next up previous contents index
Next: Performance of LAPACK Up: Computational Routines Previous: Deflating Subspaces and Condition   Contents   Index


Generalized (or Quotient) Singular Value Decomposition

The generalized (or quotient) singular value decomposition of an m-by-n matrix A and a p-by-n matrix B is described in section 2.3.5. The routines described in this section, are used to compute the decomposition. The computation proceeds in the following two stages:

1.
xGGSVP is used to reduce the matrices A and B to triangular form:

\begin{eqnarray*}
U^T_1 A Q_1 & = & \bordermatrix{ & n-k-l & k & l \cr
\hfill k...
...-l & k & l \cr
\hfill l & 0 & 0 & B_{13} \cr
p-l & 0 & 0 & 0 }
\end{eqnarray*}


where A12 and B13 are nonsingular upper triangular, and A23 is upper triangular. If m-k-l < 0, the bottom zero block of U1T A Q1 does not appear, and A23 is upper trapezoidal. U1, V1 and Q1 are orthogonal matrices (or unitary matrices if A and B are complex). l is the rank of B, and k+l is the rank of $ \left( \begin{array}{c}
A \\
B
\end{array} \right) $.

2.
The generalized singular value decomposition of two l-by-l upper triangular matrices A23 and B13 is computed using xTGSJA2.2:

\begin{displaymath}
A_{23} = U_2 C R Q^T_2 \quad \mbox{and} \quad
B_{13} = V_2 S R Q^T_2 \; \; .
\end{displaymath}

Here U2, V2 and Q2 are orthogonal (or unitary) matrices, C and S are both real nonnegative diagonal matrices satisfying C2 + S2 = I, S is nonsingular, and R is upper triangular and nonsingular.


Table 2.16: Computational routines for the generalized singular value decomposition
Operation Single precision Double precision
  real complex real complex
triangular reduction of A and B SGGSVP CGGSVP DGGSVP ZGGSVP
GSVD of a pair of triangular matrices STGSJA CTGSJA DTGSJA ZTGSJA

The reduction to triangular form, performed by xGGSVP, uses QR decomposition with column pivoting for numerical rank determination. See [8] for details.

The generalized singular value decomposition of two triangular matrices, performed by xTGSJA, is done using a Jacobi-like method as described in [83,10].


next up previous contents index
Next: Performance of LAPACK Up: Computational Routines Previous: Deflating Subspaces and Condition   Contents   Index
Susan Blackford
1999-10-01