1       SUBROUTINE DGEBD2( M, N, A, LDA, D, E, TAUQ, TAUP, WORK, INFO )
  2 *
  3 *  -- LAPACK routine (version 3.3.1) --
  4 *  -- LAPACK is a software package provided by Univ. of Tennessee,    --
  5 *  -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
  6 *  -- April 2011                                                      --
  7 *
  8 *     .. Scalar Arguments ..
  9       INTEGER            INFO, LDA, M, N
 10 *     ..
 11 *     .. Array Arguments ..
 12       DOUBLE PRECISION   A( LDA, * ), D( * ), E( * ), TAUP( * ),
 13      $                   TAUQ( * ), WORK( * )
 14 *     ..
 15 *
 16 *  Purpose
 17 *  =======
 18 *
 19 *  DGEBD2 reduces a real general m by n matrix A to upper or lower
 20 *  bidiagonal form B by an orthogonal transformation: Q**T * A * P = B.
 21 *
 22 *  If m >= n, B is upper bidiagonal; if m < n, B is lower bidiagonal.
 23 *
 24 *  Arguments
 25 *  =========
 26 *
 27 *  M       (input) INTEGER
 28 *          The number of rows in the matrix A.  M >= 0.
 29 *
 30 *  N       (input) INTEGER
 31 *          The number of columns in the matrix A.  N >= 0.
 32 *
 33 *  A       (input/output) DOUBLE PRECISION array, dimension (LDA,N)
 34 *          On entry, the m by n general matrix to be reduced.
 35 *          On exit,
 36 *          if m >= n, the diagonal and the first superdiagonal are
 37 *            overwritten with the upper bidiagonal matrix B; the
 38 *            elements below the diagonal, with the array TAUQ, represent
 39 *            the orthogonal matrix Q as a product of elementary
 40 *            reflectors, and the elements above the first superdiagonal,
 41 *            with the array TAUP, represent the orthogonal matrix P as
 42 *            a product of elementary reflectors;
 43 *          if m < n, the diagonal and the first subdiagonal are
 44 *            overwritten with the lower bidiagonal matrix B; the
 45 *            elements below the first subdiagonal, with the array TAUQ,
 46 *            represent the orthogonal matrix Q as a product of
 47 *            elementary reflectors, and the elements above the diagonal,
 48 *            with the array TAUP, represent the orthogonal matrix P as
 49 *            a product of elementary reflectors.
 50 *          See Further Details.
 51 *
 52 *  LDA     (input) INTEGER
 53 *          The leading dimension of the array A.  LDA >= max(1,M).
 54 *
 55 *  D       (output) DOUBLE PRECISION array, dimension (min(M,N))
 56 *          The diagonal elements of the bidiagonal matrix B:
 57 *          D(i) = A(i,i).
 58 *
 59 *  E       (output) DOUBLE PRECISION array, dimension (min(M,N)-1)
 60 *          The off-diagonal elements of the bidiagonal matrix B:
 61 *          if m >= n, E(i) = A(i,i+1) for i = 1,2,...,n-1;
 62 *          if m < n, E(i) = A(i+1,i) for i = 1,2,...,m-1.
 63 *
 64 *  TAUQ    (output) DOUBLE PRECISION array dimension (min(M,N))
 65 *          The scalar factors of the elementary reflectors which
 66 *          represent the orthogonal matrix Q. See Further Details.
 67 *
 68 *  TAUP    (output) DOUBLE PRECISION array, dimension (min(M,N))
 69 *          The scalar factors of the elementary reflectors which
 70 *          represent the orthogonal matrix P. See Further Details.
 71 *
 72 *  WORK    (workspace) DOUBLE PRECISION array, dimension (max(M,N))
 73 *
 74 *  INFO    (output) INTEGER
 75 *          = 0: successful exit.
 76 *          < 0: if INFO = -i, the i-th argument had an illegal value.
 77 *
 78 *  Further Details
 79 *  ===============
 80 *
 81 *  The matrices Q and P are represented as products of elementary
 82 *  reflectors:
 83 *
 84 *  If m >= n,
 85 *
 86 *     Q = H(1) H(2) . . . H(n)  and  P = G(1) G(2) . . . G(n-1)
 87 *
 88 *  Each H(i) and G(i) has the form:
 89 *
 90 *     H(i) = I - tauq * v * v**T  and G(i) = I - taup * u * u**T
 91 *
 92 *  where tauq and taup are real scalars, and v and u are real vectors;
 93 *  v(1:i-1) = 0, v(i) = 1, and v(i+1:m) is stored on exit in A(i+1:m,i);
 94 *  u(1:i) = 0, u(i+1) = 1, and u(i+2:n) is stored on exit in A(i,i+2:n);
 95 *  tauq is stored in TAUQ(i) and taup in TAUP(i).
 96 *
 97 *  If m < n,
 98 *
 99 *     Q = H(1) H(2) . . . H(m-1)  and  P = G(1) G(2) . . . G(m)
100 *
101 *  Each H(i) and G(i) has the form:
102 *
103 *     H(i) = I - tauq * v * v**T  and G(i) = I - taup * u * u**T
104 *
105 *  where tauq and taup are real scalars, and v and u are real vectors;
106 *  v(1:i) = 0, v(i+1) = 1, and v(i+2:m) is stored on exit in A(i+2:m,i);
107 *  u(1:i-1) = 0, u(i) = 1, and u(i+1:n) is stored on exit in A(i,i+1:n);
108 *  tauq is stored in TAUQ(i) and taup in TAUP(i).
