Re: Order of matrix multiplication

[siko1056]

Mike Miller-4 wrote
> Agreed. IMHO mtimes and the * operator would retain the expected
> left-associativity property, but another function could be introduced
> that determines a different order to perform pairwise multiplications
> for performance.

I think many Octave users know from courses about linear algebra, that the
order of operations matters (e.g. painful pen&paper computation of outer
product times vector w'*w*x or w'*(w*x)) and are able to identify such
inappropriate operation order, like in the initial example.

I agree that Octave can have linter integrated inside the editor (like with
Matlab's checkcode function) to suggest such and many other optimizations
(vectorizing statements, ...). But I would not like to apply optimizations
by default. Maybe some teaching person likes to visualize this effect, or
simply relies on the order of operations.

Kai



--
View this message in context: 
http://octave.1599824.n4.nabble.com/Order-of-matrix-multiplication-tp4679715p4679729.html
Sent from the Octave - General mailing list archive at Nabble.com.