SUBROUTINE DLATRS( UPLO, TRANS, DIAG, NORMIN, N, A, LDA, X, SCALE, $ CNORM, INFO ) * * -- LAPACK auxiliary routine (version 3.1) -- * Univ. of Tennessee, Univ. of California Berkeley and NAG Ltd.. * November 2006 * * .. Scalar Arguments .. CHARACTER DIAG, NORMIN, TRANS, UPLO INTEGER INFO, LDA, N DOUBLE PRECISION SCALE * .. * .. Array Arguments .. DOUBLE PRECISION A( LDA, * ), CNORM( * ), X( * ) * .. * * Purpose * ======= * * DLATRS solves one of the triangular systems * * A *x = s*b or A'*x = s*b * * with scaling to prevent overflow. Here A is an upper or lower * triangular matrix, A' denotes the transpose of A, x and b are * n-element vectors, and s is a scaling factor, usually less than * or equal to 1, chosen so that the components of x will be less than * the overflow threshold. If the unscaled problem will not cause * overflow, the Level 2 BLAS routine DTRSV is called. If the matrix A * is singular (A(j,j) = 0 for some j), then s is set to 0 and a * non-trivial solution to A*x = 0 is returned. * * Arguments * ========= * * UPLO (input) CHARACTER*1 * Specifies whether the matrix A is upper or lower triangular. * = 'U': Upper triangular * = 'L': Lower triangular * * TRANS (input) CHARACTER*1 * Specifies the operation applied to A. * = 'N': Solve A * x = s*b (No transpose) * = 'T': Solve A'* x = s*b (Transpose) * = 'C': Solve A'* x = s*b (Conjugate transpose = Transpose) * * DIAG (input) CHARACTER*1 * Specifies whether or not the matrix A is unit triangular. * = 'N': Non-unit triangular * = 'U': Unit triangular * * NORMIN (input) CHARACTER*1 * Specifies whether CNORM has been set or not. * = 'Y': CNORM contains the column norms on entry * = 'N': CNORM is not set on entry. On exit, the norms will * be computed and stored in CNORM. * * N (input) INTEGER * The order of the matrix A. N >= 0. * * A (input) DOUBLE PRECISION array, dimension (LDA,N) * The triangular matrix A. If UPLO = 'U', the leading n by n * upper triangular part of the array A contains the upper * triangular matrix, and the strictly lower triangular part of * A is not referenced. If UPLO = 'L', the leading n by n lower * triangular part of the array A contains the lower triangular * matrix, and the strictly upper triangular part of A is not * referenced. If DIAG = 'U', the diagonal elements of A are * also not referenced and are assumed to be 1. * * LDA (input) INTEGER * The leading dimension of the array A. LDA >= max (1,N). * * X (input/output) DOUBLE PRECISION array, dimension (N) * On entry, the right hand side b of the triangular system. * On exit, X is overwritten by the solution vector x. * * SCALE (output) DOUBLE PRECISION * The scaling factor s for the triangular system * A * x = s*b or A'* x = s*b. * If SCALE = 0, the matrix A is singular or badly scaled, and * the vector x is an exact or approximate solution to A*x = 0. * * CNORM (input or output) DOUBLE PRECISION array, dimension (N) * * If NORMIN = 'Y', CNORM is an input argument and CNORM(j) * contains the norm of the off-diagonal part of the j-th column * of A. If TRANS = 'N', CNORM(j) must