Lines Matching full:matrix
148 // Find where in the covariance matrix the block is located.
218 Matrix block1_jacobian(block1_size, block1_local_size);
225 Matrix block2_jacobian(block2_size, block2_local_size);
252 // Determine the sparsity pattern of the covariance matrix based on
281 // first row corresponding to it in the covariance matrix using the
293 // Compute the number of non-zeros in the covariance matrix. Along
295 // triangular part of the matrix.
312 // Make sure we are constructing a block upper triangular matrix.
332 // Fill the sparsity pattern of the covariance matrix.
340 // covariance matrix. For each parameter block, look in the upper
341 // triangular part of the covariance matrix to see if there are any
346 // row/columns of the covariance matrix are ordered by the pointer
349 // rows of the covariance matrix in order.
351 int cursor = 0; // index into the covariance matrix.
416 // Nothing to do, all zeros covariance matrix.
436 Matrix is not symmetric.
479 // Since the covariance matrix is symmetric, the i^th row and column
596 // Nothing to do, all zeros covariance matrix.
645 cholmod_jacobian.stype = 0; // Matrix is not symmetric.
661 // sparse matrix.
666 // matrix. When using AMD, we have observed in the past that
667 // computing the ordering with the block matrix is significantly
684 LOG(WARNING) << "Jacobian matrix is rank deficient."
709 // Since the covariance matrix is symmetric, the i^th row and column
759 // Nothing to do, all zeros covariance matrix.
767 Matrix dense_jacobian(jacobian.num_rows, jacobian.num_cols);
777 Eigen::JacobiSVD<Matrix> svd(dense_jacobian,
798 // the resulting covariance matrix is a Moore-Penrose inverse
823 Matrix dense_covariance =
