This email message is a notification to let you know that
a file has been uploaded to the Files area of the anyscript
group.
File : /AnyBody.3.0_files_txt_JavierMarin_2008Jan03.rar
Uploaded by : jjmarinz <jjmarin@unizar.es>
Description : For Søren. Files “txt” with AnyBody ver 3.0
I have downloaded and tried to run the code you uploaded.
As described in Page 184 of the reference manual,
the MinMaxNRSimplex is the simplex algorithm based
on algorithms in the book Numerical Recipes by Press et. al. (You can read
this using this link: http://www.nrbook.com/a/bookcpdf/c10-8.pdf).
In particular, as noted in the book, this method may
encounter endless cycling if input data is not
properly scale - most probably this is what happen
in your case. To overcome this, you can change the
tolerance value (which default value is 1e-6) for
zero testing in the algorithm to a larger value,
say 1e-4. You also need to change the relative zero
testing tolerance use while finding pivot coefficient
(default is 100) to a similar smaller value 1.
This relative zero testing tolerance will scale
the remaining tolerance value for zero testing
in the code correspondingly. Value of 1 set all of them
to 1e-4, value of 100 set all of them to 1e-2 (which
is not desired since it may give inaccurate result).
In order to do so, simply add the following two lines
of codes:
right after line “RecruitmentSolver = MinMaxNRSimplex;”
where you define your recruitment solver in file:
“_Model-LEMUS.Main_ver3.0.any”
Another fix to this problem is to change the
recruitment solver type to “MinMaxSimplex” instead
of “MinMaxNRSimplex” that you currently have.
By doing so, you should be able to run the
inverse dynamic analysis without any problem.
The MinMaxSimplex solver uses the numerical library
OOSol, which is more robust in this case.
This email message is a notification to let you know that
a file has been uploaded to the Files area of the anyscript
group.
File : /AnyInputFile_Demo.zip
Uploaded by : lengfenglee <lengfenglee@yahoo.com>
Description : A short demo to show how AnyInputFile read a .txt file (For
Javier).
Thank you very much for your explications…
We have tested with the new configuration for RecruitmentSolver
variable and we can run inverse dynamic analysis without problems.
Thank you for you help.
Javier Marin (Grupo ID-ERGO. University of Zaragoza)
— In anyscript@yahoogroups.com, “Leng-Feng Lee” <lengfenglee@…>
wrote:
>
> Hi Javier,
>
> I have downloaded and tried to run the code you uploaded.
>
> As described in Page 184 of the reference manual,
> the MinMaxNRSimplex is the simplex algorithm based
> on algorithms in the book Numerical Recipes by Press et. al. (You
can read
> this using this link:
> http://www.nrbook.com/a/bookcpdf/c10-8.pdf).
>
> In particular, as noted in the book, this method may
> encounter endless cycling if input data is not
> properly scale - most probably this is what happen
> in your case. To overcome this, you can change the
> tolerance value (which default value is 1e-6) for
> zero testing in the algorithm to a larger value,
> say 1e-4. You also need to change the relative zero
> testing tolerance use while finding pivot coefficient
> (default is 100) to a similar smaller value 1.
> This relative zero testing tolerance will scale
> the remaining tolerance value for zero testing
> in the code correspondingly. Value of 1 set all of them
> to 1e-4, value of 100 set all of them to 1e-2 (which
> is not desired since it may give inaccurate result).
> In order to do so, simply add the following two lines
> of codes:
>
> Main.Study.NRSimplex.Zero=1e-4;
> Main.Study.NRSimplex.CoefZeroRel=1;
>
> right after line “RecruitmentSolver = MinMaxNRSimplex;”
> - where you define your recruitment solver in file:
> “_Model-LEMUS.Main_ver3.0.any”
>
> Another fix to this problem is to change the
> recruitment solver type to “MinMaxSimplex” instead
> of “MinMaxNRSimplex” that you currently have.
> By doing so, you should be able to run the
> inverse dynamic analysis without any problem.
> The MinMaxSimplex solver uses the numerical library
> OOSol, which is more robust in this case.
>
> Hope this helps.
>
> Thanks,
> Leng-feng
>
>
> — In anyscript@yahoogroups.com, anyscript@yahoogroups.com wrote:
> >
> >
> > Hello,
> >
> > This email message is a notification to let you know that
> > a file has been uploaded to the Files area of the anyscript
> > group.
> >
> > File : /AnyBody.3.0_files_txt_JavierMarin_2008Jan03.rar
> > Uploaded by : jjmarinz <jjmarin@>
> > Description : For Søren. Files “txt” with AnyBody ver 3.0
> >
> > You can access this file at the URL:
> >
> http://groups.yahoo.com/group/anyscript/files/AnyBody.3.0_files_txt_Ja
vierMarin_2008Jan03.rar
>
> >
> > To learn more about file sharing for your group, please visit:
> >
> http://help.yahoo.com/l/us/yahoo/groups/original/members/web/index.htm
lfiles
> >
> > Regards,
> >
> > jjmarinz <jjmarin@>
> >
>