just because the Eval type has packet access
doesn't imply the block xpr should have it too.
* make the conj functor vectorizable: it is just identity in real case,
and complex doesn't use the vectorized path anyway.
* fix bug in Block: a 3x1 block in a 4x4 matrix (all fixed-size)
should not be vectorizable, since in fixed-size we are assuming
the size to be a multiple of packet size. (Or would you prefer
Vector3d to be flagged "packetaccess" even though no packet access
is possible on vectors of that type?)
* rename:
isOrtho for vectors ---> isOrthogonal
isOrtho for matrices ---> isUnitary
* add normalize()
* reimplement normalized with quotient1 functor
- uses the common "Compressed Column Storage" scheme
- supports every unary and binary operators with xpr template
assuming binaryOp(0,0) == 0 and unaryOp(0) = 0 (otherwise a sparse
matrix doesnot make sense)
- this is the first commit, so of course, there are still several shorcommings !
(could come back to redux after it has been vectorized,
and could serve as a starting point for that)
also make the abs2 functor vectorizable (for real types).
packet access, it is not certain that it will bring a performance
improvement: benchmarking needed.
* improve logic choosing slice vectorization.
* fix typo in SSE packet math, causing crash in unaligned case.
* fix bug in Product, causing crash in unaligned case.
* add TEST_SSE3 CMake option.
* make Matrix2f (and similar) vectorized using linear path
* fix a couple of warnings and compilation issues with ICC and gcc 3.3/3.4
(cannot get Transform compiles with gcc 3.3/3.4, see the FIXME)
* use ProductReturnType<>::Type to get the correct Product xpr type
* Product is no longer instanciated for xpr types which are evaluated
* vectorization of "a.transpose() * b" for the normal product (small and fixed-size matrix)
* some cleanning
* removed ArrayBase
** Much better organization
** Fix a few bugs
** Add the ability to unroll only the inner loop
** Add an unrolled path to the Like1D vectorization. Not well tested.
** Add placeholder for sliced vectorization. Unimplemented.
* Rework of corrected_flags:
** improve rules determining vectorizability
** for vectors, the storage-order is indifferent, so we tweak it
to allow vectorization of row-vectors.
* fix compilation in benchmark, and a warning in Transpose.
to optimize matrix-diag and diag-matrix products without
making Product over complicated.
* compilation fixes in Tridiagonalization and HessenbergDecomposition
in the case of 2x2 matrices.
* added an Orientation2D small class with similar interface than Quaternion
(used by Transform to handle 2D and 3D orientations seamlessly)
* added a couple of features in Transform.
them in the ei_traits, so that they're guaranteed even if the user
specified his own non-default flags (like before).
Measured to not make compilation any slower.
flags. This ensures that unless explicitly messed up otherwise,
a Matrix type is equal to its own Eval type. This seriously reduces
the number of types instantiated. Measured +13% compile speed, -7%
binary size.
* Improve doc of Matrix template parameters.
This is the first step towards a non-selfadjoint eigen solver.
Notes:
- We might consider merging Tridiagonalization and Hessenberg toghether ?
- Or we could factorize some code into a Householder class (could also be shared with QR)
which is even better optimized by the compiler.
* Quaternion no longer inherits MatrixBase. Instead it stores the coefficients
using a Matrix<> and provides only relevant methods.
tends to show L2 norm works very well here.
(the legacy implementation is still available via a preprocessor token
to allow further experiments if needed...)
as an argument of a function. Other possibilities for the name could be "end" or "matrix" ??
* various update in Quaternion, in particular I added a lot of FIXME about the API options,
these have to be discussed and fixed.
