Code Generation Workflows
A Modelica model is a description of the physics. These tutorials show what happens after Rumoca turns that description into code: how the same model becomes a JAX module you can differentiate through, a SymPy expression you can analyze by hand, a CasADi graph you can hand to a nonlinear solver, or a SymForce program you can run on an embedded estimator.
Each tutorial embeds a full Jupyter notebook with rendered outputs, and the Open in Colab badge launches the notebook for live execution — no local Python install required.
Parameter Identification with JAX
Fit unknown physical parameters by gradient descent through a differentiable rollout.
JAX + diffrax
Symbolic Linearization with SymPy
Build the symbolic right-hand side, differentiate to closed-form
A,
B, and inspect controllability analytically.
SymPy
Nonlinear MPC with CasADi
Drop the generated dynamics into a direct-transcription MPC and solve with IPOPT for an aggressive fixed-wing maneuver.
CasADi + IPOPT
On-Manifold EKF with SymForce
Fuse IMU and GPS with an attitude estimator that lives on SO(3) rather than Euler angles.
SymForce