diff options
| author |  A. Unique TensorFlower <nobody@tensorflow.org> | 2016-02-11 15:14:11 -0800 |
|---|---|---|
| committer |  TensorFlower Gardener <gardener@tensorflow.org> | 2016-02-11 17:14:08 -0800 |
| commit | 80e682562a202d5ed57c871aa76f0faec4061590 (patch) | |
| tree | 1cfdc58299fe61b7a9e6648150d6e525acd7b5cc /WORKSPACE | |
| parent | 3c933761e40101952fdcc0896bb3de1c5654192b (diff) | |
Switch the slow path in matrix_solve_ls to using Eigen::CompleteOrthogonalDecomposition (COD), which I recently contributed to Eigen in https://bitbucket.org/eigen/eigen/pull-requests/163/implement-complete-orthogonal/diff
The advantage of COD over column pivoted QR is that it is able to compute the minimum-norm solution when the matrix is rank-deficient, which is usually the desired behavior and makes it consistent with the fast path.
Change: 114483303
Diffstat (limited to 'WORKSPACE')
| -rw-r--r-- | WORKSPACE | 4 |
1 files changed, 2 insertions, 2 deletions
@@ -21,8 +21,8 @@ new_http_archive( new_http_archive( name = "eigen_archive", - url = "https://bitbucket.org/eigen/eigen/get/726c779.tar.gz", - sha256 = "30e0c5d84cfefc6a0bf7ae1e682b22788b5b2e408e7db7d9ea2d2aa9f70a72a9", + url = "https://bitbucket.org/eigen/eigen/get/0b9ab889fac2.tar.gz", + sha256 = "b9cff4ca8eb4889b1f52316b9f7362eec177898323c14d60d9fdb5ad2649c301", build_file = "eigen.BUILD", ) |