109 *
110 *  The contents of A on exit are illustrated by the following examples:
111 *
112 *  m = 6 and n = 5 (m > n):          m = 5 and n = 6 (m < n):
113 *
114 *    (  d   e   u1  u1  u1 )           (  d   u1  u1  u1  u1  u1 )
115 *    (  v1  d   e   u2  u2 )           (  e   d   u2  u2  u2  u2 )
116 *    (  v1  v2  d   e   u3 )           (  v1  e   d   u3  u3  u3 )
117 *    (  v1  v2  v3  d   e  )           (  v1  v2  e   d   u4  u4 )
118 *    (  v1  v2  v3  v4  d  )           (  v1  v2  v3  e   d   u5 )
119 *    (  v1  v2  v3  v4  v5 )
120 *
121 *  where d and e denote diagonal and off-diagonal elements of B, vi
122 *  denotes an element of the vector defining H(i), and ui an element of
123 *  the vector defining G(i).
124 *
125 *  =====================================================================
126 *
127 *     .. Parameters ..
128       DOUBLE PRECISION   ZERO, ONE
129       PARAMETER          ( ZERO = 0.0D+0, ONE = 1.0D+0 )
130 *     ..
131 *     .. Local Scalars ..
132       INTEGER            I
133 *     ..
134 *     .. External Subroutines ..
135       EXTERNAL           DLARF, DLARFG, XERBLA
136 *     ..
137 *     .. Intrinsic Functions ..
138       INTRINSIC          MAXMIN
139 *     ..
140 *     .. Executable Statements ..
141 *
142 *     Test the input parameters
143 *
144       INFO = 0
145       IF( M.LT.0 ) THEN
146          INFO = -1
147       ELSE IF( N.LT.0 ) THEN
148          INFO = -2
149       ELSE IF( LDA.LT.MAX1, M ) ) THEN
150          INFO = -4
151       END IF
152       IF( INFO.LT.0 ) THEN
153          CALL XERBLA( 'DGEBD2'-INFO )
154          RETURN
155       END IF
156 *
157       IF( M.GE.N ) THEN
158 *
159 *        Reduce to upper bidiagonal form
160 *
161          DO 10 I = 1, N
162 *
163 *           Generate elementary reflector H(i) to annihilate A(i+1:m,i)
164 *
165             CALL DLARFG( M-I+1, A( I, I ), A( MIN( I+1, M ), I ), 1,
166      $                   TAUQ( I ) )
167             D( I ) = A( I, I )
168             A( I, I ) = ONE
169 *
170 *           Apply H(i) to A(i:m,i+1:n) from the left
171 *
172             IF( I.LT.N )
173      $         CALL DLARF( 'Left', M-I+1, N-I, A( I, I ), 1, TAUQ( I ),
174      $                     A( I, I+1 ), LDA, WORK )
175             A( I, I ) = D( I )
176 *
177             IF( I.LT.N ) THEN
178 *
179 *              Generate elementary reflector G(i) to annihilate
180 *              A(i,i+2:n)
181 *
182                CALL DLARFG( N-I, A( I, I+1 ), A( I, MIN( I+2, N ) ),
183      $                      LDA, TAUP( I ) )
184                E( I ) = A( I, I+1 )
185                A( I, I+1 ) = ONE
186 *
187 *              Apply G(i) to A(i+1:m,i+1:n) from the right
188 *
189                CALL DLARF( 'Right', M-I, N-I, A( I, I+1 ), LDA,
190      $                     TAUP( I ), A( I+1, I+1 ), LDA, WORK )
191                A( I, I+1 ) = E( I )
192             ELSE
193                TAUP( I ) = ZERO
194             END IF
195    10    CONTINUE
196       ELSE
197 *
198 *        Reduce to lower bidiagonal form
199 *
200          DO 20 I = 1, M
201 *
202 *           Generate elementary reflector G(i) to annihilate A(i,i+1:n)
203 *
204             CALL DLARFG( N-I+1, A( I, I ), A( I, MIN( I+1, N ) ), LDA,
205      $                   TAUP( I ) )
206             D( I ) = A( I, I )
207             A( I, I ) = ONE
208 *
209 *           Apply G(i) to A(i+1:m,i:n) from the right
210 *
211             IF( I.LT.M )
212      $         CALL DLARF( 'Right', M-I, N-I+1, A( I, I ), LDA,
213      $                     TAUP( I ), A( I+1, I ), LDA, WORK )
214             A( I, I ) = D( I )
215 *
216             IF( I.LT.M ) THEN
217 *
218 *              Generate elementary reflector H(i) to annihilate
219 *              A(i+2:m,i)
220 *
221                CALL DLARFG( M-I, A( I+1, I ), A( MIN( I+2, M ), I ), 1,
222      $                      TAUQ( I ) )
223                E( I ) = A( I+1, I )
224                A( I+1, I ) = ONE
225 *
226 *              Apply H(i) to A(i+1:m,i+1:n) from the left
227 *
228                CALL DLARF( 'Left', M-I, N-I, A( I+1, I ), 1, TAUQ( I ),
229      $                     A( I+1, I+1 ), LDA, WORK )
230                A( I+1, I ) = E( I )
231             ELSE
232                TAUQ( I ) = ZERO
233             END IF
234    20    CONTINUE
235       END IF
236       RETURN
237 *
238 *     End of DGEBD2
239 *
240       END