OpenBLAS 与 LAPACK快速上手

1. OpenBLAS

1.1 Build

make -j4
make install PREFIX=xxx

1.2 References

• Apple BLAS API Reference
• MKL-BLAS and Sparse BLAS Routines
• Numpy-Documentation
• BoostBLAS-ublas

1.3 Examples

• example.py

import numpy as np

M, N, K, alpha, beta = 3, 3, 2, 1.0, 0.0
A = np.array([1.0,2.0,1.0,-3.0,4.0,-1.0]).reshape(M, K)
B = np.array([1.0,2.0,1.0,-3.0,4.0,-1.0]).reshape(K, N)
C = np.array([.5,.5,.5,.5,.5,.5,.5,.5,.5]).reshape(M, N)

print alpha * np.dot(A, B) + beta * C

• example.c

#include 
#include 

void main() {

    int i = 0;
    double A[6] = {1.0,2.0,1.0,-3.0,4.0,-1.0};         
    double B[6] = {1.0,2.0,1.0,-3.0,4.0,-1.0};  
    double C[9] = {.5,.5,.5,.5,.5,.5,.5,.5,.5}; 

    int M = 3; // row of A and C
    int N = 3; // col of B and C
    int K = 2; // col of A and row of B

    double alpha = 1.0;
    double beta = 0.0;

    cblas_dgemm(CblasRowMajor, CblasNoTrans, CblasNoTrans, M, N, K, alpha, A, K, B, N, beta, C, N);

    for (i = 0; i < 9; i++) {
        printf("%lf ", C[i]);
    }
    printf("\n");
}

$ gcc -o example example.c -I ./include -L ./lib -lopenblas -lpthread -lgfortran

$ python example.py 
[[ -5.  10.  -1.]
 [ 10. -10.   4.]
 [  7.   4.   5.]]

$ ./example 
-5.000000 10.000000 -1.000000 10.000000 -10.000000 4.000000 7.000000 4.000000 5.000000

---

2. LAPACK

2.1 Build

mkdir build; cd build
cmake -DCMAKE_INSTALL_PREFIX=/home/zhubolong/openblas/lapack-3.7.0-release ..
make -j4
make test
make install


cp LAPACK/make.inc.example LAPACK/make.inc
cd LAPACKE
make lapacke

2.2 Reference

• LAPACK - Linear Algebra PACKage
• LAPACK Users’ Guide 3rd
• LAPACKE C Interface to LAPACK

2.3 Examples

• example.py

import numpy as np

M, N, K = 3, 2, 5
A = np.array([1,1,1,2,3,4,3,5,2,4,2,5,5,4,3]).reshape(K, M)
B = np.array([-10,-3,12,14,14,12,16,16,18,16]).reshape(K, N)

print np.linalg.lstsq(A, B)

• example.c

#include 
#include 

int main (int argc, const char * argv[]) {
    double a[5][3] = {1,1,1,2,3,4,3,5,2,4,2,5,5,4,3};
    double b[5][2] = {-10,-3,12,14,14,12,16,16,18,16};
    lapack_int info,m,n,lda,ldb,nrhs;
    int i,j;

    m = 5;
    n = 3;
    nrhs = 2;
    lda = 3;
    ldb = 2;

    info = LAPACKE_dgels(LAPACK_ROW_MAJOR,'N',m,n,nrhs,*a,lda,*b,ldb);

    for(i=0;ifor(j=0;jprintf("%lf ",b[i][j]);
        }
        printf("\n");
    }
    return(info);
}

$ gcc -o example example.c -I ./include -L ./lib -llapacke -llapack -lblas -ltmglib -lpthread -lgfortran -lm

$ python example.py 
(array([[ 2.,  1.],
       [ 1.,  1.],
       [ 1.,  2.]]), array([ 200.,   51.]), 3, array([ 12.31682225,   3.16227766,   1.81545851]))

$ ./example 
2.000000 1.000000 
1.000000 1.000000 
1.000000 2.000000