Jason K. Moore
Jason K. Moore
I forgot this detail to get the fastest speeds and remove as much overhead: ``` python import numpy as np import sympy as sm from pydy.models import n_link_pendulum_on_cart from pydy.codegen.ode_function_generators...
For 5 link pendulum if you solve for x' symoblically you get speed ups (`F**`): ``` In [2]: %timeit f(x, t, r, p) The slowest run took 16.94 times longer...
This info should be added here: https://github.com/pydy/pydy/wiki/PyDy-Speed These matrices are not sparse, they are dense. Sparse solvers are generally slower given a dense matrix. They typically use slower iterative methods....
> Isn't the difference between a minimum mass matrix formulation and a full mass matrix formulation of the same system that the full mass matrix is block diagonal, and one...
Also, you should report any performance issues with SymPy to the issue tracker on that repo. We will look into it. We can also add a benchmark to our benchmarking...
@sixpearls Just came across this: https://github.com/bjodah/pyodesys, which you may find more suitable than PyDy's System class.
Please update with master so the builds stop failing.
As of SciPy 2016 ipywidgets 5 doesn't yet seem to be in the official anaconda repos. But it is on pypi.
This is a duplicate of #133 I believe.
Thanks for that. Yes, I think you may want to have two visualization in one notebook. You may create two models in a notebook or two simulations of one model,...