Found here: https://github.com/sympy/sympy/blob/master/sympy/simplify/_cse_diff.py
It is a simple swap to use it and it is include in SymPy 1.14. I tried it and all the opty tests passed. But the examples that use the gait2d model failed to differentiate. I don't think the assumptions of the original symbols are taken into account because I get this error:
/home/moorepants/miniconda/envs/opty-dev/lib/python3.14/site-packages/sympy/printing/codeprinter.py(610)_print_not_supported()
608 def _print_not_supported(self, expr):
609 if self._settings.get('strict', False):
--> 610 raise PrintMethodNotImplementedError(
611 f"Unsupported by {type(self)}: {type(expr)}" +
612 "\nSet the printer option 'strict' to False in order to generate partially printed code."
ipdb> expr
re(x102)
ipdb>
showing that the derivative of the foot penetration force is being treated as complex.
The Jacobian in cse_diff.py may be faster than the one in opty.
Found here: https://github.com/sympy/sympy/blob/master/sympy/simplify/_cse_diff.py
It is a simple swap to use it and it is include in SymPy 1.14. I tried it and all the opty tests passed. But the examples that use the gait2d model failed to differentiate. I don't think the assumptions of the original symbols are taken into account because I get this error:
showing that the derivative of the foot penetration force is being treated as complex.
The Jacobian in cse_diff.py may be faster than the one in opty.