be greater than or equal * to the infinity-norm, and if TRANS = 'T' or 'C', CNORM(j) * must be greater than or equal to the 1-norm. * * If NORMIN = 'N', CNORM is an output argument and CNORM(j) * returns the 1-norm of the offdiagonal part of the j-th column * of A. * * INFO (output) INTEGER * = 0: successful exit * < 0: if INFO = -k, the k-th argument had an illegal value * * Further Details * ======= ======= * * A rough bound on x is computed; if that is less than overflow, DTRSV * is called, otherwise, specific code is used which checks for possible * overflow or divide-by-zero at every operation. * * A columnwise scheme is used for solving A*x = b. The basic algorithm * if A is lower triangular is * * x[1:n] := b[1:n] * for j = 1, ..., n * x(j) := x(j) / A(j,j) * x[j+1:n] := x[j+1:n] - x(j) * A[j+1:n,j] * end * * Define bounds on the components of x after j iterations of the loop: * M(j) = bound on x[1:j] * G(j) = bound on x[j+1:n] * Initially, let M(0) = 0 and G(0) = max{x(i), i=1,...,n}. * * Then for iteration j+1 we have * M(j+1) <= G(j) / | A(j+1,j+1) | * G(j+1) <= G(j) + M(j+1) * | A[j+2:n,j+1] | * <= G(j) ( 1 + CNORM(j+1) / | A(j+1,j+1) | ) * * where CNORM(j+1) is greater than or equal to the infinity-norm of * column j+1 of A, not counting the diagonal. Hence * * G(j) <= G(0) product ( 1 + CNORM(i) / | A(i,i) | ) * 1<=i<=j * and * * |x(j)| <= ( G(0) / |A(j,j)| ) product ( 1 + CNORM(i) / |A(i,i)| ) * 1<=i< j * * Since |x(j)| <= M(j), we use the Level 2 BLAS routine DTRSV if the * reciprocal of the largest M(j), j=1,..,n, is larger than * max(underflow, 1/overflow). * * The bound on x(j) is also used to determine when a step in the * columnwise method can be performed without fear of overflow. If * the computed bound is greater than a large constant, x is scaled to * prevent overflow, but if the bound overflows, x is set to 0, x(j) to * 1, and scale to 0, and a non-trivial solution to A*x = 0 is found. * * Similarly, a row-wise scheme is used to solve A'*x = b. The basic * algorithm for A upper triangular is * * for j = 1, ..., n * x(j) := ( b(j) - A[1:j-1,j]' * x[1:j-1] ) / A(j,j) * end * * We simultaneously compute two bounds * G(j) = bound on ( b(i) - A[1:i-1,i]' * x[1:i-1] ), 1<=i<=j * M(j) = bound on x(i), 1<=i<=j * * The initial values are G(0) = 0, M(0) = max{b(i), i=1,..,n}, and we * add the constraint G(j) >= G(j-1) and M(j) >= M(j-1) for j >= 1. * Then the bound on x(j) is * * M(j) <= M(j-1) * ( 1 + CNORM(j) ) / | A(j,j) | * * <= M(0) * product ( ( 1 + CNORM(i) ) / |A(i,i)| ) * 1<=i<=j * * and we can safely call DTRSV if 1/M(n) and 1/G(n) are both greater * than max(underflow, 1/overflow). * * ===================================================================== * * .. Parameters .. DOUBLE PRECISION ZERO, HALF, ONE PARAMETER ( ZERO = 0.0D+0, HALF = 0.5D+0, ONE = 1.0D+0 ) * .. * .. Local Scalars .. LOGICAL NOTRAN, NOUNIT, UPPER INTEGER I, IMAX, J, JFIRST, JINC, JLAST DOUBLE PRECISION BIGNUM, GROW, REC, SMLNUM, SUMJ, TJJ, TJJS, $ TMAX, TSCAL, USCAL, XBND, XJ, XMAX * .. * .. External Functions .. LOGICAL LSAME INTEGER IDAMAX DOUBLE PRECISION DASUM, DDOT, DLAMCH EXTERNAL LSAME, IDAMAX, DASUM, DDOT, DLAMCH * .. * .. External Subroutines .. EXTERNAL DAXPY, DSCAL, DTRSV, XERBLA * .. * .. Intrinsic Functions .. INTRINSIC ABS, MAX, MIN * .. * .. Executable Statements .. * INFO = 0 UPPER = LSAME( UPLO, 'U' ) NOTRAN = LSAME( TRANS, 'N' ) NOUNIT = LSAME( DIAG, 'N' ) * * Test the input parameters. * IF( .NOT.UPPER .AND. .NOT.LSAME( UPLO, 'L' ) ) THEN INFO = -1 ELSE IF( .NOT.NOTRAN .AND. .NOT.LSAME( TRANS, 'T' ) .AND. .NOT. $ LSAME( TRANS, 'C' ) ) THEN INFO = -2 ELSE IF( .NOT.NOUNIT .AND. .NOT.LSAME( DIAG, 'U' ) ) THEN INFO = -3 ELSE IF( .NOT.LSAME( NORMIN, 'Y' ) .AND. .NOT. $ LSAME( NORMIN, 'N' ) ) THEN INFO = -4 ELSE IF( N.LT.0 ) THEN INFO = -5 ELSE IF( LDA.LT.MAX( 1, N ) ) THEN INFO = -7 END IF IF( INFO.NE.0 ) THEN CALL XERBLA( 'DLATRS', -INFO ) RETURN END IF * * Quick return if possible * IF( N.EQ.0 ) $ RETURN * * Determine machine dependent parameters to control overflow. * SMLNUM = DLAMCH( 'Safe minimum' ) / DLAMCH( 'Precision' ) BIGNUM = ONE / SMLNUM SCALE = ONE * IF( LSAME( NORMIN, 'N' ) ) THEN * * Compute the 1-norm of each column, not including the diagonal. * IF( UPPER ) THEN * * A is upper triangular. * DO 10 J = 1, N CNORM( J ) = DASUM( J-1, A( 1, J ), 1 ) 10 CONTINUE ELSE * * A is lower triangular. * DO 20 J = 1, N - 1 CNORM( J ) = DASUM( N-J, A( J+1, J ), 1 ) 20 CONTINUE CNORM( N ) = ZERO END IF END IF * * Scale the column norms by TSCAL if the maximum element in CNORM is * greater than BIGNUM. * IMAX = IDAMAX( N, CNORM, 1 ) TMAX = CNORM( IMAX ) IF( TMAX.LE.BIGNUM ) THEN TSCAL = ONE ELSE TSCAL = ONE / ( SMLNUM*TMAX ) CALL DSCAL( N, TSCAL, CNORM, 1 ) END IF * * Compute a bound on the computed solution vector to see if the * Level 2 BLAS routine DTRSV can be used. * J = IDAMAX( N, X, 1 ) XMAX = ABS( X( J ) ) XBND = XMAX IF( NOTRAN ) THEN * * Compute the growth in A * x = b. * IF( UPPER ) THEN JFIRST = N JLAST = 1 JINC = -1 ELSE JFIRST = 1 JLAST = N JINC = 1 END IF * IF( TSCAL.NE.ONE ) THEN GROW = ZERO GO TO 50 END IF * IF( NOUNIT ) THEN * * A is non-unit triangular. * * Compute GROW = 1/G(j) and XBND = 1/M(j). * Initially, G(0) = max{x(i), i=1,...,n}. * GROW = ONE / MAX( XBND, SMLNUM ) XBND = GROW DO 30 J = JFIRST, JLAST, JINC * * Exit the loop if the growth factor is too small. * IF( GROW.LE.SMLNUM ) $ GO TO 50 * * M(j) = G(j-1) / abs(A(j,j)) * TJJ = ABS( A( J, J ) ) XBND = MIN( XBND, MIN( ONE, TJJ )*GROW ) IF( TJJ+CNORM( J ).GE.SMLNUM ) THEN * * G(j) = G(j-1)*( 1 + CNORM(j) / abs(A(j,j)) ) * GROW = GROW*( TJJ / ( TJJ+CNORM( J ) ) ) ELSE * * G(j) could overflow, set GROW to 0. * GROW = ZERO END IF 30 CONTINUE GROW = XBND ELSE * * A is unit triangular. * * Compute GROW = 1/G(j), where G(0) = max{x(i), i=1,...,n}. * GROW = MIN( ONE, ONE / MAX( XBND, SMLNUM ) ) DO 40 J = JFIRST, JLAST, JINC * * Exit the loop if the growth factor is too small. * IF( GROW.LE.SMLNUM ) $ GO TO 50 * * G(j) = G(j-1)*( 1 + CNORM(j) ) * GROW = GROW*( ONE / ( ONE+CNORM( J ) ) ) 40 CONTINUE END IF 50 CONTINUE * ELSE * * Compute the