dlasd2.f.html
来自「famous linear algebra library (LAPACK) p」· HTML 代码 · 共 537 行 · 第 1/3 页
HTML
537 行
</span><span class="comment">*</span><span class="comment"> ===============
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Based on contributions by
</span><span class="comment">*</span><span class="comment"> Ming Gu and Huan Ren, Computer Science Division, University of
</span><span class="comment">*</span><span class="comment"> California at Berkeley, USA
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> =====================================================================
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> .. Parameters ..
</span> DOUBLE PRECISION ZERO, ONE, TWO, EIGHT
PARAMETER ( ZERO = 0.0D+0, ONE = 1.0D+0, TWO = 2.0D+0,
$ EIGHT = 8.0D+0 )
<span class="comment">*</span><span class="comment"> ..
</span><span class="comment">*</span><span class="comment"> .. Local Arrays ..
</span> INTEGER CTOT( 4 ), PSM( 4 )
<span class="comment">*</span><span class="comment"> ..
</span><span class="comment">*</span><span class="comment"> .. Local Scalars ..
</span> INTEGER CT, I, IDXI, IDXJ, IDXJP, J, JP, JPREV, K2, M,
$ N, NLP1, NLP2
DOUBLE PRECISION C, EPS, HLFTOL, S, TAU, TOL, Z1
<span class="comment">*</span><span class="comment"> ..
</span><span class="comment">*</span><span class="comment"> .. External Functions ..
</span> DOUBLE PRECISION <a name="DLAMCH.183"></a><a href="dlamch.f.html#DLAMCH.1">DLAMCH</a>, <a name="DLAPY2.183"></a><a href="dlapy2.f.html#DLAPY2.1">DLAPY2</a>
EXTERNAL <a name="DLAMCH.184"></a><a href="dlamch.f.html#DLAMCH.1">DLAMCH</a>, <a name="DLAPY2.184"></a><a href="dlapy2.f.html#DLAPY2.1">DLAPY2</a>
<span class="comment">*</span><span class="comment"> ..
</span><span class="comment">*</span><span class="comment"> .. External Subroutines ..
</span> EXTERNAL DCOPY, <a name="DLACPY.187"></a><a href="dlacpy.f.html#DLACPY.1">DLACPY</a>, <a name="DLAMRG.187"></a><a href="dlamrg.f.html#DLAMRG.1">DLAMRG</a>, <a name="DLASET.187"></a><a href="dlaset.f.html#DLASET.1">DLASET</a>, DROT, <a name="XERBLA.187"></a><a href="xerbla.f.html#XERBLA.1">XERBLA</a>
<span class="comment">*</span><span class="comment"> ..
</span><span class="comment">*</span><span class="comment"> .. Intrinsic Functions ..
</span> INTRINSIC ABS, MAX
<span class="comment">*</span><span class="comment"> ..
</span><span class="comment">*</span><span class="comment"> .. Executable Statements ..
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Test the input parameters.
</span><span class="comment">*</span><span class="comment">
</span> INFO = 0
<span class="comment">*</span><span class="comment">
</span> IF( NL.LT.1 ) THEN
INFO = -1
ELSE IF( NR.LT.1 ) THEN
INFO = -2
ELSE IF( ( SQRE.NE.1 ) .AND. ( SQRE.NE.0 ) ) THEN
INFO = -3
END IF
<span class="comment">*</span><span class="comment">
</span> N = NL + NR + 1
M = N + SQRE
<span class="comment">*</span><span class="comment">
</span> IF( LDU.LT.N ) THEN
INFO = -10
ELSE IF( LDVT.LT.M ) THEN
INFO = -12
ELSE IF( LDU2.LT.N ) THEN
INFO = -15
ELSE IF( LDVT2.LT.M ) THEN
INFO = -17
END IF
IF( INFO.NE.0 ) THEN
CALL <a name="XERBLA.219"></a><a href="xerbla.f.html#XERBLA.1">XERBLA</a>( <span class="string">'<a name="DLASD2.219"></a><a href="dlasd2.f.html#DLASD2.1">DLASD2</a>'</span>, -INFO )
RETURN
END IF
<span class="comment">*</span><span class="comment">
</span> NLP1 = NL + 1
NLP2 = NL + 2
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Generate the first part of the vector Z; and move the singular
</span><span class="comment">*</span><span class="comment"> values in the first part of D one position backward.
</span><span class="comment">*</span><span class="comment">
</span> Z1 = ALPHA*VT( NLP1, NLP1 )
Z( 1 ) = Z1
DO 10 I = NL, 1, -1
Z( I+1 ) = ALPHA*VT( I, NLP1 )
D( I+1 ) = D( I )
IDXQ( I+1 ) = IDXQ( I ) + 1
10 CONTINUE
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Generate the second part of the vector Z.
