In DGEMM http://www.netlib.org/lapack/explore-html/d1/d54/group__double__blas__level3_gaeda3cbd99c8fb834a60a6412878226e1.html#gaeda3cbd99c8fb834a60a6412878226e1 one can do C = alpha * A.B + beta*C

In eimsum, there is ij,jk->ik (or more general indices) for A.B. Is there any efficient way to incorporate alpha and beta*C, presumbly A and B are general-rank arrays? or Do I need to write several lines of operations with np.dot for alpha,beta and np.add for beta*C?

I tried

na = nb = nc = nd = 50
A = np.random.random((na,nb))
B = np.random.random((nb,nc))
C = np.random.random((na,nc))
np.einsum('ab,bc + ac ->ac',A,B)

got ValueError: invalid subscript '+' in einstein sum subscripts string, subscripts must be letters

It seems scipy.linalg.blas.dgemm can do it, but only for rank-2 array https://docs.scipy.org/doc/scipy/reference/generated/scipy.linalg.blas.dgemm.html