Hi!
First and foremost, Thank you all for your work/support/community in symengine!
At our institute we all use proprietary software for symbolic computation. This particular one has a horrible syntax. Sympy as a solution to price and fitness would be great, would there not be the major drawback of speed.
Symengine to the rescue. I calculated the forward dynamics of a 7-Axis Robot manipulator using symengine and it's matrix functions (most notably jacobian). This is great and in terms of speed well comparable to the proprietary solution we would otherwise use! Both implementations run in at around 2min.
My question is twofold,
for day-to-day use I need 2 more thing out of my current "uneducated" workflow:
1) Display in Jupyter: at the moment I convert all my expressions to sympy and display those, but this often takes long - this would be one bottleneck I can live with, because for very long expressions, displaying is very hard anyway.
2) Simplify trigonometric expressions. In mechanics you have a lot of sin**2 + cos**2 expressions and other trigonometric identities which not only simplify expressions, but also make numeric evaluation faster and more precise. I found a trig_simp() function in symengine's code, but I cannot seem to think of a way to expose it to sympy. Can you tell me, if this is implemented yet, or at "worst" guide me how to extend the python wrapper to get this function into my calculation?
Thanks again,
Mike