[Scilab-users] Computing Performance and debugging in Scilab.

[Berns Buenaobra]

Hello TVit:

This seemed to me almost saying we reverse engineer the code? I do no know
if the Open Source agreement will even allow it. Decompiling it surely will
bring you to this level of detail but whats the rationale? You could do
time profile your routines like in the olden days where you can out CPU
clocks like a software a stop watch. Perhaps that should be better -
however, I don't know of any way of code profiling functions in Scilab
right now?

I think the developer are looking more and more on high level development
and its always better - decoupled, object oriented, parallel, polymorphic
etc.

This is my personal point view of it.

Regards,
Berns B.
USC Physics


On Tue, Mar 19, 2013 at 3:05 PM, TViT <tvitklg at rambler.ru> wrote:

> 1) Say to me please, as in Scilab to see quantity(amount) of operations of
> add mul and others operations which the function calculates. For example
> how
> assembler listing or logs can Scilab create?
>
> It is necessary to understand speed of algorithm and through what functions
> it is possible to increase speed.
>
> Can there is for this purpose a function - counter, to call it before
> performance of tested algorithm and after and to compare value of the
> counter of commands it is possible so to make in Scilab?
>
> 2) How to debug my algorithms it is possible to see on lines to watch(keep
> up) on steps calculations which the function carries out. It for
> understanding where a mistake and how much actions are made at calculation.
>
>
>
> --
> View this message in context:
> http://mailinglists.scilab.org/Computing-Performance-and-debugging-in-Scilab-tp4026287.html
> Sent from the Scilab users - Mailing Lists Archives mailing list archive
> at Nabble.com.
> _______________________________________________
> users mailing list
> users at lists.scilab.org
> http://lists.scilab.org/mailman/listinfo/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.scilab.org/pipermail/users/attachments/20130319/db5ffb43/attachment.htm>