djadrp(3P) Sun Performance Library djadrp(3P)NAMEdjadrp - right permutation of a jagged diagonal matrix
SYNOPSIS
SUBROUTINE DJADRP( TRANSP, M, K, VAL, INDX, PNTR, MAXNZ,
* IPERM,WORK,LWORK)
INTEGER TRANSP, M, K, MAXNZ, LWORK
INTEGER INDX(*), PNTR(MAXNZ+1), IPERM(K), WORK(LWORK)
DOUBLE PRECISION VAL(*)
SUBROUTINE DJADRP_64( TRANSP, M, K, VAL, INDX, PNTR, MAXNZ,
* IPERM,WORK,LWORK)
INTEGER*8 TRANSP, M, K, MAXNZ, LWORK
INTEGER*8 INDX(*), PNTR(MAXNZ+1), IPERM(K), WORK(LWORK)
DOUBLE PRECISION VAL(*)
F95 INTERFACE
SUBROUTINE JADRP( TRANSP, M, K, VAL, INDX, PNTR, MAXNZ,
* IPERM, [WORK], [LWORK] )
INTEGER TRANSP, M, K, MAXNZ
INTEGER, DIMENSION(:) :: INDX, PNTR, IPERM
DOUBLE PRECISION, DIMENSION(:) :: VAL
SUBROUTINE JADRP_64( TRANSP, M, K, VAL, INDX, PNTR, MAXNZ,
* IPERM, [WORK], [LWORK] )
INTEGER*8 TRANSP, M, K, MAXNZ
INTEGER*8, DIMENSION(:) :: INDX, PNTR, IPERM
DOUBLE PRECISION, DIMENSION(:) :: VAL
C INTERFACE
#include <sunperf.h>
void djadrp (const int transp, const int m, const int k, const double*
val, const int* indx, int* pntr, const int maxnz, const int*
iperm);
void djadrp_64 (const long transp, const long m, const long k, const
double* val, const long* indx, long* pntr, const long maxnz,
const long* iperm);
DESCRIPTIONdjadrp performs one of the matrix-matrix operations
A <- A P or A <- A P'
( ' indicates matrix transpose)
where A is an M-by-K sparse matrix represented in the jagged
diagonal format, the permutation matrix P is represented by an
integer vector IPERM, such that IPERM(I) is equal to the position
of the only nonzero element in row I of permutation matrix P.
NOTE: In order to get a symetrically permuted jagged diagonal
matrix P A P', one can explicitly permute the columns P A by
calling
DJADRP(0, M, M, VAL, INDX, PNTR, MAXNZ, IPERM, WORK, LWORK)
where parameters VAL, INDX, PNTR, MAXNZ, IPERM are the representation
of A in the jagged diagonal format. The operation makes sense if
the original matrix A is square.
ARGUMENTSTRANSP(input) TRANSP indicates how to operate with the permutation
matrix:
0 : operate with matrix
1 : operate with transpose matrix
Unchanged on exit.
M(input) On entry, M specifies the number of rows in
the matrix A. Unchanged on exit.
K(input) On entry, K specifies the number of columns
in the matrix A. Unchanged on exit.
VAL(input/output) On entry, VAL is a scalar array of length
NNZ=PNTR(MAXNZ+1)-PNTR(1)+1 consisting of entries of A.
VAL can be viewed as a column major ordering of a
row permutation of the Ellpack representation of A,
where the Ellpack representation is permuted so that
the rows are non-increasing in the number of nonzero
entries. Values added for padding in Ellpack are
not included in the Jagged-Diagonal format.
On exit, VAL contains non-zero entries
of the output permuted jagged diagonal matrix.
INDX(input/output) On entry, INDX is an integer array of length
NNZ=PNTR(MAXNZ+1)-PNTR(1)+1 consisting of the column
indices of the corresponding entries in VAL.
On exit, INDX is is overwritten by the column indices
of the output permuted jagged diagonal matrix.
PNTR(input) On entry, PNTR is an integer array of length
MAXNZ+1, where PNTR(I)-PNTR(1)+1 points to
the location in VAL of the first element
in the row-permuted Ellpack represenation of A.
Unchanged on exit.
MAXNZ(input) On entry, MAXNZ specifies the max number of
nonzeros elements per row. Unchanged on exit.
IPERM(input) On entry, IPERM is an integer array of length K
such that I = IPERM(I').
Array IPERM represents a permutation P, such that
IPERM(I) is equal to the position of the only nonzero
element in row I of permutation matrix P.
For example, if
⎪ 0 0 1 ⎪
P =⎪ 1 0 0 ⎪
⎪ 0 1 0 ⎪
then IPERM = (3, 1, 2). Unchanged on exit.
WORK(workspace) Scratch array of length LWORK. LWORK should be at
least K.
LWORK(input) On entry, LWORK specifies the length of the array WORK.
If LWORK=0, the routine is to allocate workspace needed.
If LWORK = -1, then a workspace query is assumed;
the routine only calculates the optimal size of the
array WORK, returns this value as the first entry of
the WORK array, and no error message related to LWORK
is issued by XERBLA.
SEE ALSO
Libsunperf SPARSE BLAS is parallelized with the help of OPENMP and it
is fully compatible with NIST FORTRAN Sparse Blas but the sources are
different. Libsunperf SPARSE BLAS is free of bugs found in NIST FOR‐
TRAN Sparse Blas. Besides several new features and routines are imple‐
mented.
NIST FORTRAN Sparse Blas User's Guide available at:
http://math.nist.gov/mcsd/Staff/KRemington/fspblas/
Based on the standard proposed in
"Document for the Basic Linear Algebra Subprograms (BLAS) Standard",
University of Tennessee, Knoxville, Tennessee, 1996:
http://www.netlib.org/utk/papers/sparse.ps
3rd Berkeley Distribution 6 Mar 2009 djadrp(3P)
