Page 1 of 1

STKO’s Domain Reduction Method

Posted: Tue Aug 05, 2025 6:59 am
by Zouatine
Dear All,

I am currently working on implementing the Domain Reduction Method (DRM) available in STKO, as demonstrated in the webinar. This approach requires an H5DRM file generated using the ShakerMaker module, which provides a set of input motions to excite the model.

I have attempted multiple times to install ShakerMaker within an Anaconda-created environment. I also tried running the code in Visual Studio Code, following the installation instructions provided on GitLab. However, I consistently encounter issues when executing the command:

sudo python setup.py install

The error appears to be related to the Fortran compiler. Since ShakerMaker contains components written in Fortran, it seems these parts are not functioning correctly. When I run the installation command, I receive the following error:

Would you assist me to resolve this issue ?

+ Can i use the DRM in STKO with another input type ? i mean like using a defined normal time serie ?

Thank you,

Re: STKO’s Domain Reduction Method

Posted: Mon Aug 25, 2025 8:47 am
by STKO Team
Dear user,
This is not an STKO-related question, so we cannot give you extensive help on this topic.

However, not that compiling shakermaker in python requires the F2PY module. It's a module used to compile fortran code and interface it with python.

As you can read in this documentation https://numpy.org/doc/stable/f2py/windows/index.html, Anaconda has F2PY with the MSYS2 MinGW compilers, which are outdated and not recommended.

I would suggest you to install Intel One API compilers and do it from your native python (not Anaconda), but be prepared because you need some experience with compiling code

Re: STKO’s Domain Reduction Method

Posted: Wed Aug 27, 2025 3:33 pm
by jaabell
Dear user. Creator of ShakerMaker here.

I do not use Windows with ShakerMaker at all. I know other users have (unsuccesfully) attempted compilation for windows. I haven't tried it myself. If you have success in this regard, I'd love to know about it. Please report as an issue on the project github.

Thank you!

Re: STKO’s Domain Reduction Method

Posted: Fri Sep 05, 2025 12:47 pm
by Zouatine
Thank you for your response.

I think the easiest option for me was to use Linux; however, it was still challenging to find compatible versions of the other modules with ShakerMaker.

After getting ShakerMaker to run, I tried to replicate what was presented in the webinar, I still ran into some issues. Furthermore, when I tried to run the TCL files by copying the Solver and Santiago GM dataset, the model still did not run.

I have also opened this issue in another section of the forum:

viewtopic.php?t=3270

Can you please take a look on the model ?

Thank you,

Re: STKO’s Domain Reduction Method

Posted: Fri Sep 12, 2025 10:03 am
by STKO Team