</span><span class="comment">*</span><span class="comment">
</span> DO 20 I = NLP2, M
Z( I ) = BETA*VT( I, NLP2 )
20 CONTINUE
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Initialize some reference arrays.
</span><span class="comment">*</span><span class="comment">
</span> DO 30 I = 2, NLP1
COLTYP( I ) = 1
30 CONTINUE
DO 40 I = NLP2, N
COLTYP( I ) = 2
40 CONTINUE
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Sort the singular values into increasing order
</span><span class="comment">*</span><span class="comment">
</span> DO 50 I = NLP2, N
IDXQ( I ) = IDXQ( I ) + NLP1
50 CONTINUE
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> DSIGMA, IDXC, IDXC, and the first column of U2
</span><span class="comment">*</span><span class="comment"> are used as storage space.
</span><span class="comment">*</span><span class="comment">
</span> DO 60 I = 2, N
DSIGMA( I ) = D( IDXQ( I ) )
U2( I, 1 ) = Z( IDXQ( I ) )
IDXC( I ) = COLTYP( IDXQ( I ) )
60 CONTINUE
<span class="comment">*</span><span class="comment">
</span> CALL <a name="DLAMRG.267"></a><a href="dlamrg.f.html#DLAMRG.1">DLAMRG</a>( NL, NR, DSIGMA( 2 ), 1, 1, IDX( 2 ) )
<span class="comment">*</span><span class="comment">
</span> DO 70 I = 2, N
IDXI = 1 + IDX( I )
D( I ) = DSIGMA( IDXI )
Z( I ) = U2( IDXI, 1 )
COLTYP( I ) = IDXC( IDXI )
70 CONTINUE
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Calculate the allowable deflation tolerance
</span><span class="comment">*</span><span class="comment">
</span> EPS = <a name="DLAMCH.278"></a><a href="dlamch.f.html#DLAMCH.1">DLAMCH</a>( <span class="string">'Epsilon'</span> )
TOL = MAX( ABS( ALPHA ), ABS( BETA ) )
TOL = EIGHT*EPS*MAX( ABS( D( N ) ), TOL )
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> There are 2 kinds of deflation -- first a value in the z-vector
</span><span class="comment">*</span><span class="comment"> is small, second two (or more) singular values are very close
</span><span class="comment">*</span><span class="comment"> together (their difference is small).
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> If the value in the z-vector is small, we simply permute the
</span><span class="comment">*</span><span class="comment"> array so that the corresponding singular value is moved to the
</span><span class="comment">*</span><span class="comment"> end.
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> If two values in the D-vector are close, we perform a two-sided
</span><span class="comment">*</span><span class="comment"> rotation designed to make one of the corresponding z-vector
</span><span class="comment">*</span><span class="comment"> entries zero, and then permute the array so that the deflated
</span><span class="comment">*</span><span class="comment"> singular value is moved to the end.
</span><span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> If there are multiple singular values then the problem deflates.
</span><span class="comment">*</span><span class="comment"> Here the number of equal singular values are found. As each equal
</span><span class="comment">*</span><span class="comment"> singular value is found, an elementary reflector is computed to
</span><span class="comment">*</span><span class="comment"> rotate the corresponding singular subspace so that the
</span><span class="comment">*</span><span class="comment"> corresponding components of Z are zero in this new basis.
</span><span class="comment">*</span><span class="comment">
</span> K = 1
K2 = N + 1
DO 80 J = 2, N
IF( ABS( Z( J ) ).LE.TOL ) THEN
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Deflate due to small z component.
</span><span class="comment">*</span><span class="comment">
</span> K2 = K2 - 1
IDXP( K2 ) = J
COLTYP( J ) = 4
IF( J.EQ.N )
$ GO TO 120
ELSE
JPREV = J
GO TO 90
END IF
80 CONTINUE
90 CONTINUE
J = JPREV
100 CONTINUE
J = J + 1
IF( J.GT.N )
$ GO TO 110
IF( ABS( Z( J ) ).LE.TOL ) THEN
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Deflate due to small z component.
</span><span class="comment">*</span><span class="comment">
</span> K2 = K2 - 1
IDXP( K2 ) = J
COLTYP( J ) = 4
ELSE
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Check if singular values are close enough to allow deflation.
</span><span class="comment">*</span><span class="comment">
</span> IF( ABS( D( J )-D( JPREV ) ).LE.TOL ) THEN
<span class="comment">*</span><span class="comment">
</span><span class="comment">*</span><span class="comment"> Deflation is possible.
</span><span class="comment">*</span><span class="comment">
</span> S = Z( JPREV )
⌨️ 快捷键说明
复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?