Johan Carlson wrote:
>
>
<snip, speed of ML vs speed of C discussion...
Of course there is a middle way. If you code your bottlenecks as
C-MEX functions, you get the benefit of hand-coded speed where it
matters without losing the creature comforts of ML's working
environment.
C-MEX functions aren't difficult to get into. The entry barrier is
pretty low and the rewards are substantial.
|
|
