Lines Matching refs:dense

406    dimensions in which a full, dense approximation to the inverse
451 residuals) with relatively dense Jacobians, ``DENSE_QR`` is the method
460 Ceres uses ``Eigen`` 's dense QR factorization routines.
468 cases, using a dense QR factorization is inefficient. Let :math:`H =
483 factorization -- sparse and dense.
485 ``DENSE_NORMAL_CHOLESKY``  as the name implies performs a dense
487 ``Eigen`` 's dense LDLT factorization routines.
580 :math:`S` as a dense matrix [TrefethenBau]_. This method has
594 up over those based on dense factorization. Ceres implements this
729 completely dense factor.
1184    This algorithm gives high quality results but for large dense
1202    Ceres supports using multiple dense linear algebra libraries for
1203    dense matrix factorizations. Currently ``EIGEN`` and ``LAPACK`` are
1291    Some non-linear least squares problems are symbolically dense but
1472       dense matrix. The vectors :math:`D`, :math:`x` and :math:`f` are
1473       printed as dense vectors. This should only be used for small
2120    Type of the dense linear algebra library used.
2215 user can use the dense algorithm. For more details see the work of
2269      problems this is done using dense linear algebra.
2316       in ``SuiteSparse``. It uses dense linear algebra and is multi