in reply to (jeffa) Re: vector math
in thread vector math
In fact, x in PDL is overloaded to achieve matrix multiplication, so the left- and right-hand operands need to be appropriately dimensioned (it will error out in a hopefully-informative way if not). But * operates on an element-by-element basis (using "broadcasting") over vectors, thus achieving part of a dot-product (there is in fact an inner operation that does an actual dot-product).