group() regression after https://codereview.scilab.org/15819
@sgougeon)
Reported by Samuel GOUGEON (BUG DESCRIPTION:
----------------
https://www.mail-archive.com/users@lists.scilab.org/msg08367.html
Hello Yann,
Le 24/02/2019 à 13:46, yann.debray@scilab-enterprises.com a écrit :
> For the signal specialists,
>
> There seems to be a problem with the function group:
> //Group function
> [h w]=wfir('lp',7,[.2,0],'hm',[0.01,-1]);
> [tg,fr]=group(100,h);
> plot2d(fr',tg',-1,'011',' ',[0,2,0.5,4.])
>
> --> [tg,fr]=group(100,h);
> à la ligne 97 de la fonction group ( C:\Program Files\scilab-6.0.2\modules\signal_processing\macros\group.sci ligne 109 )
For the record (and the report): Scilab 5.5.1 was still OK. The bug appeared in Scilab 5.5.2.
Regards
Samuel
---------
Analysis:
When fixing the bug 13205 @ https://codereview.scilab.org/13530,
the condition to select discrete systems was too selective.
This was reported as the bug 13692. However, the fix @
https://codereview.scilab.org/15819 was too intrusive and
made abusive changes, mainly ignoring the hcs=1; initialization.
Restoring group.sci commited @ https://codereview.scilab.org/13530
and just fixing the condition for the selection of discrete systems
fixes this bug.
ERROR LOG:
----------
See above
HOW TO REPRODUCE THE BUG:
-------------------------
[h, w] = wfir('lp',7,[.2,0],'hm',[0.01,-1]);
[tg,fr] = group(100,h);
OTHER INFORMATION:
------------------
Since 5.5.2 and still in 6.0.2
Edited by Vincent COUVERT