LAPACK  3.10.0
LAPACK: Linear Algebra PACKage

◆ dlaed9()

subroutine dlaed9 ( integer  K,
integer  KSTART,
integer  KSTOP,
integer  N,
double precision, dimension( * )  D,
double precision, dimension( ldq, * )  Q,
integer  LDQ,
double precision  RHO,
double precision, dimension( * )  DLAMDA,
double precision, dimension( * )  W,
double precision, dimension( lds, * )  S,
integer  LDS,
integer  INFO 
)

DLAED9 used by DSTEDC. Finds the roots of the secular equation and updates the eigenvectors. Used when the original matrix is dense.

Download DLAED9 + dependencies [TGZ] [ZIP] [TXT]

Purpose:
 DLAED9 finds the roots of the secular equation, as defined by the
 values in D, Z, and RHO, between KSTART and KSTOP.  It makes the
 appropriate calls to DLAED4 and then stores the new matrix of
 eigenvectors for use in calculating the next level of Z vectors.
Parameters
[in]K
          K is INTEGER
          The number of terms in the rational function to be solved by
          DLAED4.  K >= 0.
[in]KSTART
          KSTART is INTEGER
[in]KSTOP
          KSTOP is INTEGER
          The updated eigenvalues Lambda(I), KSTART <= I <= KSTOP
          are to be computed.  1 <= KSTART <= KSTOP <= K.
[in]N
          N is INTEGER
          The number of rows and columns in the Q matrix.
          N >= K (delation may result in N > K).
[out]D
          D is DOUBLE PRECISION array, dimension (N)
          D(I) contains the updated eigenvalues
          for KSTART <= I <= KSTOP.
[out]Q
          Q is DOUBLE PRECISION array, dimension (LDQ,N)
[in]LDQ
          LDQ is INTEGER
          The leading dimension of the array Q.  LDQ >= max( 1, N ).
[in]RHO
          RHO is DOUBLE PRECISION
          The value of the parameter in the rank one update equation.
          RHO >= 0 required.
[in]DLAMDA
          DLAMDA is DOUBLE PRECISION array, dimension (K)
          The first K elements of this array contain the old roots
          of the deflated updating problem.  These are the poles
          of the secular equation.
[in]W
          W is DOUBLE PRECISION array, dimension (K)
          The first K elements of this array contain the components
          of the deflation-adjusted updating vector.
[out]S
          S is DOUBLE PRECISION array, dimension (LDS, K)
          Will contain the eigenvectors of the repaired matrix which
          will be stored for subsequent Z vector calculation and
          multiplied by the previously accumulated eigenvectors
          to update the system.
[in]LDS
          LDS is INTEGER
          The leading dimension of S.  LDS >= max( 1, K ).
[out]INFO
          INFO is INTEGER
          = 0:  successful exit.
          < 0:  if INFO = -i, the i-th argument had an illegal value.
          > 0:  if INFO = 1, an eigenvalue did not converge
Author
Univ. of Tennessee
Univ. of California Berkeley
Univ. of Colorado Denver
NAG Ltd.
Contributors:
Jeff Rutter, Computer Science Division, University of California at Berkeley, USA

Definition at line 154 of file dlaed9.f.

