* renaming, e.g. LU ---> FullPivLU
* split tests framework: more robust, e.g. dont generate empty tests if a number is skipped
* make all remaining tests use that splitting, as needed.
* Fix 4x4 inversion (see stable branch)
* Transform::inverse() and geo_transform test : adapt to new inverse() API, it was also trying to instantiate inverse() for 3x4 matrices.
* CMakeLists: more robust regexp to parse the version number
* misc fixes in unit tests
is aligned or not. This is done using the Aligned constant:
Map<MatrixType,Aligned>::Map(data);
* rename ForceAligned to EnforceAlignedAccess, and update its doc,
and emphasize this is mainly an internal stuff.
* support resize() to same size (nop). The case of FFT was another case where that make one's life far easier.
hope that's ok with you Gael. but indeed, i don't use it in the ReturnByValue stuff.
FFT:
* Support MatrixBase (well, in the case with direct memory access such as Map)
* adapt unit test
* continue the decomposition until a pivot is exactly zero;
don't try to compute the rank in the decomposition itself.
* Instead, methods such as rank() use a new internal parameter
called 'threshold' to determine which pivots are to be
considered nonzero.
* The threshold is by default determined by defaultThreshold()
but the user can override that by calling useThreshold(value).
* In solve/kernel/image, don't assume that the diagonal of U
is sorted in decreasing order, because that's only approximately
true. Additional work was needed to extract the right pivots.
construction of generic expressions working
for both dense and sparse matrix. A nicer solution
would be to use CwiseBinaryOp for any kind of matrix.
To this end we either need to change the overall design
so that the base class(es) depends on the kind of matrix,
or we could add a template parameter to each expression
type (e.g., int Kind = ei_traits<MatrixType>::Kind)
allowing to specialize each expression for each kind of matrix.
* Extend AutoDiffScalar to work with sparse vector expression
for the derivatives.
now we also align to 8byte boundary fixed-size objects that are multiple of 8 bytes.
That's only useful for now for double, not e.g. for Vector2f, but that didn't seem to hurt. Am I missing something? Do you prefer that we don't align Vector2f at all?
Also, improvements in test_unalignedassert.
---> question: do we change the prefix eigen2/ to eigen3/ now?
no, better wait until we've also changed the repository name
* more message improvements: "Install Eigen" was unclear as it left
out other things like the BLAS library
* remove EIGEN_BUILD_TESTS and siblings
* add summary at the end of cmake run, hopefully not too verbose
* fix build of quaternion demo
* kill remnants of old binary library option
This allows to limit the number of instantiations of the big template method evalTo.
Also allows to get rid of the dummy MatrixBase::resize().
See "TODO" comment.
-- simpplify by removing the 2nd template parameter
-- rename Functor to Derived, as now it's a usual CRTP
* Homogeneous:
-- in products, honor the Max sizes etc.
that eigen2 owes all its performance to nonstandard restrict keyword.
well, this can also improve portability in case some compiler doesn't have __restrict.
