Closed Lars-DHondt-KUL closed 2 months ago
1) Changes to existing settings -> I agree with all the changes and I think that they make more sense now 2) New settings -> I also agree and I think that they are useful settings 3) Cleaner main script -> perfect!
I have added small comments to some lines.
I have not run the code, as I see that Dhruv ran it without problems.
1) Changes to existing settings
S.solver.run_as_batch_job
(false)S.OpenSimADOptions.verbose_mode
(false)S.solver.CasADi_path
to usecasadi.GlobalOptions.getCasadiPath()
.save_folder
from subject to misc.savename
andresult_filename
from post_processing to misc.IG_selection
andIG_selection_gaitCyclePercent
from subject to solver.S.subject.v_pelvis_x_trgt
toS.misc.forward_velocity
.gaitmotion_type
) are in misc. Changed name because: 1) we impose the velocity on the forward motion of the floating base, this does not have to be pelvis; 2) trgt (target) could mean that the velocity is in the objective ( $(v-v_{tgt})^2$ ), while this is not the case.S.weights.e_arm
toS.weights.e_torqAct
.S.subject.spasticity
andS.subject.muscle_coordination
.S.misc.main_path
in initializeSettings instead of in main.2) New settings
[]
[]
3) Cleaner main script
run_this_file_to_plot_figures
. Previously, running run_this_file_to_plot_figures multiple times could give problems because the array with results was not cleared between runs. If this script does not have to be called from main, we can always clear the array.To a new function,
runPredSim
. This way, users always call the same function, regardless of using batch jobs or not. This also avoids the possibility of callingadd_pred_sim_to_batch
withS.solver.run_as_batch_job = false
(this could cause conflicts).How Has This Been Tested?
Ran main script and compared result to example result.
Suggested tests for reviewers
Run main script.
To do
I will update the example results, to reflect the changes to the settings. I will do this after the changes to settings are approved.