growth in A' * x = b. * IF( UPPER ) THEN JFIRST = 1 JLAST = N JINC = 1 ELSE JFIRST = N JLAST = 1 JINC = -1 END IF * IF( TSCAL.NE.ONE ) THEN GROW = ZERO GO TO 80 END IF * IF( NOUNIT ) THEN * * A is non-unit triangular. * * Compute GROW = 1/G(j) and XBND = 1/M(j). * Initially, M(0) = max{x(i), i=1,...,n}. * GROW = ONE / MAX( XBND, SMLNUM ) XBND = GROW DO 60 J = JFIRST, JLAST, JINC * * Exit the loop if the growth factor is too small. * IF( GROW.LE.SMLNUM ) $ GO TO 80 * * G(j) = max( G(j-1), M(j-1)*( 1 + CNORM(j) ) ) * XJ = ONE + CNORM( J ) GROW = MIN( GROW, XBND / XJ ) * * M(j) = M(j-1)*( 1 + CNORM(j) ) / abs(A(j,j)) * TJJ = ABS( A( J, J ) ) IF( XJ.GT.TJJ ) $ XBND = XBND*( TJJ / XJ ) 60 CONTINUE GROW = MIN( GROW, XBND ) ELSE * * A is unit triangular. * * Compute GROW = 1/G(j), where G(0) = max{x(i), i=1,...,n}. * GROW = MIN( ONE, ONE / MAX( XBND, SMLNUM ) ) DO 70 J = JFIRST, JLAST, JINC * * Exit the loop if the growth factor is too small. * IF( GROW.LE.SMLNUM ) $ GO TO 80 * * G(j) = ( 1 + CNORM(j) )*G(j-1) * XJ = ONE + CNORM( J ) GROW = GROW / XJ 70 CONTINUE END IF 80 CONTINUE END IF * IF( ( GROW*TSCAL ).GT.SMLNUM ) THEN * * Use the Level 2 BLAS solve if the reciprocal of the bound on * elements of X is not too small. * CALL DTRSV( UPLO, TRANS, DIAG, N, A, LDA, X, 1 ) ELSE * * Use a Level 1 BLAS solve, scaling intermediate results. * IF( XMAX.GT.BIGNUM ) THEN * * Scale X so that its components are less than or equal to * BIGNUM in absolute value. * SCALE = BIGNUM / XMAX CALL DSCAL( N, SCALE, X, 1 ) XMAX = BIGNUM END IF * IF( NOTRAN ) THEN * * Solve A * x = b * DO 110 J = JFIRST, JLAST, JINC * * Compute x(j) = b(j) / A(j,j), scaling x if necessary. * XJ = ABS( X( J ) ) IF( NOUNIT ) THEN TJJS = A( J, J )*TSCAL ELSE TJJS = TSCAL IF( TSCAL.EQ.ONE ) $ GO TO 100 END IF TJJ = ABS( TJJS ) IF( TJJ.GT.SMLNUM ) THEN * * abs(A(j,j)) > SMLNUM: * IF( TJJ.LT.ONE ) THEN IF( XJ.GT.TJJ*BIGNUM ) THEN * * Scale x by 1/b(j). * REC = ONE / XJ CALL DSCAL( N, REC, X, 1 ) SCALE = SCALE*REC XMAX = XMAX*REC END IF END IF X( J ) = X( J ) / TJJS XJ = ABS( X( J ) ) ELSE IF( TJJ.GT.ZERO ) THEN * * 0 < abs(A(j,j)) <= SMLNUM: * IF( XJ.GT.TJJ*BIGNUM ) THEN * * Scale x by (1/abs(x(j)))*abs(A(j,j))*BIGNUM * to avoid overflow when dividing by A(j,j). * REC = ( TJJ*BIGNUM ) / XJ IF( CNORM( J ).GT.ONE ) THEN * * Scale by 1/CNORM(j) to avoid overflow when * multiplying x(j) times column j. * REC = REC / CNORM( J ) END IF CALL DSCAL( N, REC, X, 1 ) SCALE = SCALE*REC XMAX = XMAX*REC END IF X( J ) = X( J ) / TJJS XJ = ABS( X( J ) ) ELSE * * A(j,j) = 0: Set x(1:n) = 0, x(j) = 1, and * scale = 0, and compute a solution to A*x = 0. * DO 90 I = 1, N X( I ) = ZERO 90 CONTINUE X( J ) = ONE XJ = ONE SCALE = ZERO XMAX = ZERO END IF 100 CONTINUE * * Scale x if necessary to avoid overflow when adding a * multiple of column j of A. * IF( XJ.GT.ONE ) THEN REC = ONE / XJ IF( CNORM( J ).GT.( BIGNUM-XMAX )*REC ) THEN * * Scale x by 1/(2*abs(x(j))). * REC = REC*HALF CALL DSCAL( N, REC, X, 1 ) SCALE = SCALE*REC END IF ELSE IF( XJ*CNORM( J ).GT.