forum.alglib.net

I was wanting to know if ALGLIB distinguishes between row vectors and column vectors. For instance if
a is a row vector
b is a col vector

a*b is a scalar (also the dot product)

b*a is a square matrix.

how would you do a computation like b*a to produce a matrix in ALGLIB?

Strictly speaking, ALGLIB works with arrays, 1-dimensional or 2-dimensional - not vectors (or matrices). It means that ALGLIB has no overloaded operators to performs calculations like "a:=b+3*c" with vectors or matrices. It concentrates on higher-lever operations, like matrix decompositions.

If you want, you may use RMatrixGEMM to calculate product of two matrices - Mx1 and 1xN, but this function is optimized for large operands. It is not optimized to calculate such "thin", low-rank products, so its performance may be several times lower than that of two nested loops without any optimizations at all.

Sergey,

Thanks for the reply. It seems that it will be simpler (and more efficient) to write my own outer product function for 2 vectors.

Regards,