BLAS1_D
Basic Linear Algebra Subprograms
Level 1
Double Precision Real Arithmetic

BLAS1_D is a library of C++ routines, using double precision real arithmetic, which implement the Level 1 BLAS, or Basic Linear Algebra Subprograms.

The BLAS are a small core library of linear algebra utilities, which can be highly optimized for various architectures. Software that relies on the BLAS is thus highly portable, and will typically run very efficiently.

The Level 1 BLAS are primarily for use in vector operations. In certain cases, they may also be used to operate on the rows or columns of a two-dimensional array.

On the SGI, a precompiled optimized BLAS library may be linked to via a command like:

g++ myprog.C -lblas

On the Compaq ALPHA, a precompiled optimized BLAS library is available in CXML, the "Compaq Extended Math Library". A C++ program can link with the CXML library, but the CXML routine names must have an underscore appended to them, and they must be declared as extern "C" to avoid the dreaded C++ name-mangling. The linking command would be something like:

g++ myprog.C -lcxml

Related Data and Programs:

BLAS1_C is a C++ library of routines which is a version of BLAS1 for single precision complex arithmetic.

BLAS1_D is also available in a C version and a FORTRAN77 version and a FORTRAN90 version and a MATLAB version.

BLAS1_S is a C++ library of routines which is a version of BLAS1 for single precision real arithmetic.

BLAS1_Z is a C++ library of routines which is a version of BLAS1 for double precision complex arithmetic.

BLAS2 is a FORTRAN90 library of routines which is a version of the Level 2 BLAS, for matrix-vector operations.

BLAS3 is a FORTRAN90 library of routines which is a version of the Level 3 BLAS, for matrix-matrix operations.

GSL is the Gnu Scientific Library which includes an implementation of the BLAS1 routines.

LAPACK is a FORTRAN90 library of routines which is a linear algebra package that uses BLAS1 routines

LINPACK is a C++ library of routines which is a linear algebra library that uses the BLAS1 routines

SUPER_BLAS is a C library of routines which is a version of some of the BLAS1 routines written in C.

Reference:

1. Jack Dongarra, Jim Bunch, Cleve Moler, Pete Stewart,
   LINPACK User's Guide,
   SIAM, 1979,
   ISBN13: 978-0-898711-72-1,
   LC: QA214.L56.
2. Charles Lawson, Richard Hanson, David Kincaid and Fred Krogh,
   Algorithm 539, Basic Linear Algebra Subprograms for Fortran Usage,
   ACM Transactions on Mathematical Software,
   Volume 5, Number 3, September 1979, pages 308-323.
3. Thomas Coleman and Charles Van Loan,
   Handbook for Matrix Computations,
   Society for Industrial and Applied Mathematics,
   3600 University City Science Center,
   Philadelphia, PA 19104-2688.
4. Compaq Computer Corporation,
   Compaq Extended Math Library,
   CXML Reference Guide

Source Code:

• blas1_d.C, the source code.
• blas1_d.H, the include file.
• blas1_d.csh commands to compile the source code.

Examples and Tests:

• blas1_d_prb.C, the calling program.
• blas1_d_prb.csh compiles, links and runs the calling program.
• blas1_d_prb.out the resulting output file.

The ALPHA has a compiled library of BLAS routines available, but they were compiled as FORTRAN routines. In order for a C++ routine to call them, special steps must be taken to avoid "name mangling" and to append an underscore to the name of each FORTRAN routine.

• blas1_d_prb_alpha.C, the calling program.
• blas1_d_prb_alpha.csh compiles the calling program, links it with the CXML library on a Compaq Alpha, and runs the program.
• blas1_d_prb_alpha.out the resulting output file.

List of Routines:

• DASUM takes the sum of the absolute values of a double vector.
• DAXPY computes double constant times a vector plus a vector.
• DCOPY copies a double vector X to a vector Y.
• DDOT forms the dot product of two double vectors.
• DMACH computes machine parameters of double arithmetic.
• DNRM2 returns the euclidean norm of a double vector.
• DROT applies a double plane rotation.
• DROTG constructs a double Givens plane rotation.
• DSCAL scales a double vector by a constant.
• DSWAP interchanges two double vectors.
• I4_MAX returns the maximum of two integers.
• I4_MIN returns the smaller of two integers.
• IDAMAX finds the index of the double vector element of maximum absolute value.
• LSAME returns TRUE if CA is the same letter as CB regardless of case.
• R9_ABS returns the absolute value of a double.
• R8_MAX returns the maximum of two doubles.
• R8_SIGN returns the sign of a double.
• XERBLA is an error handler for the LAPACK routines.

You can go up one level to the C++ source codes.