Error in sundialsTB/kinsol

[Michel Kern]

Hello,

there seems to be an error in the Matlab interface to kinsol, in file
kimOpts.c. The value of EtaForm is ignored, and always taken as 1.
The mistake is in lines 129-131, which say:
    if(strcmp(bufval,"Type1"))         *etachoice = KIN_ETACHOICE1;
    else if(strcmp(bufval,"Type2"))    *etachoice = KIN_ETACHOICE2;
    else if(strcmp(bufval,"Constant")) *etachoice = KIN_ETACONSTANT;
whereas the test should be if "(!strcmp(...."  (as is done elsewhere in
the file). This is for version 2.4.

Thanks for the nice software anyway,
Regards,
Michel

--
Michel Kern
INRIA-Rocquencourt, B.P. 105, F-78153 Le Chesnay Cedex.
tel. : [phone], fax : [phone]
e-mail : Michel.Kern@...

[Alan Hindmarsh]

Hello Michel,

Sorry to be so late with this reply.

You are quite right.  We have made this fix here, so it will be
in the next release.

-Alan H



On Thu, 13 Aug 2009, Michel Kern wrote:

> Hello,
>
> there seems to be an error in the Matlab interface to kinsol, in file
> kimOpts.c. The value of EtaForm is ignored, and always taken as 1.
> The mistake is in lines 129-131, which say:
>     if(strcmp(bufval,"Type1"))         *etachoice = KIN_ETACHOICE1;
>     else if(strcmp(bufval,"Type2"))    *etachoice = KIN_ETACHOICE2;
>     else if(strcmp(bufval,"Constant")) *etachoice = KIN_ETACONSTANT;
> whereas the test should be if "(!strcmp(...."  (as is done elsewhere in
> the file). This is for version 2.4.
>
> Thanks for the nice software anyway,
> Regards,
> Michel
>
> --
> Michel Kern
> INRIA-Rocquencourt, B.P. 105, F-78153 Le Chesnay Cedex.
> tel. : [phone], fax : [phone]
> e-mail : Michel.Kern@...
>
>
>