Static marker error

[Ursula Trinler]

Dear OpenSim community,

Does someone know how to output single error values of all markers (experimental vs. virtual) used in a static trial and not only the overal RMS marker error and the maximum marker error which are given by default in the message window of the GUI using the scaling tool? Is there a simple command I can include in my .xml file to use that in the GUI?

Thank you a lot!
Ursula

[Ursula Trinler]

I found an older post explaining how to receivce the positions of the virtual markers in IK through following command:

<report_marker_locations> true </report_marker_locations>

https://simtk.org/forums/viewtopic.php? ... ons#p10342

So it is easy to calculate the RMS and individual errors for every single marker between virtual and experimental. But how can I do exactly the same thing for the static pose before markers get finally placed (after ModelScaler and before MarkerPlacer)? Is that somehow possible?

Thank you
Ursula

[Ursula Trinler]

I just got the tip from one of the OpenSim guys to only run the scaling part of the scale model tool and then run the IK-tool with the command

<report_marker_locations> true </report_marker_locations>

Through this I can get the marker location on the generic model before the markers get accordingly adjusted to the experimental markers and I can compare them to the experimental markers and calculate myself the marker errors.

Best
Ursula

[jimmy d]

Hi all-

Just to reiterate....

(1) run through the entire scale process (ModelScaler and MarkerPlacer) and get a scaled model with markers registered.
(2) Run IK using this model and the static marker data that was just used to scale your model, adding the tag Ursula mentioned to the IK setup file.

Let us know how that goes

Thanks Ursula
-James

[Ursula Trinler]

James,

It works perfectly! I used for the weighting of the markers in the IK-tool the same markers so the positions of the segments stay the same (normaly I include in the IK task of the scaling-tool different markers than in the IK-tool). If not the position of the model changes slightly.

Best
Ursula