( BIGNUM-XMAX ) ) THEN * * Scale x by 1/2. * CALL DSCAL( N, HALF, X, 1 ) SCALE = SCALE*HALF END IF * IF( UPPER ) THEN IF( J.GT.1 ) THEN * * Compute the update * x(1:j-1) := x(1:j-1) - x(j) * A(1:j-1,j) * CALL DAXPY( J-1, -X( J )*TSCAL, A( 1, J ), 1, X, $ 1 ) I = IDAMAX( J-1, X, 1 ) XMAX = ABS( X( I ) ) END IF ELSE IF( J.LT.N ) THEN * * Compute the update * x(j+1:n) := x(j+1:n) - x(j) * A(j+1:n,j) * CALL DAXPY( N-J, -X( J )*TSCAL, A( J+1, J ), 1, $ X( J+1 ), 1 ) I = J + IDAMAX( N-J, X( J+1 ), 1 ) XMAX = ABS( X( I ) ) END IF END IF 110 CONTINUE * ELSE * * Solve A' * x = b * DO 160 J = JFIRST, JLAST, JINC * * Compute x(j) = b(j) - sum A(k,j)*x(k). * k<>j * XJ = ABS( X( J ) ) USCAL = TSCAL REC = ONE / MAX( XMAX, ONE ) IF( CNORM( J ).GT.( BIGNUM-XJ )*REC ) THEN * * If x(j) could overflow, scale x by 1/(2*XMAX). * REC = REC*HALF IF( NOUNIT ) THEN TJJS = A( J, J )*TSCAL ELSE TJJS = TSCAL END IF TJJ = ABS( TJJS ) IF( TJJ.GT.ONE ) THEN * * Divide by A(j,j) when scaling x if A(j,j) > 1. * REC = MIN( ONE, REC*TJJ ) USCAL = USCAL / TJJS END IF IF( REC.LT.ONE ) THEN CALL DSCAL( N, REC, X, 1 ) SCALE = SCALE*REC XMAX = XMAX*REC END IF END IF * SUMJ = ZERO IF( USCAL.EQ.ONE ) THEN * * If the scaling needed for A in the dot product is 1, * call DDOT to perform the dot product. * IF( UPPER ) THEN SUMJ = DDOT( J-1, A( 1, J ), 1, X, 1 ) ELSE IF( J.LT.N ) THEN SUMJ = DDOT( N-J, A( J+1, J ), 1, X( J+1 ), 1 ) END IF ELSE * * Otherwise, use in-line code for the dot product. * IF( UPPER ) THEN DO 120 I = 1, J - 1 SUMJ = SUMJ + ( A( I, J )*USCAL )*X( I ) 120 CONTINUE ELSE IF( J.LT.N ) THEN DO 130 I = J + 1, N SUMJ = SUMJ + ( A( I, J )*USCAL )*X( I ) 130 CONTINUE END IF END IF * IF( USCAL.EQ.TSCAL ) THEN * * Compute x(j) := ( x(j) - sumj ) / A(j,j) if 1/A(j,j) * was not used to scale the dotproduct. * X( J ) = X( J ) - SUMJ XJ = ABS( X( J ) ) IF( NOUNIT ) THEN TJJS = A( J, J )*TSCAL ELSE TJJS = TSCAL IF( TSCAL.EQ.ONE ) $ GO TO 150 END IF * * Compute x(j) = x(j) / A(j,j), scaling if necessary. * TJJ = ABS( TJJS ) IF( TJJ.GT.SMLNUM ) THEN * * abs(A(j,j)) > SMLNUM: * IF( TJJ.LT.ONE ) THEN IF( XJ.GT.TJJ*BIGNUM ) THEN * * Scale X by 1/abs(x(j)). * REC = ONE / XJ CALL DSCAL( N, REC, X, 1 ) SCALE = SCALE*REC XMAX = XMAX*REC END IF END IF X( J ) = X( J ) / TJJS ELSE IF( TJJ.GT.ZERO ) THEN * * 0 < abs(A(j,j)) <= SMLNUM: * IF( XJ.GT.TJJ*BIGNUM ) THEN * * Scale x by (1/abs(x(j)))*abs(A(j,j))*BIGNUM. * REC = ( TJJ*BIGNUM ) / XJ CALL DSCAL( N, REC, X, 1 ) SCALE = SCALE*REC XMAX = XMAX*REC END IF X( J ) = X( J ) / TJJS ELSE * * A(j,j) = 0: Set x(1:n) = 0, x(j) = 1, and * scale = 0, and compute a solution to A'*x = 0. * DO 140 I = 1, N X( I ) = ZERO 140 CONTINUE X( J ) = ONE SCALE = ZERO XMAX = ZERO END IF 150 CONTINUE ELSE * * Compute x(j) := x(j) / A(j,j) - sumj if the dot * product has already been divided by 1/A(j,j). * X( J ) = X( J ) / TJJS - SUMJ END IF XMAX = MAX( XMAX, ABS( X( J ) ) ) 160 CONTINUE END IF SCALE = SCALE / TSCAL END IF * * Scale the column norms by 1/TSCAL for return. * IF( TSCAL.NE.ONE ) THEN CALL DSCAL( N, ONE / TSCAL, CNORM, 1 ) END IF * RETURN * * End of DLATRS * END