Symmetry of muscle-tendon parameters

[MSShourijeh]

Depending on the type and number of trials the user has, s/he may want to assume a bilateral symmetry for activation dynamics and/or muscle-tendon parameters. This can potentially reduce the quality of joint moment matching in #MTP, which needs to be clearly explained to the user.

Assume the total number of design variables is n, and the user wants to make m parameters symmetric. There are at least two ways to implement this: 

1. A slow but clean approach 
   The optimizer will still tweak n parameters where the actual number of parameters is n-m. The m number of parameters are overwritten from the other side at each cost function evaluation.

2. A fast but relatively messy approach 
   For this approach, we need functionality that compresses n design variables to n-m so that the parameters assumed to be symmetric between the two sides are optimized once. Would we prefer one side over the other, i.e., would we optimize m parameters for one side and iteratively impose those values on the contralateral side? Alternatively, we optimize both sides together so that m parameters are found such that the average error of both sides is minimized.