saberbud/ceres-solver
1
0
Fork 0
mirror of https://github.com/ceres-solver/ceres-solver.git synced 2026-08-30 00:50:37 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add Apple's Accelerate framework as a sparse linear algebra library.

Browse Source 
- Currently DynamicSparseNormalCholeskySolver is unsupported for
  Accelerate.

Change-Id: I03b5a86bb22fef249c4aecd48947a613e8eff7a5
This commit is contained in:
Alex Stewart
2018-06-23 20:17:34 +01:00
parent 357288008e
commit 8f41ca6abc
44 changed files with 1835 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/source/installation.rst
+14 -2
View File
@@ -89,6 +89,10 @@ optional. For details on customizing the build process, see
no dependencies on ``LAPACK`` and ``BLAS``. This makes for a simpler
build process and a smaller binary. **Optional**
- `Apple's Accelerate sparse solvers <https://developer.apple.com/documentation/accelerate/sparse_solvers>`_.
As of Xcode 9.0, Apple's Accelerate framework includes support for
solving sparse linear systems across macOS, iOS et al. **Optional**
- `BLAS <http://www.netlib.org/blas/>`_ and `LAPACK
<http://www.netlib.org/lapack/>`_ routines are needed by
``SuiteSparse``, and optionally used by Ceres directly for some
@@ -691,7 +695,13 @@ Options controlling Ceres configuration
CXSparse is licensed under the LGPL.
#. ``EIGENSPARSE [Default: OFF]``: By default, Ceres will not use
#. ``ACCELERATESPARSE [Default: ON]``: By default, Ceres will link to
Apple's Accelerate framework directly if a version of it is detected
which supports solving sparse linear systems. Note that on Apple OSs
Accelerate usually also provides the BLAS/LAPACK implementations and
so would be linked against irrespective of the value of ``ACCELERATESPARSE``.
#. ``EIGENSPARSE [Default: ON]``: By default, Ceres will not use
Eigen's sparse Cholesky factorization.
.. NOTE::
@@ -896,12 +906,14 @@ The Ceres components which can be specified are:
#. ``CXSparse``: Ceres built with CXSparse (``CXSPARSE=ON``).
#. ``AccelerateSparse``: Ceres built with Apple's Accelerate sparse solvers (``ACCELERATESPARSE=ON``).
#. ``EigenSparse``: Ceres built with Eigen's sparse Cholesky factorization
(``EIGENSPARSE=ON``).
#. ``SparseLinearAlgebraLibrary``: Ceres built with *at least one* sparse linear
algebra library. This is equivalent to ``SuiteSparse`` **OR** ``CXSparse``
**OR** ``EigenSparse``.
**OR** ``AccelerateSparse`` **OR** ``EigenSparse``.
#. ``SchurSpecializations``: Ceres built with Schur specializations
(``SCHUR_SPECIALIZATIONS=ON``).