running CMC from command line

Provide easy-to-use, extensible software for modeling, simulating, controlling, and analyzing the neuromusculoskeletal system.
POST REPLY
User avatar
Moe Curtin
Posts: 43
Joined: Fri Oct 02, 2009 5:40 am

running CMC from command line

Post by Moe Curtin » Thu Oct 04, 2012 6:31 am

I have noticed that CMC runs differently from the command line on 2 of my PCs. OpenSim both been installed on both PCs (both running windows 7 64bit) from the same file and I am using the exact same CMC setup files on each.

One runs fine with no problems, but when I start the other running I get a warning for every muscle in my model saying "CMC::Warning: CMC cannot compute controls for muscles with muscle controls < 0.02. The minumum control limit for muscle 'muscle_name' has been reset to 0.02.". The constraints file is definitely there and includes all muscles.

I have tried it on a couple different CMC setups and the same thing happens. I have reinstalled OpenSim, but the same problem occurs. Anyone familiar with this issue?

User avatar
Ayman Habib
Posts: 2235
Joined: Fri Apr 01, 2005 12:24 pm

Re: running CMC from command line

Post by Ayman Habib » Mon Oct 08, 2012 10:12 am

Hi Moe,

The only explanation for this (since the GUI is just another interface to execute the same code) is that the libraries found by the GUI (which doesn't need a PATH to be set) are different from command line executables (these require env. var PATH to point to proper installDir/bin to work).

Did you check that?

Best regards,
-Ayman

User avatar
Moe Curtin
Posts: 43
Joined: Fri Oct 02, 2009 5:40 am

Re: running CMC from command line

Post by Moe Curtin » Fri Oct 19, 2012 6:12 am

Hi Ayman,

It turns out that they must have been installed from different files because the build date is actually different on both of the command windows by a month or so. They are both version 2.4.0. The one from Sept 2011 is working ok, but the one from Oct 2011 is the one that I am having this issue with. Could there be a bug perhaps?

User avatar
Ayman Habib
Posts: 2235
Joined: Fri Apr 01, 2005 12:24 pm

Re: running CMC from command line

Post by Ayman Habib » Sun Oct 21, 2012 5:59 pm

Hi Moe,

Same version (2.4.0) corresponds to same code base. The two installers were likely built with different compilers (Visual Studio version). Normally this makes no difference except in execution speed (due to different optimizations etc.). If you can reproduce this behavior consistently (hopefully from the GUI so that Path setting is not an issue) then please file a bug report and attach all necessary model, data and setup files to help us reproduce the problem.

Can you check the behavior in version 3.0 as well?

Best regards,
-Ayman

User avatar
Moe Curtin
Posts: 43
Joined: Fri Oct 02, 2009 5:40 am

Re: running CMC from command line

Post by Moe Curtin » Mon Oct 22, 2012 3:49 am

I have just installed 3.0 and am running the example files. The message appears for all muscles in every CMC example in the GUI. Is this a problem or can it just be ignored?

User avatar
Ajay Seth
Posts: 136
Joined: Thu Mar 15, 2007 10:39 am

Re: running CMC from command line

Post by Ajay Seth » Tue Oct 23, 2012 9:33 am

In general, you shouldn't ignore the small force range warnings because they indicate where muscles are at some physiological limit (to short or shortening too fast) such that min and max excitation is unable to generate a significant difference (force range). This often happens in the plantarflexors for fast walking where soleus is shortening rapidly at push-off. This is likely occurring more in OpenSim 3.0 since the Thelen2003Muscle was restored to the f-v curve defined in the Thelen 2003 paper, which has lower force generation ability for the same fiber velocity. To verify, you can replace Thelen2003Muscle with Thelen2003Muscle _Deprecated and the results should be virtually identical to 2.4 results.

User avatar
Moe Curtin
Posts: 43
Joined: Fri Oct 02, 2009 5:40 am

Re: running CMC from command line

Post by Moe Curtin » Wed Oct 24, 2012 5:50 am

Hi Ajay,

It isn't a small force range warning. It's acting as if I haven't defined constraints for my muscles and is saying that it it resetting all minimum constraints to 0.02, even though all of them are already set to minimum values of 0.02 or greater.

POST REPLY