156 *
157 * -- LAPACK computational routine --
158 * -- LAPACK is a software package provided by Univ. of Tennessee, --
159 * -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..--
160 *
161 * .. Scalar Arguments ..
162  INTEGER INFO, K, KSTART, KSTOP, LDQ, LDS, N
163  DOUBLE PRECISION RHO
164 * ..
165 * .. Array Arguments ..
166  DOUBLE PRECISION D( * ), DLAMDA( * ), Q( LDQ, * ), S( LDS, * ),
167  $ W( * )
168 * ..
169 *
170 * =====================================================================
171 *
172 * .. Local Scalars ..
173  INTEGER I, J
174  DOUBLE PRECISION TEMP
175 * ..
176 * .. External Functions ..
177  DOUBLE PRECISION DLAMC3, DNRM2
178  EXTERNAL dlamc3, dnrm2
179 * ..
180 * .. External Subroutines ..
181  EXTERNAL dcopy, dlaed4, xerbla
182 * ..
183 * .. Intrinsic Functions ..
184  INTRINSIC max, sign, sqrt
185 * ..
186 * .. Executable Statements ..
187 *
188 * Test the input parameters.
189 *
190  info = 0
191 *
192  IF( k.LT.0 ) THEN
193  info = -1
194  ELSE IF( kstart.LT.1 .OR. kstart.GT.max( 1, k ) ) THEN
195  info = -2
196  ELSE IF( max( 1, kstop ).LT.kstart .OR. kstop.GT.max( 1, k ) )
197  $ THEN
198  info = -3
199  ELSE IF( n.LT.k ) THEN
200  info = -4
201  ELSE IF( ldq.LT.max( 1, k ) ) THEN
202  info = -7
203  ELSE IF( lds.LT.max( 1, k ) ) THEN
204  info = -12
205  END IF
206  IF( info.NE.0 ) THEN
207  CALL xerbla( 'DLAED9', -info )
208  RETURN
209  END IF
210 *
211 * Quick return if possible
212 *
213  IF( k.EQ.0 )
214  $ RETURN
215 *
216 * Modify values DLAMDA(i) to make sure all DLAMDA(i)-DLAMDA(j) can
217 * be computed with high relative accuracy (barring over/underflow).
218 * This is a problem on machines without a guard digit in
219 * add/subtract (Cray XMP, Cray YMP, Cray C 90 and Cray 2).
220 * The following code replaces DLAMDA(I) by 2*DLAMDA(I)-DLAMDA(I),
221 * which on any of these machines zeros out the bottommost
222 * bit of DLAMDA(I) if it is 1; this makes the subsequent
223 * subtractions DLAMDA(I)-DLAMDA(J) unproblematic when cancellation
224 * occurs. On binary machines with a guard digit (almost all
225 * machines) it does not change DLAMDA(I) at all. On hexadecimal
226 * and decimal machines with a guard digit, it slightly
227 * changes the bottommost bits of DLAMDA(I). It does not account
228 * for hexadecimal or decimal machines without guard digits
229 * (we know of none). We use a subroutine call to compute
230 * 2*DLAMBDA(I) to prevent optimizing compilers from eliminating
231 * this code.
232 *
233  DO 10 i = 1, n
234  dlamda( i ) = dlamc3( dlamda( i ), dlamda( i ) ) - dlamda( i )
235  10 CONTINUE
236 *
237  DO 20 j = kstart, kstop
238  CALL dlaed4( k, j, dlamda, w, q( 1, j ), rho, d( j ), info )
239 *
240 * If the zero finder fails, the computation is terminated.
241 *
242  IF( info.NE.0 )
243  $ GO TO 120
244  20 CONTINUE
245 *
246  IF( k.EQ.1 .OR. k.EQ.2 ) THEN
247  DO 40 i = 1, k
248  DO 30 j = 1, k
249  s( j, i ) = q( j, i )
250  30 CONTINUE
251  40 CONTINUE
252  GO TO 120
253  END IF
254 *
255 * Compute updated W.
256 *
257  CALL dcopy( k, w, 1, s, 1 )
258 *
259 * Initialize W(I) = Q(I,I)
260 *
261  CALL dcopy( k, q, ldq+1, w, 1 )
262  DO 70 j = 1, k
263  DO 50 i = 1, j - 1
264  w( i ) = w( i )*( q( i, j ) / ( dlamda( i )-dlamda( j ) ) )
265  50 CONTINUE
266  DO 60 i = j + 1, k
267  w( i ) = w( i )*( q( i, j ) / ( dlamda( i )-dlamda( j ) ) )
268  60 CONTINUE
269  70 CONTINUE
270  DO 80 i = 1, k
271  w( i ) = sign( sqrt( -w( i ) ), s( i, 1 ) )
272  80 CONTINUE
273 *
274 * Compute eigenvectors of the modified rank-1 modification.
275 *
276  DO 110 j = 1, k
277  DO 90 i = 1, k
278  q( i, j ) = w( i ) / q( i, j )
279  90 CONTINUE
280  temp = dnrm2( k, q( 1, j ), 1 )
281  DO 100 i = 1, k
282  s( i, j ) = q( i, j ) / temp
283  100 CONTINUE
284  110 CONTINUE
285 *
286  120 CONTINUE
287  RETURN
288 *
289 * End of DLAED9
290 *
double precision function dlamc3(A, B)
DLAMC3
Definition: dlamch.f:169
subroutine xerbla(SRNAME, INFO)
XERBLA
Definition: xerbla.f:60
subroutine dlaed4(N, I, D, Z, DELTA, RHO, DLAM, INFO)
DLAED4 used by DSTEDC. Finds a single root of the secular equation.
Definition: dlaed4.f:145
subroutine dcopy(N, DX, INCX, DY, INCY)
DCOPY
Definition: dcopy.f:82
real(wp) function dnrm2(n, x, incx)
DNRM2
Definition: dnrm2.f90:89
Here is the call graph for this function:
Here is the caller graph for this function: